RI: Medium: Collaborative Research: Physically Grounded Object Recognition Proposal Title: RI: Medium: Collaborative Research: Physically Grounded Object
Recognition
Institution: Carnegie Mellon University
Abstract Date: 05/05/09
This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).
Although the world is very much three dimensional, most of today s approaches to
visual object recognition essentially reduce the problem to one of 2D pattern
classification, where rectangular image patches are independently compared to stored
templates to produce isolated object labels within the image. This project aims to
account for the three dimensional nature of the real world by exploring qualitative
geometric reasoning in terms of 3D spatial relationships between scene components,
category level object models, and global scene understanding.
The project is organized around two major research areas. Qualitative 3D scene
parsing: A central part of our effort will be to develop qualitative 3D models of the scene
that describe the depicted objects and surfaces and their physical relations. Grounding
objects in the scene: We integrate the geometric representation of the scene and the
corresponding 3D spatial relations with the object recognition process by (1) inferring
the set of likely object identities based on 3D relations among scene components; (2)
predicting the most likely object locations from the scene layout; and (3) using the
occlusion relations and depth ordering to predict the parts of objects that may be visible
in the scene.
The project is anticipated to result in major advances in 3D scene understanding from
photographs, a critical enabling technology for a wide range of applications including
autonomous systems, health care, human computer interaction, assistive technology,
image retrieval, industrial and personal robotics, manufacturing, scientific image
analysis, surveillance and security, and transportation.
NATIONAL SCIENCE FOUNDATION
Proposal Abstract
Proposal:0905402 PI Name:Hebert, Martial
Printed from eJacket: 05/06/09 Page 1 of 1 NeTS ProWin: COLLABORATIVE RESEARCH: A new taxonomy for cooperative wireless networking It is well documented how cooperative links can offer considerable performance gains at the physical layer, but it is unclear what kind of network support would be required to attain the sought gains. Cooperative links violate the simple collision model for broadcast transmission, a model that has been instrumental so far in allowing the parallel evolution of communication theory and network theory.
Recognizing the absence of a correct taxonomy to use cooperative links at the network and multiple access layer, the objective of this collaborative project is to investigate theoretically and experimentally the interplay between a cooperative decentralized physical layer and the wireless network architecture as a whole. More specifically, the project will develop viable link abstractions, multiple access protocols, end to end network transport models, appropriate algorithms to support the introduction in wireless mobile networks of two technologies that are rapidly advancing: 1) cooperative transmission, that consists of multiple network nodes operating as a decentralized multi antenna modem and, 2) distributed source coding, that allows the decentralized compression of correlated observations and, thus, is relevant to the design of a decentralized receiver.
The project will also use the GNU software radio platform to test cooperative links and assess their feasibility and degradation when facing real limitations of transceiver synchronization, carrier offset, clock jitter and computation delays.
Algorithms and theoretical results will be disseminated through the standard tools of research publications. Experimental results will be also documented online where the software will be shared to serve as an educational tool as well as to foster new technological advances in mesh networks.
This project will bring future wireless networks closer to achieving the physical limits of communications. AF: Medium: Collaborative Research: The Role of Order in Search Searching for a string in a digital library is a fundamental operation. One may seek a phrase in a file that is being manipulated by a text editor, a sentence on the web via a search engine, or a genome in the human DNA sequence. This basic tool has been well researched and, indeed, is ubiquitous in data processing.

However, string matching and motif discovery algorithms ( Stringology ) that have dealt with searches in general databases, have always assumed that the sought string is ordered. The research thrusts were in directions such as exact matching, approximate matching, or grappling with the challenge of coping with errors in the data. But always the order of the data was assumed to be iron clad.

Nevertheless, some non conforming problems have been gnawing at the walls of this assumption. Some of the areas where applications assume erroneous order of the data are: Text Editing, where errors such as swaps or transpositions, assume that the data has not been changed it has rather been rearranged; Computational Biology, where reversals or transpositions of genome subsequences are part of the evolutionary process; and Linguistics, where the task of lexical categorization is aided by considering sets of parts of speech in various orders.

The above application areas suggest that the wonderfully ordered world of pattern matching may be too rigid to handle new challenges. This project introduces a fundamentally new model of string matching, where the order of symbols in the input pattern may be perturbed, but the content remains unchanged. This model studies unordered pattern matching universes in order to supply the above mentioned application areas with appropriate computational tools. A theory of unordered matching will give a general framework for all the above problems.

This project considers permuted Stringology as a line leading from fully ordered data all the way to data with no order at all. The ordered side has been historically amply researched. The specific technical goals of the project are:

1. To develop a theory and framework for permuted Stringology.
2. To identify the types of unordered problems that are more difficult than ordered Stringology, the conditions that make them harder, and the reason why.
3. To define the term ``Similarity in unordered clusters.
4. To define the world of ``almost ordered sequences , and to compare it with unordered and ordered sequences.
5. To design tools that will be key elements in the solution for the proposed problems, and will be shared by more than one algorithm. The list of problems includes Indexing, Dictionary and Approximations.

This project s aim is to study the fundamental problems arising from a model of pattern matching with rearrangement. The immediate benefit to the field is a totally new, yet very basic, research direction that seems to defy the state of the art toolkit. The historical tools of pattern matching, such as dynamic programming, FFT, sub word trees, renaming, encodings, and embeddings do not seem suitable to handle these problems. New algorithmic tools and data structures are required. In fact, this direction already bore unexpected interesting fruits the solution of an open problem in graph theory posed by the mathematician Cayley in 1849! Techniques such as non standard convolutions, group testing, and graph theoretic algorithms, have been necessary to solve some of the problems thus far.

Since this project defines a new model, there are many directions to explore. It is expected that this project will mark the beginning of intensive research in a new paradigm. RI: Medium: Collaborative Research:From Text to Pictures Last Modified Date: 05/15/09 Last Modified By: Tatiana D. Korelsky

Abstract
This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The researchers are developing new theoretical models and technology to automatically convert descriptive text into 3D scenes representing the text?s meaning. They do this via the Scenario Based Lexical Knowledge Resource (SBLR), a resource they are creating from existing sources (PropBank, WordNet, FrameNet) and from automated mining of Wikipedia and other un annotated text. In addition to predicate argument structure and semantic roles, the SBLR includes necessary roles, typical role fillers, contextual elements, and activity poses which enables analysis of input sentences at a deep level and assembly of appropriate elements from libraries of 3D objects to depict the fuller scene implied by a sentence. For example, ?Terry ate breakfast? does not tell us where (kitchen, dining room, restaurant) or what he ate (cereal, doughnut, or rice, umeboshi, and natto). These elements must be supplied from knowledge about typical role fillers appropriate for the information that is specified in the input. Note that the SBLR has a component that varies by cultural context.

Textually generated 3D scenes will have a profound, paradigm shifting effect in human computer interaction, giving people unskilled in graphical design the ability to directly express intentions and constraints in natural language bypassing standard low level direct manipulation techniques. This research will open up the world of 3D scene creation to a much larger group of people and a much wider set of applications. In particular, the research will target middle school age students who need to improve their communicative skills, including those whose first language is not English or who have learning difficulties: a field study in a New York after school program will test whether use of the system can improve literacy skills. The technology also has the potential for interesting a more diverse population in computer science at an early age, as interactions with K 12 teachers have indicated. CIF:Medium:Collaborative Research:Understanding and Managing Interference in Communications Networks The ubiquity of wireless devices and services and the ever increasing bandwidth demand make it imperative to improve spectrum utilization. Key to improving spectrum efficiency in a multi user networks is understanding and managing interference. This collaborative project studies the phenomenon of interference in communication networks and develops a theoretical foundation on how to deal with interference under realistic operating conditions. The study addresses several long standing open problems; solutions to those problems should collectively advance our understanding on interference and provide guidance on the design of future wireless networks.

This project pursues a broad range of topics that are of great theoretical and practical significance. Conventional wisdom suggests that, since interference has structure that is not present in thermal noise, this structure should be exploited by transceivers. On the other hand, there are situations where interference can be essentially ignored without compromising system throughput, as evidenced by recent breakthroughs in the study of the sum rate capacity of Gaussian interference channels. This project demonstrates that there are different regimes for interference management and provide means of analysis for various multi user communication systems. The transformative nature of this project lies in its ambitious goal of establishing a solid theoretical foundation on how interference should be dealt with in complex networks, and in providing guidance on system designs that achieve optimal performance. RI: Medium: Collaborative Research: From Text to Pictures This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The researchers are developing new theoretical models and technology to automatically convert descriptive text into 3D scenes representing the text?s meaning. They do this via the Scenario Based Lexical Knowledge Resource (SBLR), a resource they are creating from existing sources (PropBank, WordNet, FrameNet) and from automated mining of Wikipedia and other un annotated text. In addition to predicate argument structure and semantic roles, the SBLR includes necessary roles, typical role fillers, contextual elements, and activity poses which enables analysis of input sentences at a deep level and assembly of appropriate elements from libraries of 3D objects to depict the fuller scene implied by a sentence. For example, ?Terry ate breakfast? does not tell us where (kitchen, dining room, restaurant) or what he ate (cereal, doughnut, or rice, umeboshi, and natto). These elements must be supplied from knowledge about typical role fillers appropriate for the information that is specified in the input. Note that the SBLR has a component that varies by cultural context.

Textually generated 3D scenes will have a profound, paradigm shifting effect in human computer interaction, giving people unskilled in graphical design the ability to directly express intentions and constraints in natural language bypassing standard low level direct manipulation techniques. This research will open up the world of 3D scene creation to a much larger group of people and a much wider set of applications. In particular, the research will target middle school age students who need to improve their communicative skills, including those whose first language is not English or who have learning difficulties: a field study in a New York after school program will test whether use of the system can improve literacy skills. The technology also has the potential for interesting a more diverse population in computer science at an early age, as interactions with K 12 teachers have indicated. SHF: Medium: MACANTOK a MAchine Code ANalysis TOol Kit and its Applications Recent work has revealed how important it is to examine the
properties of programs after they have been translated to machine
code. For instance, many security exploits depend on
platform specific features that are not visible at the source code
level, such as memory layout details (e.g., the offsets of variables
in activation records and padding between fields of a structure). The
expected contributions of the project include (i) a
language independent tool generator that, from a formal specification
of a given instruction set s syntax and semantics, generates
implementations of dynamic analysis, static analysis, and
symbolic evaluation components tailored to that instruction set, and
(ii) a variety of prototype language specific applications (i.e.,
specific machine code analysis tools), including

o A tool to automate the detection of bugs and security
vulnerabilities in machine code. The aim is to identify definite
bugs and vulnerabilities, and information about what is required to
trigger them.
o A tool to check sequencing properties on machine code.
o A tool that can aid in detecting interoperability problems among
components by inferring input/output and network communication
formats, and by summarizing the behavior of a component s client.

The results will help programmers create correct, reliable, and secure
software systems by providing them with new kinds of tools to (a) verify
properties of a program?s behavior, and (b) find potential bugs and
security vulnerabilities. TC: Medium: Reimagining Cryptography by Identifying its Culturally Rooted Assumptions This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).


In the last few years, universally composable (UC) security, defined and extensively investigated by Ran Canetti, has become a popular and important topic in modern cryptography. Canetti s notion has enormous appeal, promising a unified framework under which one can define virtually any cryptographic protocol goal, and further promising that the resulting definitions will be composable in the sense that a protocol solution for some goal will comprise a suitable primitive to use within any other protocol that desires its abstracted functionality. But UC goals are definitionally complex in fact, the definitions can take tens of pages of English prose to simply write down, and, even then, significant ambiguities may remain.

In this project, Dr. Phillip Rogaway explores an alternative descriptive language for specifying UC goals. Instead of describing the execution model in English, it will be described within the framework of code based game playing. Under this descriptive language, a UC definition will be described using a program, the program setting variables that induce a clear and precise notion of adversarial advantage.

Dr. Rogaway will demonstrate the feasibility and practical value of his approach by applying it to universally composable signature schemes. This will be an important first step towards making unambiguous and verifiable UC definitions accessible to the general cryptographic community and those they serve. III: Medium: Better Information Integration through Uncertainty This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).

The problem of providing seamless, integrated querying over multiple
interrelated sources of information has been plaguing the database,
information management, information retrieval, and artificial
intelligence research communities for decades. There have been
successful lines of research addressing specific components of the
data integration problem, and large one off systems have been built
that successfully integrate specific information sources in specific
domains. However, a completely general solution to the data
integration problem is thought not to be realizable.

The investigator is developing a new type of information integration
system that is both novel and realizable. It is based on the following
premises and components.

1) The system provides a general data integration solution targeted
for a certain type of environment: when multiple sources have joining,
overlapping, and potentially conflicting information about the same or
closely related real world entities.

2) The system permits and exploits uncertainty as an integral part of
data integration: Uncertainty may be present in source data, source
schemata, the integration process, integrated schemata, and integrated
data. In fact, uncertainty can play a key role successful information
integration.

3) The system relies on general purpose entity resolution as a
fundamental building block of the integration process and the
integrated information. Furthermore, the system retains both the
uncertainty and the lineage associated with the entity resolution
process.

4) The system incorporates powerful lineage capabilities, tracking
where, when, and how data was produced, how it has evolved over time,
and how it has been combined and manipulated as part of the
integration process. Lineage is used to enhance the integration
process, and it is offered to the end user in a variety of forms for
data understanding and conflict resolution purposes.

Further information on the project can be found at the project web
page: http://infolab.stanford.edu/udi/ RI: Medium: Collaborative Research: Minimalist Mapping and Monitoring Proposal Title: RI: Medium Collaborative Research: Minimalist Mapping and
Monitoring
Institution: University of Illinois at Urbana Champaign
Abstract Date: 05/05/09
This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).
This project addresses fundamental and challenging questions that are common to
robotic systems that build their own maps and solve monitoring tasks. In particular, the
work contributes to our general understanding of the interplay between sensing, control,
and computation as people attempt to design systems that minimize costs and
maximize robustness.
Powerful new abstractions, planning algorithms, and control laws that accomplish basic
mapping and monitoring operations are being developed in this effort. This is expected
to lead to improved technologies in numerous settings where mapping and monotoring
are basic components.
Ample motivation is provided by technological challenges that involve searching,
tracking, and monitoring the behavior of people, wildlife, and robots. Examples include
search and rescue, security sweeps, mapping abandoned mines, scientific study of
endangered species, assisted living, ground based military operations, and even
analysis of shopping habits.
The work is particularly transformative because it lives outside of the traditional
boundaries of algorithms, computational geometry, sensor networks, control theory, and
robotics. Furthermore, national interest continues to grow in the direction of developing
distributed robotic systems that combine sensing, actuation, and computation. By
helping to break down traditional academic and scientific barriers, it is expected that the
work will transform the way we think about robotics algorithms, the engineering design
process, and the education of students across the robotics, computational geometry,
and control disciplines.
NATIONAL SCIENCE FOUNDATION
Proposal Abstract
Proposal:0905523 PI Name:LaValle, Steven
Printed from eJacket: 05/06/09 Page 1 of 1 RI: Medium: Collaborative Research: Robotic Hands: Understanding and Implementing Adaptive Grasping This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

This project is defining the basis for lower complexity robotic hands that can grasp a wide variety of objects in noisy and unstructured environments. The new generation of mobile and humanoid robots still lacks basic ?hands? that can reliably grasp objects. Robot hands have been traditionally built as anthropomorphic, high degree of freedom (DOF) mechanisms that are expensive and difficult to control. The research team is developing technologies based on defining hand mechanisms that capture two key features of human grasping, versatility and low dimensionality of hand postures. Reducing complexity brings major benefits. Determining the minimal number of hand joints, sensors and actuators can reduce costs and speed research as low complexity hands can be easily fabricated, designs can be quickly iterated, and control can be simplified. These ideas are used to build a low cost, low DOF grasping device that is based on hard human grasping data. Further, the new hand designs are being tested in simulation so as to build hardware that is functionally proven for robotic grasping tasks. Important research outcomes include: development of a new low dimensional, low cost robotic hand; experiments to gain insights from human grasping and adaptive compliance; and machine learning algorithms for grasping. Broader impacts include: collaboration between neuroscience and robotics; hardware design methods and computational tools for hand researchers; providing robust grasping capabilities in real environments such as robots for home care and assistance for the elderly and disabled; establishing links between neural control and prosthetic devices based on dimensionality reduction; and dissemination of modeling and simulation grasping software. NeTS: Medium: Collaborative Research: Designing a Content Aware Internet Ecosystem This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

A significant majority of current Internet traffic is due to distributing content, yet the Internet was designed to be largely agnostic to characteristics of the content flowing over it. This research investigates the design and operation of a content aware Internet ecosystem, which thrives on the interaction between users (seeking seek fast and correct downloads), content providers (seeking to minimize network congestion and transit traffic), and network providers (who generate content, and seek the cost and resource efficient dissemination).

This research takes a two pronged approach. On one hand, it explores novel analysis of fundamental performance limits for a content aware Internet ecosystem that rigorously characterizes the benefits of an intelligently designed cross layer architecture. On the other hand, it includes developing mechanisms and practical implementation of a content distribution system, by which involved parties can interact constructively to achieve these gains yet respect each others interests. This approach combines a range of techniques, including modeling and theoretical analysis, measurement and data analysis, system design, simulation, and system implementation. Affordable and ready access to digital content helps inform, educate, and entertain society as a whole. Additionally, by developing cost and resource effective delivery techniques, the friction continuing to build between involved parties can be reduced and the technical side of the network neutrality debate can be better informed. To enhance this impact, the project includes an educational component involving local universities from under represented groups, curriculum development and interactions with industry. SHF: MEDIUM: Semiconductor Life Extenstion through Reconfiguration Industry will soon manufacture transistors whose sizes are on the order of atoms. These tiny transistors are fragile, meaning each transistor will be different and will change during use. Many transistors will be unusable when first built; still others will degrade or fail with use. Like people or snowflakes, each of our components (e.g. microprocessors, graphic chips, memories) and systems (e.g. cell phones, mp3 players, anti lock brakes) built with these tiny transistors will be unique. Traditional, one size fits all approaches assign tasks to transistors oblivious of their unique strengths and weaknesses; these approaches waste much of the potential benefits of these tiny transistors, leading to systems that cost too much, use too much energy, and fail too soon. This research explores a novel assignment approach that assigns tasks adaptively based on measured transistor characteristics. The fastest transistors are assigned where they most accelerate performance, while the slower transistors can still be used for less time critical tasks. Assignments are further re evaluated during system operation, allowing fresh transistors to replace transistors that wear out.

Practically, this means IC manufacturers can produce smaller transistors and continue to deliver more capable electronics (e.g. digital video recorders, cell phones, laptops, supercomputers) for fixed dollar budgets. These capabilities continue to improve our quality of life, providing richer media, better communication, greater automation, and greater safety. This work will reduce the energy per computational task thereby extending battery life, reducing energy bills, and facilitating cooler operation. Replacement and reassignment mean electronic components will last longer and degrade gracefully. SHF: Medium: Collaborative Research: Throughput Driven Multi Core Architecture and a Compilation System Computing substrates such as multi core processors or Field Programmable Gate Arrays (FPGAs) share the characteristic of having two dimensional arrays of processing elements interconnected by a routing fabric. At one end of the spectrum, FPGAs have single output programmable logic functions, and at the other end, multi core chips have complex 32/64 bit processing cores. For different applications, different programmable substrates produce the best area power performance tradeoffs. This project is developing a large scale multi core substrate that has hundreds or thousands of simple processing cores along with a compilation system that maps computations onto this fabric. This many core architecture, named Diastolic Array, is coarser grained than FPGAs but finer grained than conventional multi cores. To efficiently exploit such a large number of processing cores, the architecture needs spatially mapping a computation to processing cores and communication to the point to point interconnect network. To be practically viable, this mapping process must be automated and effective. The project addresses this challenge by simultaneously developing hardware architecture and a compilation system.


A diastolic array chip is expected to outperform FPGAs or general purpose processors on an interesting class of applications, enabling more efficient prototyping and low volume production. The outcomes of this project such as statically configured interconnection architecture with associated algorithms for routing and resource allocation will also be applicable to other multi core designs. Finally, the project is developing a new parallel processing module for an undergraduate computer architecture class to give sophomores early exposure to parallel hardware, experience with writing parallel programs and using compilers that exploit parallelism. RI: Medium: Hierarchical Decision Making for Physical Agents Research under this award addresses a core problem in the development of intelligent systems: the generation of effective, deliberate activity over substantial time scales in complex state spaces. Whereas current methods for plan generation are limited either to very short plans or to very repetitive behavior in highly structured and simplified environments, the investigators are developing a new mathematical framework for hierarchical decision making. The framework, based on so called angelic nondeterministic semantics for high level actions, underpins new algorithms capable of generating provably optimal high level plans without expanding those plans into primitive actions. The algorithms therefore satisfy the downward refinement property, resolving a problem that has been open for over 30 years at the heart of AI planning research.

By combining hierarchical deliberation, both offline and online, with hierarchical reinforcement learning and apprenticeship learning for low level physical skill acquisition, the research seeks to enable dexterous human scale robots to operate in unstructured environments such as kitchens, offices, and environments of still greater complexity. The research promises to yield a deeper understanding of, and better tools for, large scale decision making in general, with implications in the social, governmental, corporate, and military spheres.

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5). NeTS:Medium:Collaborative Research: Exploiting Battery Supply Nonlinearities in Optimal Resource Management and Protocol Design for Wireless Sensor Networks This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Wireless sensor networks (WSNs) are becoming pervasive in both civilian and military domains. Low cost and ease of deployment of these networks necessitate the use of batteries as the primary source of power. As a result, battery capacity has emerged as a critical design parameter for maximizing the operational lifetime of the network. In this project, a comprehensive battery charge oriented framework for energy management in WSNs is developed. Its key novelty is its accounting of unique nonlinear battery characteristics, including passive recharge, load profile dependence, and capacity fading. Such characteristics have significant impact on the usable battery capacity, and consequently on the network lifetime. Novel, physically justified analytical models for battery charge/discharge are exploited in designing adaptive control strategies for data processing and communications in a WSN, with the aim of maximizing the network lifetime. These strategies are used to operate individual nodes as well as a hierarchical network of nodes. Battery aware adaptivity is performed on CPU voltage/frequency, RF transmission power, transmission rate/modulation scheme, sleep/wakeup scheduling, cluster head assignment, cover selection, etc. Models and algorithms developed in this project are validated and their feasibility demonstrated through simulations and experimentation. The activity includes an education component involving undergraduate and graduate students, and a strong technology transfer plan. The project is expected to lead to novel designs and control strategies for sensor networks, with significantly longer operational lifetime and highly efficient energy management. RI: Medium: Deciphering Natural Language (DECIPHER) This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

This project takes on two problems: (1) deciphering ancient texts using computers, and (2) training automated language translation systems without using parallel texts. Statistical language processing software has played little role to date in the analysis of ancient texts, where data is limited and human intuition has so far ruled. Data for automated language translation is more plentiful, and research has made great strides in the 21st century. However, researchers are addicted to training on large parallel texts, which are limited for the diversity of languages and domains for which people need automated translation.

The project develops unsupervised methods that compensate for the lack of parallel data, using alternative sources of linguistic knowledge. For ancient languages, these sources include known languages as decipherment targets, capitalizing on tight connections within a language family. In translation, large quantities of untranslated data are exploited to induce strong bilingual connections. Formulating these tasks in a decipherment framework brings powerful cryptographic theory and algorithms to bear. Such theory also helps estimate expected translation accuracy given fixed data resources, and gauge whether a lost language is decipherable, given a fixed amount of script.

Computational analysis of ancient scripts offers a better understanding of ancient cultures, and unsupervised techniques construct language connections of great interest to historical linguists. Applying such techniques to automated language translation offers the chance to bring many more language pairs and domains to the population at large. AF: Medium: Collaborative Research: Approximate Computational Geometry via Controlled Linear Perturbation The investigators will develop an approximate computational geometry that is algorithm independent, accurate, and fast. Geometric predicate evaluation and element construction will be performed approximately using floating point arithmetic. Degeneracy will be handled transparently. The evaluation and construction techniques will be encapsulated in a software library that will be free for nonprofit use.

The research challenge is robustness: the output of an approximate algorithm must be correct for a small perturbation of the given input. This definition extends the numerical analysis definition of a stable algorithm to cover combinatorial error. Robustness is a fundamental computer science problem that is a major challenge in computational geometry. The predominant strategy in computational geometry, exact computation using algebraic geometry, has high computational complexity and contradicts the standard scientific and engineering strategy of approximate computation with error bounds. The investigators will adapt approximate computation to the special needs of computational geometry, which is primarily combinatorial. This task involves core research at the interface between computational geometry and numerical computing.

Robust approximate computation will transform how computational geometry is taught, how algorithms are developed and implemented, and how the field interacts with the wider scientific and engineering community. Introductory courses will present a rigorous, practical robustness theory, instead of treating robustness in an ad hoc, incomplete way. Programmers will implement real RAM algorithms as stated, using our library to ensure robustness and to handle degeneracy, instead of addressing these problems anew for every algorithm, which is often a major research challenge. Computational geometry will be available to other disciplines in the form of high quality software libraries, akin to modern applied mathematics libraries. NeTS: Medium: A SCAFFOLD for Service Centric Networking This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

This research proposes a new network architecture, SCAFFOLD, that directly supports the need of wide area services. SCAFFOLD treats service level objects (rather than hosts) as first class citizens and explores a tighter coupling between object based naming and routing. A clean slate, scalable version of the federated SCAFFOLD architecture is being designed and prototyped. System components include programmable routers/switches, resolution services for object based lookup and forwarding, and integrated end hosts.

The center of people s digital lives today are online services not the networks or computers on which they run. The research ultimately explores what abstractions and mechanisms that will make the future network a powerful, flexible hosting platform for wide area services (the so called ``cloud ). In doing so, SCAFFOLD would lower the barrier to deploying networked services that are scalable, reliable, secure, energy efficient, and easy to manage.

The project includes a summer camp outreach activity with schools serving under represented groups to build services on top of SCAFFOLD, new special course development, and technology transfer with industry. NeTS: Medium: Collaborative Research: Towards Versatile and Programmable Measurement Architecture for Future Networks This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Traffic measurement is central to network operation, management, and security. Yet, support for measurement was not an integral part of the original Internet architecture. This project aims to develop a programmable measurement architecture that is versatile enough to support current and future measurement needs, adaptable to dynamic network conditions, modular/lightweight, and scalable to high link speeds.

This project proposes a new flow abstraction module and query language that can specify arbitrary traffic sub populations for statistics collection. Efficient data structures to encode these queries will be developed. The team also strives to identify a core set of data streaming and sampling primitives that can be composed together to satisfy most of the queries. Efficient hardware implementation for these core set of primitives will constitute the basic measurement modules that can be easily reconfigured to measure traffic at different desired granularity. Measurement application case studies will be carried out to evaluate and showcase the capabilities of the proposed approach.

This project has great potential in guiding the design of a clean slate measurement instrumentation for future Internet. It will provide both graduate and undergraduate students with training that span multiple disciplines, from fundamental statistical theory, algorithm design, to hardware implementation. The results (including the query language and underlying data structures, sampling/streaming algorithms, and hardware building blocks) will be broadly disseminated through publications, invited talks/tutorials, and open sourcing software distribution. TC: Medium: Collaborative Research: Unification Laboratory: Increasing the Power of Cryptographic Protocol Analysis Tools This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The project develops cryptographic protocol reasoning techniques
that take into account algebraic properties of cryptosystems.
Traditionally, formal methods for cryptographic protocol
verification view cryptographic operations as a black box,
ignoring the properties of cryptographic algorithms that can be
exploited to design attacks. The proposed research uses a novel
approach based on equational unification to build new more
expressive and efficient search algorithms for algebraic theories
relevant to cryptographic protocols. Equational unification gives
a compact representation of all circumstances under which two
different terms correspond to the same behavior. The algorithms
are implemented and integrated into Maude NPA, a system that has
been successful in symbolic protocol analysis. It is demonstrated
that Maude NPA when enriched with such powerful unification
algorithms can analyze protocols and ensure their reliability,
which could not be done otherwise.

Improved techniques for analyzing security are helpful both in
assuring that systems are free of bugs, and in speeding up the
acceptance of new systems based on the confidence gained by a
formal analysis. This research will lead to the design and
implementation of next generation tools for protocol analysis.
Algorithms developed will be made available to researchers as a
library suitable for use with protocol analysis tools. Tools from
the project will help students understand concepts relevant to
protocol design and get hands on experience. Equational
unification for algebraic theories is not only useful for
protocol analysis, but also for program analysis in general, thus
making the results of this project to be widely relevant. III: Medium: A Machine Learning Approach to Computational Understanding of Skill Criteria in Surgical Training This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

A recent study by the Agency for Healthcare Research and Quality (AHRQ) documented over 32,000 mostly surgery related deaths, costing nine billion dollars and accounting for 2.4 million extra days in hospital in 2000. At the same time, economic pressures influence medical schools to reduce the cost in training surgeons. The success of simulation based surgical education and training will not only shorten the time involving a faculty surgeon in various stages of training (hence reducing cost), but also will improve the quality of training. Reviewing the state of the art research, there are two research challenges: (1) to automatically rate the proficiency of a resident surgeon in simulation based training, and (2) to associate skill ratings with correction procedures. This award will explore a machine learning based approach to computational understanding of surgical skills based on temporal inference of visual and motion capture data from surgical simulation. The approach employs latent space analysis that exploits intrinsic correlations among multiple data sources of a surgical action. This learning approach is enabled by the simulation and data acquisition design that ensures clinical meaningfulness of the acquired data.

Intellectual Merit:
The intellectual merit of the proposed work lies in the exploration of the answers to two key research questions: (1) how to develop a computational understanding of surgical skill defining criteria (high level knowledge) from low level, raw sensory observations; and (2) how to efficiently perform temporal inference from correlated data sources of disparate nature with unknown structures. Expertise related to motion is a well observed phenomenon but the task of inferring motion expertise patterns from captured data especially video is largely unexplored. This award presents a controlled methodology to extract expertise patterns from raw sensory data. This interdisciplinary team with collaboration from medical professionals is well positioned to address these fundamental research issues.

Broad Impact:
The broad impact of the proposed work is threefold. First, the ability of the proposed system to create descriptive, mathematical models from recorded video and other data in surgical training opens up the possibility of new paradigms that can significantly improve current practice in surgeon education and training. Second, the proposed approaches may enable development of lower cost surgical training systems. Third, the novel formulation of and solution to the fundamental problem of expertise inference from disparate temporal measurements with intrinsic correlation can find many other applications such as sports (e.g., baseball and golf training), rehabilitation, and surveillance. The impact of the interdisciplinary project on education manifests in its direct contribution to better training and education of surgeons and to nurturing a new generation of students who possesses strong interdisciplinary background and skills.

Key Words: Surgical Simulation, Surgical Education and Training, Complementary Feature Selection, Visual and Temporal Inference, Learning in Latent Space. TC: Medium: Collaborative Research: Novel Forensic Analysis for Crimes Involving Mobile Systems TC: Medium: Collaborative Research: Novel Forensic Analysis for Crimes Involving Mobile Systems

Abstract:

Our project will significantly advance forensic methods of investigating mobile devices used for trafficking in digital contraband. While current methods and legislation focus heavily on logical identifiers, we will design, evaluate, and deploy new forensic techniques that focus on consistent and trackable characteristics of mobile computing. Additionally, our work will play an important role in understanding the limits of personal privacy in these settings.

We will develop new radio fingerprinting techniques that detect identifying information present in a radio s low level components. We seek the comprehensive understanding available from an accurate model of these processes so that we can both determine the key causes of anonymity loss and investigate new countermeasures.

We will develop novel techniques of traffic analysis that determine the source of encrypted Web traffic. Our focus will be on real world traffic scenarios, where background traffic is present and the entire Internet is a potential source.

We will empirically evaluate our methods in a real world setting by using two large, outdoor and indoor wireless testbeds that we have deployed.

Our research will directly assist law enforcement that investigate network trafficking of images of child sexual exploitation, demonstrating the usability of trustworthy computing. We will disseminate our results to the Internet Crimes Against Children Task Force and the Massachusetts State Police. Additionally, this project will define research pathways to allow students who complete their BS and MS degrees at John Jay (a minority serving institution) to continue their PhD work at UMass Amherst. III: Medium: Collaborative Research: Towards On Line Analytical Mining of Heterogeneous Information Networks This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5). This proposal seeks to address a wide range of research challenges involving OLAP (OnLine Analytical Processing). This collaborative project is concerned with increasing the ability for analytical processing for information networks. Information networks have been expanding rapidly and attract broad scholarly interest ranging from intrusion pattern detection to social community discovery. Typical information networks include communication networks, social networks, the Web, and biological networks. In contrast to the rising popularity and increasing scale of information networks, there is a lack of general analytical processing frameworks for exploiting the information contained in the networks. The lack of such frameworks inhibits personalized navigation and interactive knowledge exploration. The work supports the Infonet OLAP framework by extending methods for structure discovery, network summarizations, and self quality assurance of underlying networks. If successful, the techniques would simplify information network analytical processing and transform existing ad hoc graph exploratory work into a unified framework as traditional OLAP does to multidimensional data analysis. TC: Medium: Collaborative Research: Mobile Personal Privacy and Security A New Framework and Technology to Account for Human Values This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Significant progress has been made on technical solutions for implementing security while preserving varying degrees of privacy for mobile electronic devices. A largely unsolved problem involves knowing what type of technical solutions to implement and under what circumstances. This challenge arises because in actuality there is in general no such thing as a clear optimal or right solution. One way to view this problem is by focusing on the fact that such devices potentially play an integral role in people s lives and impact important human values: security and privacy of course, but also other values such as autonomy, trust, and physical well being. Further, these values are often in tension. This project will develop a principled and systematic conceptual framework for analyzing these value implications and tensions in the context of two archetypical examples of future mobile devices: implantable medical devices and mobile phone safety applications. A key component of our approach will be the application and extension of the Value Sensitive Design theory and methodology. Researchers will develop and evaluate new technology, as well as undertake empirical work with a range of stakeholders, including Futures Workshops and semi structured interviews. Expected outcomes include a framework for analyzing the relationships and trade offs among privacy, security, autonomy, and other values for mobile devices that accounts for both situational and embodied dimensions of these values; a pallet of key technical solutions to these problems; a set of case studies; and finally design recommendations for use by other researchers and practitioners.

Results of this work will not only develop technical mechanisms for providing the right level of security and privacy for new mobile technologies, but can help influence the definition of what right means in the context of the broader set of goals and values. In society more generally, there is an increasing use of technologies that are focused upon here: implantable medical devices and highly capable cell phones used for personal safety. Improving human health and personal safety, while not significantly undermining other key values such as privacy, trust, and autonomy, will be important in the development of such technologies. It is anticipated that the ways of analyzing such problems will be applicable to other technologies as well, such as RFID tags on personal possessions, Smart Cards for paying for tolls or transit fares, and many others. TC: Medium: Collaborative Research: Towards Self Protecting Data Centers: A Systematic Approach This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Data centers using virtual machine (VM) consolidation are taking over old computer rooms run by individual companies. However, consolidating services and resources does not consolidate security automatically. To meet the top two requirements for modern data centers, namely business continuity and information security, this research will take a systematic approach that leverages the emerging VM technologies to consolidate four areas of systems security research: redundancy, microscopic intrusion analysis and detection, automatic response, and diversity driven protection. We will make innovative contributions on various aspects of security consolidation, including (1) An architecture and underlying techniques based on diversified replication towards defensive protection against unknown attacks; (2) Novel cross layer and cross VM methods for causal relation logging, event correlation, damage assessment, and forensics; (3) New intrusion detection techniques based on unique cross VM replica inconsistency checking techniques, and new cross layer inconsistency checking methods; (4) A novel pipelining approach towards automated intrusion response; and (5) New techniques for on the fly data center intrusion confinement and recovery.

Our research will result in significant advances in helping mission/life/business critical applications and information systems reduce risk, increase business continuity, and deliver data assurance in the presence of severe cyber attacks. Broader impact will also result from the education, outreach, and dissemination initiatives. Educational resources from this project, including course modules and teaching laboratory designs, will be disseminated through a dedicated Website.

Key Words: self protection; recovery; virtual machine monitor; causal relations; availability RI: Medium: Collaborative Research: Methods for Empirical Mechanism Design The design of computational systems that support group decisions, allocate resources to distributed tasks, or mediate social interactions is fundamentally different from the corresponding design problem serving individual or centralized users. When multiple parties, or agents, are involved, the designer s objectives are complicated by the fact that the interests of these parties are rarely, if ever, perfectly aligned. The field of mechanism design offers a theoretical framework that directly addresses the issue of incentives as it relates to the design of multiagent systems. However, this purely analytical approach carries with it inherent practical limitations. The investigators introduce a new approach, empirical mechanism design (EMD), whose premise is to extend the basic foundation of mechanism design with empirical methods such as simulation and statistical analysis. These extensions promise to dramatically expand the scope of mechanism design beyond the small scale, stylized, or idealized domains to which it has been predominantly limited to date.

The project will investigate several concrete EMD problems, within the general theme of market design. Improved market design has significant implications for the public and private sectors. In public policy, market based approaches are likely to play a major role in, for example, instituting measures to cope with climate change, banking reform and regulation, and adoption of new energy sources. In the commercial domain, new markets for advertising placement, computational services, and other goods will also entail significant mechanism design efforts. Regardless of the sector, design outcomes bear on important social objectives including efficiency, transparency, and stability (e.g., of financial relationships). An empirical basis for evaluating candidate mechanisms will complement existing theoretical perspectives, enriching the tools available to designers and other stakeholders. TC: Medium: Self Securing Services for Mobile Handsets The main objective of the proposed research is to develop a self securing framework for smart mobile handsets, called S3Mobile (Self Securing Services for Mobile handsets), that protects the handsets against known and unknown malware. The key research components of S3Mobile include: (1) collect malware samples for mobile handsets, (2) record the hardware and software usage logs of the samples that represent the corresponding run time behavior, (3) develop an algorithm for extracting features from the logs, (4) analyze similarities, features, and exploit vectors, and (5) implement and evaluate S3Mobile on the Android platform.

The general approach to the problem is to conduct some monitoring on the handset, but, recognizing the more limited computational and electrical resources available there, to use a remote server to conduct more computationally intensive activities and to maintain repositories of malware and normal application behavior. The server side facilities are referred to as a second line of defense. Four particular challenges are noted for the work: accurately specifying malware behavior, accurately detecting such behavior, testing the constructed system to evaluate its performance correctly, and the difficulty of obtaining malware samples. The proposal documents approaches to each of these challenges, including the use of a temporal logic based notation for describing malware behavior and a combination of honeypots and industrial collaboration to provide malware samples.

If successful, the work could lead to a cellphone / smartphone infrastructure with much improved resilience to software based attacks that aim to steal information or drain power from the phone. RI: Medium: Building Flexible, Robust, and Autonomous Agents This project is developing computational agents that operate for extended periods of time in rich and dynamic environments, and achieve mastery of many aspects of their environments without task specific programming. To accomplish these goals, research is exploring a space of cognitive architectures that incorporate four fundamental features of real neural circuitry: (1) reinforcing behaviors that lead to intrinsic rewards, (2) executing and learning over mental, as well as, motor actions, (3) extracting regularities in mental representations, whether derived from perception or cognitive operations, and (4) continuously encoding and retrieving episodic memories of past events. A software framework called Storm facilitates this exploration by enabling the integration of independent functional subsystems, allowing researchers to easily plug in and remove different subsystems in order to assess their impact on the overall behavior of the system.


Cognitive architectures are being tested by exposing them to a wide variety of novel environments with unpredictable (and non repeatable) extrinsic rewards, but in which many actions could lead to intrinsic rewards (e.g., surprise). To assess flexibility, an automated environment generator exposes agents to environments that are unknown in advance to the artificial agent or human researcher. To assess robustness, cognitive systems are being exposed to many variants of the same environment to ensure that the systems can learn from past experience and generalize when appropriate. And to assess autonomy, systems must operate effectively for extended periods of time in a dynamic environment.

In the longer term, flexible and robust cognitive architectures being devloped under this research will have application as the brains of robotic and software systems in emergency, miltary, and a wide variety of other societal and service realms.

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5). AF: Medium: Collaborative Research: Integral Equation Based Fast Algorithms and Graph Theoretic Methods for Large Scale Simulations The phenomenal advance in computer technology in terms of processing
speed and capacity, closely described by Moore s law, in the last four
decades has been outpaced by the explosive amount of data that are used
to describe more realistic models in scientific computing. For instance,
the number of unknowns in a linear system has grown from hundreds in
the past to tens of millions nowadays. Fast algorithms such as the
celebrated fast multipole method (FMM) have provided a computational
tool for narrowing the gap. At the same time, there is a great need
and challenge to develop better computation techniques and utilize the
present and emerging computers, with the gain in speed up to a couple
of orders of magnitude. The goal of the proposed research is to
advance computational theories and techniques, in order to meet the
demand and challenge for large scale simulations of complex
systems in scientific, medical and engineering studies.

The research team proposed to investigate, innovate and integrate the
key simulation steps, from analytic re formulation of system models with
complex geometries to combinatorial optimization in mapping numerical
algorithms to computing architectures. Many traditional models are
formulated in terms of linear or nonlinear partial differential
equations (PDEs) with boundary conditions on complex geometries. By
the work of other researchers and principal investigators,
integral equation (IE) formulations have lead to better numerical
algorithms in both efficiency and stability, and more importantly
enabled certain important large scale simulations. It is proposed first
to study the reformulation of traditional PDE models into IE models, as
a direct and analytical approach to innovative algorithm design. Next,
preconditioning techniques will be studied as an indirect and
stabilization approach. Furthermore, Graph theoretic methods will be
applied to optimize the FMM based algorithms on various modern
computer architectures, especially, parallel architectures. These key
components will be studied in conjunction, not in isolation.

The intellectual merits of the proposed work are three fold.
It sheds lights on (1) the model reformulation into IEs of the second
kind as a fundamental analytic algorithmic approach to accelerating
and stabilizing numerical computation, (2) the connection between
reformulation and preconditioning, and (3) on the mutual dependence
of numerical algorithms and computer architectures. The proposed
work will have broader impacts on various applications through
timely dissemination with demonstration of case studies. Three
application areas of specific concern are electrostatics calculation
in molecular dynamics simulations, computational fluid dynamics, and
the study of oxygen delivery in tissues and tumors via microvascular
networks. The proposed work involves interdisciplinary research
collaboration and cultivation of young and new researchers with
multi disciplinary backgrounds. Finally, the findings and
algorithms will be embodied in open source high performance
software to facilitate research computing by and large and to
be used in classrooms. TC: Medium: Collaborative Research: Securing Concurrency in Modern Systems This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Concurrency related vulnerabilities are pervasive in modern computing
systems. Concurrency exploits include time of check to time of use
(TOCTTOU) race conditions in file systems, attacks on signal handlers,
and evasive malware that uses concurrency to escape sandboxing
mechanisms. As processors feature ever more parallelism, and
computers process more of our sensitive data, defending against
concurrency attacks is a key challenge for the coming decade.

The first goal is to protect legitimate applications from concurrency
attacks when they access system resources (e.g., prevent TOCTTOU
attacks on file accesses and exploitable race conditions in signal
handlers). The objective is to provide application programmers with
mechanisms and policies for synchronizing access to system resources
so they can avoid unintentional vulnerabilities.

The second goal is to provide strong confinement of untrusted code in
the presence of concurrency, i.e., blocking intentionally malicious
behavior. Today s malware abuses concurrency mechanisms to bypass and
circumvent containment mechanisms like reference monitors and system
call wrappers. Providing robust system support for containing
malicious code is a critical challenge in intrusion detection and
prevention.

Modern computing systems fundamentally depend on concurrency for their
performance and functionality. Making sure that concurrency is used
securely is essential for building a trusted cyber infrastructure.
This research will have a significant impact on the practical
development of secure software, and enable security critical
applications to realize the performance benefits of today s highly
parallel systems. TC: Medium: Collaborative Research: Wide Aperture Traffic Analysis for Internet Security TC: Medium: Collaborative Research: Wide Aperture Traffic Analysis for Internet Security

Among emerging network threats, some of the most pernicious and elusive are stealthy attacks that take place at very low rates and in a targeted fashion. This project is developing methods for identifying malicious and unwanted activity in the Internet specifically, traffic that is low volume and well hidden among normal traffic. The approach being taken is to develop new methods for direct analysis of Internet traffic of unprecedented scope and scale. In particular, the project is designing and implementing a system that leverages high performance cluster computing to allow application of sophisticated pattern analysis and machine learning algorithms to network traffic at the packet and flow level.

An organizing principle of the system is its decomposition into data parallel lenses and more computationally challenging pattern analysis components. The project is investigating the application of this architecture to dark address monitoring in traffic from core networks a capability that has not been possible to date.
The end result of this project will be a set of tools and a running system that may be used by researchers to enable new investigations into traffic analysis, and may be used by network operators on an ongoing basis to help protect their networks. III: Medium: Collaborative Research: Integration, Prediction, and Generation of Mixed Mode Information using Graphical Models, with Applications to Protein Protein Interactions This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Probabilistic graphical models provide a powerful mechanism for representing and reasoning with uncertain information. These methods have been successfully applied in diverse domains such as bioinformatics, social networks, sensor networks, robotics, and web mining; in turn, such application areas have posed new computational challenges driving graphical model research. This project is motivated by challenges in emerging application areas such as epidemiological simulation, geoscience modeling, and studies of interacting proteins, where there are rich sets of information of multiple types and at multiple levels of granularity. While the methods developed will be general, the research will focus on protein protein interactions, which drive the molecular machinery of the cell by forming transient or persistent complexes to propagate signals, catalyze reactions, transport molecules, and so forth. The mixed mode information available includes amino acid sequences, three dimensional structures and associated physical models, and binary, rank ordered, or even quantitative interaction data. The proposed techniques address key challenges in information integration, prediction, and generation using graphical models.

Intellectual merits:
The intellectual merits of this work derive both from the new capabilities for information integration and for reasoning with probabilistic graphical models, as well as their application to the study of protein protein interactions. Proteins offer, by far, some of the most complex, multi faceted datasets for integration using computational methods; hence the lessons learned here can be applied to similarly rich information spaces, such as epidemiology and geosciences. These integrated models of interacting proteins and new algorithms for prediction and generation will also support significant applications such as protein engineering and systems biology, bridging interaction networks to the underlying residue level interactions in order to better understand and control them.

Broader impacts:
This project will reach out to both the bioinformatics and larger computer science communities to maximize the impact of our contributions. An open source integrator platform will be developed, aimed at integrating protein datasets and which can be extended to information integration in other domains as well. To stimulate community building and foster discovery, the research team will advocate situating computer science research in the context of concrete applications. Building on prior successes, the team will organize a workshop at a suitable venue such as ICML/AAAI/NIPS/KDD focused on an information integration challenge dataset involving protein modeling. Finally, through programs such as Women@SCS at Carnegie Mellon, WISP (Women in Science Program) at Dartmouth, Howard Hughes education grant internships at Purdue, and the MAOP/VTURCS (Minority Academic Opportunities Program and VT Undergraduate Research in Computer Science) program at Virginia Tech, the team will provide cross disciplinary training to undergraduate students from underrepresented groups.


Keywords: Probabilistic Graphical Models, Information Integration, Mixed Mode Datasets, Bioinformatics, Proteins, Markov Chain Monte Carlo (MCMC) methods. CIF: Medium: Collaborative Research: Cooperative Networking Across the Layers CIF: Medium: Collaborative Research: Cooperative Networking Across the Layers

This research goes beyond the physical layer in defining and analyzing cooperative techniques for wireless networks. By incorporating higher layer properties such as traffic dynamics and access control, the investigators develop a new theoretical framework for analyzing and designing cooperative networking algorithms across the layers, which includes existing cooperative techniques such as cooperative relaying and network coding. The basis of this research rests on two major points:: first, the realization that cooperative communication at the physical layer cannot be viewed in isolation, since it has implications at the access and network layers, and second, the recognition . that cooperation at the higher layers, in its own right, can significantly impact overall network performance.

This project has three inter related thrusts: The first thrust studies resource allocation policies which stabilize the queues within various classes of cooperative networks, if stability is indeed attainable. The second thrust determines a family of scheduling algorithms which maximize the volume of traffic served by a cooperative network within a finite horizon. Finally, game theoretic models are developed for cooperative networks where nodes in the network are allowed to pursue differing objectives, and come to a distributed agreement on the (locally) optimal operating point for the overall network.
This research also considers non stationary and non ergodic environments that are more appropriate representations of the wireless channel in a network. RI: Medium: Collaborative Research: Methods of Empirical Mechanism Design The design of computational systems that support group decisions, allocate resources to distributed tasks, or mediate social interactions is fundamentally different from the corresponding design problem serving individual or centralized users. When multiple parties, or agents, are involved, the designer s objectives are complicated by the fact that the interests of these parties are rarely, if ever, perfectly aligned. The field of mechanism design offers a theoretical framework that directly addresses the issue of incentives as it relates to the design of multiagent systems. However, this purely analytical approach carries with it inherent practical limitations. The investigators introduce a new approach, empirical mechanism design (EMD), whose premise is to extend the basic foundation of mechanism design with empirical methods such as simulation and statistical analysis. These extensions promise to dramatically expand the scope of mechanism design beyond the small scale, stylized, or idealized domains to which it has been predominantly limited to date.

The project will investigate several concrete EMD problems, within the general theme of market design. Improved market design has significant implications for the public and private sectors. In public policy, market based approaches are likely to play a major role in, for example, instituting measures to cope with climate change, banking reform and regulation, and adoption of new energy sources. In the commercial domain, new markets for advertising placement, computational services, and other goods will also entail significant mechanism design efforts. Regardless of the sector, design outcomes bear on important social objectives including efficiency, transparency, and stability (e.g., of financial relationships). An empirical basis for evaluating candidate mechanisms will complement existing theoretical perspectives, enriching the tools available to designers and other stakeholders. SHF: Medium: Collaborative Research: Fixing Real Bugs in Real Programs Using Evolutionary Algorithms Fixing software bugs is a difficult and time consuming process, accounting for up to 90% of the lifetime cost of a typical program. Because the number of defects outstrips the resources available for repairing them, most software is shipped with both known and unknown bugs. All previous approaches to debugging multiple types of defects have been manual. This research will develop a fully automated method for repairing bugs in existing software, thus reducing some of the cost of software maintenance.

The research will be broadly applicable, targeting off the shelf, legacy applications created without program annotations or special coding practices. The technical focus of the work is an automated repair approach using evolutionary algorithms. Program variants are evolved, using analogues of biological processes such as mutation, until one is found that both retains required functionality and avoids the defect. Standard software test cases are used to represent the bug and to encode program requirements. Generated repairs can be presented to developers or applied to the program directly. Significant potential outcomes include: an automated program repair methodology and freely available tools; advances in program analyses; advances in evolutionary algorithms; significant efforts in outreach and education; and dissemination of the results. SHF: Medium: Exposing and Eliminating Errors at Component Boundaries The research investigates a new method for detecting errors that occur at module boundaries involving complex application program interfaces. The method first dynamically observes running programs to obtain constraints that characterize successful interactions at module boundaries. It then uses symbolic dynamic taint tracing to obtain symbolic expressions that characterize how regions of the input map to specific values that appear at module boundaries. A constraint solver then generates new input regions that produce values at module boundaries that violate the observed constraints. The final step is to run the program on the resulting new inputs to see if the inputs expose errors involving interactions between modules.

The significance of the research is that many reusable modules present complex interfaces that are difficult for developers to understand.
Module boundaries therefore comprise a prime location for errors and security vulnerabilities in software systems. The research promises to develop new testing techniques for finding and eliminating these errors and vulnerabilities. Broader impacts include more reliable software infrastructure for our society and the education of a skilled workforce. Intellectual merit includes a better understanding of errors in software systems and new techniques for finding and eliminating these errors. RI: Medium: Collaborative Research: The Effect of Subglottal Resonances on Machine and Human Speaker Normalization Last Modified Date: 05/02/09 Last Modified By: Tatiana D. Korelsky

Abstract
This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Despite large acoustic differences in the speech of various talkers, humans are generally able to understand each other quickly and easily. The mechanisms by which humans map such variability onto a set of phonemes has been the subject of research for more than 50 years. This speaker normalization problem has generally been thought of in terms of normalizing the formant frequencies of a particular speaker with a reference set of formants. In this project, a novel approach to speaker normalization is explored, in which not formants but subglottal resonances
(SGRs) are normalized. SGRs have previously been shown to define a set of frequency bands within which formants may vary, yet retaining the same phonemic vowel quality. Normalizing SGRs (and associated frequency bands) therefore reduces formant variability in an effective way. In this project, effects of SGR normalization on automatic speech recognition
(ASR) performance are evaluated for both adult and child speakers of English and Spanish. In parallel, effects on human speech perception in multi talker conditions are explored. Results are expected to improve ASR performance and shed light on human speech production and perception. The project will result in speech databases (including direct recordings of SGR acoustics) and ASR tools, which are critically useful for research in speech production, perception, speaker identification, and speech processing algorithms for cochlear implants and multi lingual ASR. The collaboration in Engineering, Linguistics, Speech & Hearing, and Psychology facilitates a multidisciplinary learning environment.
Publications, results, databases, and tools will be disseminated to the research community. NeTS: Medium: Collaborative Research: Secure Networking Using Network Coding This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

This project determines the fundamental limits of network secrecy from a network coding perspective, and then applies this theory to improve security guarantees in peer to peer and wireless networks. As network coding gains prominence as an important strategy for both wired and wireless networks, the project identifies both the advantages and vulnerabilities from using network coding. Subsequently, the effort develops a design methodology that exploits the advantages while carefully compensating for the vulnerabilities.

This project analyzes networks under both outsider and insider attacks. Specifically, coding mechanisms are developed to combat an external eavesdropper. Also, a combination of cryptographic and information theoretic tools are used to combat internal modification attacks on the network. The results are then used in two case studies: eavesdropper attacks on wireless mesh networks and pollution attacks on P2P content distribution systems.

Secure network coded systems, once well understood, can greatly impact how networks are designed and deployed. Nearly every network setting (wireless, wired or heterogeneous) can benefit in terms of improved resilience (in addition to other performance benefits such as throughput) in its design. Case studies in this effort are designed to help transition the theoretical principles developed into practical algorithms.

The research team includes an industry member which will aid in transitioning our research ideas from theory to practice. The team will disseminate its findings through traditional scholarly venues, through the web and to the local community at each partner institution. NeTS: Medium: Collaborative Research: Unlocking Capacity for Wireless Access Networks through Robust Cooperative Cross Layer Design Cooperative networking exploits the broadcast nature of the wireless channel by effectively pooling the overheard information, which is traditionally treated as harmful interference. While there is a mature suite of tools at the physical (PHY) layer to harvest cooperative gains, it is still unclear how these tools can be employed to deliver significant network capacity gains. The goal of this project is to design and implement cross layer mechanisms for cooperative networking. By integrating PHY layer cooperation with Medium Access Control (MAC) and application layers, the project will provide higher network capacity and improved multimedia quality.

The project has two interrelated components investigating basic architectures for next generation cooperative networks: (i) Cooperative data transmission, which focuses on a robust cooperative MAC PHY incorporating multiple relays under mobility and loose requirements on synchronization and network topology. (ii) Cooperative video transmission, which exploits the synergy between cooperation and layered compression in providing unequal error protection, as well as differential quality in multicast.

Apart from potential impacts on the theory and practice of new wireless technologies, this project will train undergraduate and graduate students in all aspects of wireless communications. The impact on industry will be facilitated by the close relationship of NYU Poly with WICAT member companies.


This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5). III: Medium: Data Interoperability via Schema Mappings This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Interoperability of heterogeneous data is a critical problem faced by every modern enterprise that is concerned with data analysis, data migration, and data evolution. The fundamental goal in data interoperability is to facilitate and make transparent to end users the extraction of information from multiple heterogeneous data sources that reside in different locations. At the heart of achieving data interoperability is the design and management of schema mappings. A schema mapping is a specification of the relationship between two database schemas. Schema mappings are the essential building blocks in specifying how data from different sources are to be integrated into a unified format or exchanged (i.e., translated) into a different format.

The intellectual merit of this project is the development of a solid foundation and a suite of techniques and tools for designing, understanding, and managing schema mappings. Earlier foundational work on schema mappings has mainly focused on the semantics and algorithmic issues of some of the basic operators for manipulating schema mappings with emphasis on the composition operator and the inverse operator. While the composition operator is well understood by now, much more remains to be done in the study of the inverse operator. One of the main goals of this project is to investigate in depth the inverse operator and also the difference operator, which remains largely unexplored to date. This project addresses several fundamental questions for the inverse and the difference operators, including the following: What is the right semantics for these two operators? What is the exact language for expressing these operators? Are there efficient algorithms for computing the result of the inverse operator and the difference operator? A parallel goal of this project is the development of a set of concepts and techniques for optimizing schema mapping and transforming more complex schema mappings into simpler, yet equivalent, ones. The final main goal of this project is to study the problem of using data examples to explain and illustrate schema mappings. The design of schema mappings between two schemas has been known to be one of the most costly and time consuming tasks in achieving data interoperability. Prior studies have suggested that (familiar) data examples can be extremely powerful aids in designing schema mappings. This project addresses the following questions: What is the right notion or notions of illustrative data examples for schema mappings? How easy or difficult it is to compute small examples for illustrating schema mappings? How can one illustrate large and complex networks of schema mappings with data examples? How can one depict the similarities and differences among multiple schema mappings?

The broader impact of this project is the development of human resources in science and engineering through the teaching, mentoring, and research training of graduate and undergraduate students on the foundational and system development work of this project. Further information about publications, course material, and software prototypes and tools developed through this project can be found at the project web page http://datainterop.cs.ucsc.edu CIF:Medium:Collaborative Research:Integrating and Mining Bio Data from Multi Sources in Biological Networks Networks are a natural, powerful, and versatile tool representing the structure of complex systems and have been widely used in many disciplines, ranging from sociology to physics to biology. Massive amounts of biological data from multiple sources await interpretation. This calls for formal information integration, modeling and mining methods. The functioning of complex biological systems demands the intricate coordination of various cellular processes and their participating components. As biological networks grow in size and complexity, the model of a biological network must become more rigorous to keep track of all the components and their interactions, and in general this presents the need for new methods and technologies in information acquisition, transmission and processing.

This research develops a set of novel methods and algorithms of integrating, analyzing and mining biological networks that include multiple sources of biomolecular information such as gene and protein expressions, interactions, and regulations, the formation and dissolution of protein complexes, and interactions between proteins and small molecules. Quantitative and qualitative bio data from multiple sources are iteratively integrated, mined and organized to generate scalable hypothetical biomolecular network structures. The dynamics of these computational hypotheses are tested and refined through dynamic simulation and laboratory experiments. The investigators will develop an integrated modeling and mining method for the biomolecular networks governing cellular processes, with natural connections to both the biological knowledge base and the design and analysis of biological experiments. We will also plan to investigate novel graph based theoretical models and algorithms for biological data represented as networks or graphs. NeTS: Medium: Collaborative Research: Unifying Network Coding and Cross Layer Optimization for Wireless Mesh Networks: From Theory to Distributed Algorithms to Implementation Wireless mesh networks promise a flexible and cost effective solution for bringing high bandwidth low latency applications to the home. Two orthogonal but immensely attractive approaches for designing high performance mesh networks are network coding and cross layer optimization. The former exploits the interfering wireless media as a broadcast channel by ingenious information processing, while the latter intelligently allocates and shares network resources across the layers. This project investigates the long overdue and highly challenging synergistic union of network coding and cross layer optimization.

The characterization of intersession network coding (INC), coding across different network flows, is NP hard. Nonetheless, the problem can be made tractable under important practical settings. Based on new path based characterizations, this project rigorously quantifies the network capacity in presence of different practical INC schemes. A new network coding based cross layer paradigm is developed for controlling mesh networks that takes into account practical considerations such as robustness, scalability, and time varying wireless broadcast environment. New distributed, low overhead protocols are constructed and validated over a mesh testbed on the Purdue University campus. Undergraduates and under represented students are actively recruited for the project. Research results are disseminated via the web and through publications in conferences and journals. The results of the project, spanning from theoretic exploration to protocol deployment on a practical testbed lead to the significant performance gains necessary to realize the full capability of mesh networks, which in turn has the potential to revolutionize the telecommunications industry.


This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5). CIF:Medium:Collaborative Research: Integrating and Mining Bio Data from Multiple Sources in Biological Networks Networks are a natural, powerful, and versatile tool representing the structure of complex systems and have been widely used in many disciplines, ranging from sociology to physics to biology. Massive amounts of biological data from multiple sources await interpretation. This calls for formal information integration, modeling and mining methods. The functioning of complex biological systems demands the intricate coordination of various cellular processes and their participating components. As biological networks grow in size and complexity, the model of a biological network must become more rigorous to keep track of all the components and their interactions, and in general this presents the need for new methods and technologies in information acquisition, transmission and processing.

This research develops a set of novel methods and algorithms of integrating, analyzing and mining biological networks that include multiple sources of biomolecular information such as gene and protein expressions, interactions, and regulations, the formation and dissolution of protein complexes, and interactions between proteins and small molecules. Quantitative and qualitative bio data from multiple sources are iteratively integrated, mined and organized to generate scalable hypothetical biomolecular network structures. The dynamics of these computational hypotheses are tested and refined through dynamic simulation and laboratory experiments. The investigators will develop an integrated modeling and mining method for the biomolecular networks governing cellular processes, with natural connections to both the biological knowledge base and the design and analysis of biological experiments. We will also plan to investigate novel graph based theoretical models and algorithms for biological data represented as networks or graphs. RI: Medium: Collaborative Research: Explicit Articulatory Models of Spoken Language, with Application to Automatic Speech Recognition Proposal Title: RI: Medium: Collaborative Research: Explicit Articulatory Models of
Spoken Language, with Application to Automatic Speech
Recognition
Institution: Toyota Technological Institute at Chicago
Abstract Date: 05/22/09
This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).
One of the main challenges in automatic speech recognition is variability in speaking
style, including speaking rate changes and coarticulation. Models of the articulators
(such as the lips and tongue) can succinctly represent much of this variability. Most
previous work on articulatory models has focused on the relationship between acoustics
and articulation, but more significant improvements require models of the hidden
articulatory state structure. This work has both a technological goal of improving
recognition and a scientific goal of better understanding articulatory phenomena.
The project considers larger model classes than previously studied. In particular, the
project develops graphical models, including dynamic Bayesian networks and
conditional random fields, designed to take advantage of articulatory knowledge. A new
framework for hybrid directed and undirected graphical models is being developed, in
recognition of the benefits of both directed and undirected models, and of both
generative and discriminative training. The project activities include major extension of
earlier articulatory models with context modeling, asynchrony structures, and
specialized training; development of factored conditional random field models of
articulatory variables; and discriminative training to alleviate word confusability.
The scientific goal addresses questions about the ways in which articulatory trajectories
vary in different contexts. Existing databases are used, and initial work in manual
articulatory annotation is being extended. In addition, the project uses articulatory
models to perform forced transcription of larger data sets, providing an additional
resource for the research community. Other broad impacts include new models and
techniques with applicability to other time series modeling problems. Extending the
applicability of speech recognition will help it fulfill its promise of enabling more efficient
storage of and access to spoken information, and equalizing the technological playing
field for those with hearing or motor disabilities.
NATIONAL SCIENCE FOUNDATION
Proposal Abstract
Proposal:0905633 PI Name:Livescu, Karen
Printed from eJacket: 06/10/09 Page 1 of 1 TC: Medium: Collaborative Research: Techniques to Retrofit Legacy Code with Security This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Though perhaps unfortunate, as a practical matter software is often
built with functionality as a primary goal, and security features are
only added later, often after vulnerabilities have been identified.
To reduce the cost and increase assurance in the process of security
retrofitting, the aim to develop a methodology involving automated and
semi automated tools and techniques to add authorization policy
enforcement functionality to legacy software systems.

The main insight is that major portions of the tasks involved in
retrofitting code can be or already have been automated, so the design
process focuses on enabling further automation and aggregating these
tasks into a single, coherent approach.

More specifically, techniques and tools are being developed to: (1)
identify and label security relevant objects and I/O channels by
analyzing and instrumenting annotated application source code; (2)
insert code to mediate access to labeled entities; (3) abstract the
inserted checks into policy relevant, security sensitive operations
that are authorized (or denied) by the application s security policy;
(4) integrate the retrofitted legacy code with the site s specific
policy at deployment time to ensure, through advanced policy analysis,
that the application enforces that site s policy correctly, and (5)
verify correct enforcement of OS policy delegation by the retrofitted
application.

The techniques and tools being developed are useful not only
for retrofitting, but also for augmenting and verifying existing code
already outfitted with security functionality; hence improving the
state of the art in creating more secure software. RI: Medium: Robust Intelligent Manipulation and Apprenticeship Learning for Robotic Surgical Assistants This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

This project aims to develop models, algorithms, and testbeds for robust intelligent manipulation that will enable supervisory control for robotic surgical assistants (RSAs). Drawing on the combined expertise of the investigators, the research is focusing on both analytic and empirical approaches.

Using analytic approaches, the investigators are defining mathematical models of system state, deformable tissue dynamics, and stochastic uncertainty. These models are used to develop robotic motion planning and control algorithms to plan and perform grasping and manipulation of deformable objects under uncertainty. Using empirical approaches, the investigators are developing machine learning techniques to efficiently find dynamic control policies and characterize performance metrics based on expert demonstrations. Simulation and hardware testbeds for a common set of benchmark problems are being developed to evaluate these algorithms and methods.

This project will advance basic scientific understanding by developing new analytic methods for robust grasping and manipulation of deformable objects. This project will also advance empirical approaches by developing controllers that learn deformable object manipulation skills by observing expert demonstrations. The project will extend, compare, and evaluate analytic and empirical methods and seek to develop new hybrid methods within the focused context of providing robust intelligence for RSAs.

Robust intelligent manipulation for RSAs will improve patient health and reduce costs by enhancing surgeon performance, reducing tedium, and decreasing operation time. Outreach to local girls high schools and predominantly minority Cleveland high schools is pursued as part of the project to encourage participation of underrepresented groups in engineering and computer science. CIF: Medium: Analog Architectures for Optimization in Signal Processing Abstract:

Modern techniques giving the best performance for acquiring and processing signals/images rely on repeatedly solving mathematical optimization problems which can be computationally expensive. This research involves advancing, by orders of magnitude, the state of the art for solving an important class of these problems. Rather than developing algorithms tailored to current digital computational platform, the investigators depart completely from this current line of research to study analog architectures for solving these problems. These analog architectures, when fully developed, have the potential for dramatic gains in speed and power efficiency over their digital counterparts. This research project is inherently multidisciplinary, as it combines recent advances in computational neuroscience, signal processing, and reconfigurable VLSI architectures. Among other applications, these systems enable reductions in the time needed to acquire a magnetic resonance image (MRI).

This project focuses primarily on solving optimization programs combining a mean squared error data fidelity term with a sparsity inducing cost function (e.g., the L1 norm) via an analog dynamical system architecture. Specifically , the project contains two intertwined threads: circuit implementation and mathematical analysis.
The goal of the circuit implementation thread is to produce a analog circuit which solves significant optimization programs (e.g., tens of thousands of variables) substantially faster than state of the art digital solutions. The investigators leverage recent advances in reconfigurable analog architectures to achieve efficient designs at this large scale. The analysis thread includes deriving bounds on the circuit convergence time and generalizing the architecture to include
other relevant signal processing problems. The research also involves
applying this analog architecture as a nonlinear filter which continuously reacts to changes in the input. RI: Medium: Collaborative Research: Solving Stochastic Planning Problems Through Principled Determinization Probabilistic and decision theoretic planning, which operates under conditions of uncertainty, has important applications in science and engineering, but such stochastic methods have been under utilized because these planners do not scale to large, complicated problems.

In contrast, the past decade has seen tremendous scale up in deterministic planning techniques, which do not directly address the same challenges of uncertainty head on. This project provides a systematic framework for exploiting this progress in deterministic planning technology be it classical, temporal or partial satisfaction planning problems in stochastic planning as well, by novel adaptation of theory and/or methods of determinization, hindsight optimization, and machine learning.

This project is helping to bridge the traditional divide between deterministic and stochastic planning communities, and it is bringing stochastic planning technology to real world applications.

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5). SHF: Medium: Assurance Based Development: A Rational Approach To Creating High Assurance Software The objective of this research is to create an approach to software development for critical systems where a high level of assurance is essential to prevent failures from having serious consequences. The approach being developed, Assurance Based Development (ABD), is based on two rigorous arguments that evolve throughout development. A fitness argument shows that the system has the functional, non functional (include legal and ethical) and dependability properties necessary to satisfy all stakeholders, and a success argument shows how the development activities will yield a satisfactory system within time and budget constraints. Because these arguments capture the concerns of all stakeholders, their state at any given time reveals the obligations incident on the developers. Choosing development activities to meet these obligations facilitates early detection and avoidance of potential assurance difficulties. Choice also allows the developer to deploy expensive technology, such as formal verification, only on components whose assurance needs demand it.Evaluation and assessment of ABD is being conducted using case studies of a prototype artificial heart pump and a security critical application. TC:Medium:Collaborative Research: Mobile Personal Privacy and Security A New Framework and Technology to Account for Human Values This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Significant progress has been made on technical solutions for implementing security while preserving varying degrees of privacy for mobile electronic devices. A largely unsolved problem involves knowing what type of technical solutions to implement and under what circumstances. This challenge arises because in actuality there is in general no such thing as a clear optimal or right solution. One way to view this problem is by focusing on the fact that such devices potentially play an integral role in people s lives and impact important human values: security and privacy of course, but also other values such as autonomy, trust, and physical well being. Further, these values are often in tension. This project will develop a principled and systematic conceptual framework for analyzing these value implications and tensions in the context of two archetypical examples of future mobile devices: implantable medical devices and mobile phone safety applications. A key component of our approach will be the application and extension of the Value Sensitive Design theory and methodology. Researchers will develop and evaluate new technology, as well as undertake empirical work with a range of stakeholders, including Futures Workshops and semi structured interviews. Expected outcomes include a framework for analyzing the relationships and trade offs among privacy, security, autonomy, and other values for mobile devices that accounts for both situational and embodied dimensions of these values; a pallet of key technical solutions to these problems; a set of case studies; and finally design recommendations for use by other researchers and practitioners.

Results of this work will not only develop technical mechanisms for providing the right level of security and privacy for new mobile technologies, but can help influence the definition of what right means in the context of the broader set of goals and values. In society more generally, there is an increasing use of technologies that are focused upon here: implantable medical devices and highly capable cell phones used for personal safety. Improving human health and personal safety, while not significantly undermining other key values such as privacy, trust, and autonomy, will be important in the development of such technologies. It is anticipated that the ways of analyzing such problems will be applicable to other technologies as well, such as RFID tags on personal possessions, Smart Cards for paying for tolls or transit fares, and many others. III: Medium: Provenance Analytics: Exploring Computational Tasks and their History This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5). This proposal will address the provenance of computational processes and the data they manipulate. These are of fundamental importance in maintaining scientific process. Provenance (also referred to as audit trail, lineage, and pedigree) captures information about the steps used to generate a given data product. Such information provides documentation that is key to preserving the data and determining the data s quality and authorship as well as interpreting, reproducing, sharing and publishing results. It also seeks to produce algorithms and techniques for extracting and reusing useful knowledge embedded in workflow specifications. Workflows and workflow based systems have proven to be successful in capturing computational tasks at various levels of detail and automatically record provenance information. They have recently emerged as an alternative to ad hoc approaches to assembling computational tasks that are widely used in the scientific community. In order to effectively use provenance information and to deal with a potential information overload, we need novel tools and techniques that help users. The ability to explore the knowledge available in the provenance of computational tasks has the potential to foster large scale collaborations, expedite scientific training in disciplinary and inter disciplinary settings, as well as to reduce the lag time between data acquisition and scientific insight. RI: Medium: Algorithms for Robust Barter Exchanges, with Application to Kidneys In the US alone, over 30,000 fall sick with lethal kidney disease each year, and over 77,000 await for a kidney transplant, many more than any other transplant. That demand far exceeds the supply of cadaver kidneys. It is possible for a living person to donate a kidney, e.g., to a relative or a friend, and both can live well. Unfortunately, it is unlikely that a given donor can donate to a given patient due to blood type and tissue type incompatibilities. This opens the door for kidney exchange. Consider two donor patient pairs, A and B. Say that each of the pairs is incompatible. Yet Donor A may be able to donate to Patient B, and Donor B to Patient A. This constitutes a cycle of length two. Somewhat longer cycles are also practical. The principal investigator s prior research has proved that the exchange clearing problem is NP complete, and developed an exchange algorithm that scales to the nationwide level.

Research under this award is scaling up and expanding the functionality of the exchange algorithms to develop (a) faster exhange algorithms (e.g., through search tree reorganization), (b) online algorithms for deciding which cycles to select in light of changing donors and recipients, (c) algorithms that are robust to last minute failures, and (d) using machine learning to predict the survival duration of kidney transplants and failure probabilities of last minute tests. Some of the techniques and theory developed to address these issues also apply to other combinatorial search problems.

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5). NeTS: Medium: Collaborative Research: MIMO Pipe Modeling, Scheduling and Delay Analysis in Multi hop MIMO Networks This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The fundamental differences between multi hop networks and point to point settings indicate that leveraging MIMO gains in multi hop networks requires a paradigm shift from high SNR regimes to interference limited regimes. This project undertakes a broad research agenda centered around developing fundamental theory towards achieving optimal throughput and delay performance in wireless networks. The first key step is to take a bottom up approach for solid model abstraction of MIMO links while taking into account interference, and to extract a set of feasible rate/reliability requirements, corresponding to meaningful MIMO stream configurations. Under a common thread of MIMO pipe scheduling, this project focuses on tackling the following challenges: 1) Developing rate/reliability models for ``MIMO pipes in multi hop networks; 2) MIMO pipe scheduling for throughput maximization and delay minimization; and 3) Real time scheduling of MIMO pipes with delay constraints (for time critical traffic).

This project contributes to the formulation of new fundamental theories for multi hop MIMO networks, which have direct impacts on many wireless applications. Particularly, real time scheduling sheds much light on leveraging MIMO gains in VANET to deliver timely information reliably to save lives and improve quality of life. Underrepresented undergraduate students as well as graduate students participate in this project. NeTS: Medium:Collaborative Research: Cooperative beamforming for efficient and secure wireless communication This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

There is a growing need for wireless networks that can sustain high data rates, are robust to interference, make efficient use of battery resources, and offer secure communications. This project introduces cooperative beamforming (CB), a novel technique that enables high throughput and power efficient communications in a secure manner. CB consists of two stages. In the first stage, the sources share their data with neighboring nodes via low power communications. Various approaches for such information sharing are considered, with a goal to minimize queuing delays, conserve energy, and achieve high throughput. In the second stage, the cooperative nodes apply a weight to the signal received during first stage, and transmit. The weights are such that a specific objective criterion (e.g., signal to interference at the destination) is maximized. In CB, although each node uses low power, all nodes together can deliver high power to a faraway destination. This increase in power offsets power reduction due to propagation attenuation. CB can be viewed as an alternative to multihop transmission and, unlike multihop transmission, does not deplete the power resources of other nodes. Since CB can achieve long distance communication, new paths can be found to improve the overall network performance. Also, CB improves network security by avoiding eavesdroppers; unlike traditional cryptographic based protocols that operate at higher layers and are sensitive to the broadcast nature of the transmission medium, CB improves security at the physical layer. CB will be implemented on a hardware network testbed to demonstrate how the developed techniques can revolutionize wireless communications. CSR:Medium:Collaborative Research: Stochastically Robust Resource Allocation for Computing Stochastically Robust Resource Allocation for Computing

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Parallel, distributed, and Internet based computing, communication, and information systems are heterogeneous mixtures of machines and networks. They frequently experience degraded performance due to uncertainties, such as unexpected machine failures, changes in system workload, or inaccurate estimates of system parameters. It is important for system performance to be robust against uncertainties.
What does it mean for a computer system to be ?robust?? How can robustness be described? How does one determine if a claim of robustness is true, or if a system will fail? How can one decide which of two systems is more robust? These are the types of issues we address in this project, with our team of faculty, graduate students, and undergraduate students from Colorado State University and the University of Colorado, and colleagues in industry (DigitalGlobe) and a national laboratory (NCAR, National Center for Atmospheric Research).
We are designing models, metrics, mathematical and algorithmic tools, and strategies for (1) deriving system resource management schemes that are robust, and (2) quantifying the probability of meeting performance requirements given uncertainties. We are validating our research by working with DigitalGlobe, which supplies images to Google Maps and Microsoft Virtual Earth, and NCAR, whose research activities include the prediction of severe and catastrophic weather. The robustness concepts being developed have broad applicability, and will significantly contribute to meeting national needs to build and maintain robust information technology infrastructures. TC: Medium: Collaborative Research: User Controllable Policy Learning As both corporate and consumer oriented applications introduce new functionality and increased levels of customization and delegation, they inevitably give rise to more complex security and privacy policies. Yet, studies have repeatedly shown that both lay and expert users are not good at configuring policies, rendering the human element an important, yet often overlooked source of vulnerability.

This project aims to develop and evaluate a new family of user controllable policy learning techniques capable of leveraging user feedback and presenting them with incremental, user understandable suggestions on how to improve their security or privacy policies. In contrast to traditional machine learning techniques, which are generally configured as ?black boxes? that take over from the user, user controllable policy learning aims to ensure that users continue to understand their policies and remain in control of policy changes. As a result, this family of policy learning techniques offers the prospect of empowering lay and expert users to more effectively configure a broad range of security and privacy policies.

The techniques to be developed in this project will be evaluated and refined in the context of two strategically important domains, namely privacy policies in social networks and firewall policies.
In the process, work to be conducted in this project is also expected to lead to a significantly deeper understanding of (1) the difficulties experienced by users as they try to specify and refine security and privacy policies, and (2) what it takes to overcome these difficulties. The latter includes developing models of the types of policy modifications users can relate to and exploit as well as an understanding of the tradeoffs between usability and the number of policy modifications users are presented with. It also includes understanding how the effectiveness of user controllable policy learning is impacted by the expressiveness of underlying policy languages, modes of interaction with the user (e.g. graphical versus text based), and the topologies across which policies are deployed, NeTS:medium: Design of Dynamic Spectrum Markets for Wireless Networks This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

It is widely recognized that current wireless spectrum policy has been an impediment to the continued growth of high capacity wireless networks. This project is investigating the consequences of lifting current restrictions on spectrum allocation and ownership, and allowing for extensive spectrum markets for allocating spectrum across locations, times and diverse applications. Various market structures that may emerge in such a setting are being studied using multi disciplinary techniques including ideas from micro economics, optimization theory and wireless networking. A key issue being explored is how to define the spectrum assets that will be traded in such a market taking into account both the interference among different users of wireless spectrum and the performance of the resulting market mechanisms. This results in a characterization of the trade offs, in terms of efficiency and complexity of different asset definitions and market mechanisms. These results provide new insights into market based allocation for wireless spectrum.

Results disseminated via publications could help facilitate a transition to new market structures that may lower the barriers to entry into the wireless services market thereby facilitating competition and the introduction of new services. The infusion of the cross disciplinary ideas developed through this project into graduate classes also broadens the training of graduate students. NeTS Medium: Collaborative Research: Unifying Network Coding and Cross Layer Optimization for Wireless Mesh Networks: From Theory to Distributed Algorithms to Implementation This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Wireless mesh networks promise a flexible and cost effective solution for bringing high bandwidth low latency applications to the home. Two orthogonal but immensely attractive approaches for designing high performance mesh networks are network coding and cross layer optimization. The former exploits the interfering wireless media as a broadcast channel by ingenious information processing, while the latter intelligently allocates and shares network resources across the layers. This project investigates the long overdue and highly challenging synergistic union of network coding and cross layer optimization.

The characterization of intersession network coding (INC), coding across different network flows, is NP hard. Nonetheless, the problem can be made tractable under important practical settings. Based on new path based characterizations, this project rigorously quantifies the network capacity in presence of different practical INC schemes. A new network coding based cross layer paradigm is developed for controlling mesh networks that takes into account practical considerations such as robustness, scalability, and time varying wireless broadcast environment. New distributed, low overhead protocols are constructed and validated over a mesh testbed on the Purdue University campus. Undergraduates and under represented students are actively recruited for the project. Research results are disseminated via the web and through publications in conferences and journals. The results of the project, spanning from theoretic exploration to protocol deployment on a practical testbed lead to the significant performance gains necessary to realize the full capability of mesh networks, which in turn has the potential to revolutionize the telecommunications industry. TC: Medium: Collaborative Research: Security Services in Open Telecommunications Networks This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The nature of telecommunications networks is rapidly changing.
Commodity smart
mobile phone frameworks such as Android and Openmoko invite developers and end users to build applications, modify the behavior of the phone, and use network services in novel ways. However, while simultaneously spurring incredible innovation, the move to open systems alters the underlying performance and security assumptions upon which the network was designed. Such changes invite vulnerabilities ranging from merely vexing phone glitches to catastrophic network failures. The current infrastructure lacks the basic protections needed to protect an increasingly open network, and it is unclear what new stresses and threats open systems and services will introduce.

This research analytically and experimentally investigates defensive infrastructure addressing vulnerabilities in open cellular operating systems and telecommunications networks. In this, we are exploring the requirements and design of such defenses in three coordinated efforts; a) extending and applying formal policy models for telecommunication systems, and provide tools for phone manufacturer, provider, developer, and end user policy compliance verification, b) building a security conscious distribution of the open source Android operating system, and c) explore the needs and designs of overload controls in telecommunications networks needed to absorb changes in mobile phone behavior, traffic models, and the diversity of communication end points.

This research symbiotically supports educational goals at the constituent institutions by supporting graduate and undergraduate student research, and is integral to the security and network curricula. SHF: Medium: Heterogeneous Virtual Machine: Future Execution Environments for Heterogeneous Many core Architectures A current industry trend aimed at addressing platform performance/power requirements is to create heterogeneous manycore systems, comprised of general purpose and specialized cores designed to accelerate certain application or system functions. A second trend, designed to make it easier to map a wide variety of functions and components to manycore platforms, is platform level support for system virtualization. This research innovates, implements, and evaluates new virtualization technologies for heterogeneous manycore architectures composed of commodity general purpose and accelerator cores. The goal is to realize an efficient execution environment for composing and executing a range of computationally and data intensive applications.

The system abstractions innovated include (i) the HVM (heterogeneous virtual machine) platform abstraction for dynamic composition of resources (e.g., cores, accelerators, memory, I/O) (ii) new methods for managing heterogeneous manycore resources, including power, and (iii) specialized execution environments for optimizing accelerator interactions. These components are implicitly integrated through an execution model wherein the same abstractions and mechanisms are used to dynamically manage diverse accelerator platforms, thereby realizing our vision of freely shared and customized platform resources provided to applications. SHF: Medium: Formal Analysis of Concurrent Software on Relaxed Memory Models Programmers are increasingly designing concurrent software to effectively harness the computational power of multi processor and multi core architectures. Writing correct concurrent software is challenging, and system specific concurrency libraries are particularly vulnerable in that they are affected by the subtle and complex rules governing the relationship among reads and writes to shared memory in multi processor systems. The goal of this project is to develop technology that will assist programmers in building high performance and correct system level concurrent software with respect to precise modeling of the essential details of the underlying architecture. The investigators will explore specifications for accurate machine readable descriptions of experimental and commercial memory models in both constraint based and operational styles. To allow developers to understand subtleties of specific memory models, this project will investigate algorithms and tools for checking equivalence between two specifications and for automatically generating test programs that exhibit the differences. Tools for verifying concurrency libraries with respect to memory model specifications and for automatic insertion of memory ordering fences, will be developed and evaluated on lock free implementations of commonly used data structures. The proposed research will be integrated in a new upper level course on multiprocessor programming. III: Medium: Learning from Implicit Feedback Through Online Experimentation This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The goal of the project is to harness the information contained in users interactions with information systems (e.g., query reformulations, clicks, dwell time) to train those systems to better serve their users information needs. The key challenge lies in properly interpreting this implicit feedback and collecting it in a way that provides valid training data. Moving beyond existing passive data collection methods, the project draws on multi armed bandit algorithms, experiment design, and machine learning to actively collect implicit feedback data. Developing these interactive experimentation methods goes hand in hand with developing machine learning algorithms that can use the resulting training data, and empirical evaluations that validate the models of user behavior assumed by the algorithms.

This research will improve retrieval quality for important applications like intranet search and desktop search. Additionally, the project will provide an operational full text search engine for the Physics E Print ArXiv and potentially other digital libraries, thus forming a test bed for the research while also providing a valuable service and dissemination tool to the academic community beyond computer science. The project provides interesting and motivating research opportunities to undergrads and international exchange students, and the PIs will include relevant material into the undergraduate and graduate curriculum. Finally, following their prior work on the Support Vector Machine, SVM light (http://svmlight.joachims.org/) and an open source search engine for learning ranked retrieval functions and evaluating the learned rankings, OSMOT (http://radlinski.org/osmot/), the PIs will continue to provide easy to use software that enables research and teaching, via the project website (http://www.cs.cornell.edu/People/tj/implicit/). TC: Medium: Collaborative Research: Multi Perspective Bayesian Learning for Automated Diagnosis of Advanced Malware Contemporary Internet malware is constantly evolving and making antivirus and intrusion detection systems increasingly obsolete. It is no longer acceptable to simply rely on binary signatures for malware identification. Both current and future generations of malware will require entirely new detection strategies that can tolerate the rapid perturbations in binary structure and payload delivery mechanisms. A promising direction to this end is the use of multi perspective, behavioral oriented paradigms for malware identification. In this project, we propose a new approach to 1) automatically extract infection knowledge, based on a multi perspective, behavior oriented view, and 2) rapidly apply this gained knowledge to diagnose the presence of malware in host computer systems. For each malware family, a probabilistic profile will be automatically extracted, which captures the invariant behavioral features of its members. This envisioned knowledge extraction process should provide sufficient abstraction in its invariant behavior characterization such that future malware variants can be recognized. We also propose a Bayesian framework for diagnosing live malware infections on fielded computer systems. If successful, this research will introduce a new complementary strategy for diagnosing malware infections in ways that cannot be defeated through the current suite of antivirus countermeasures. SHF: Medium: Hardware/Software Partitioning for Hybrid Shared Memory Multiprocessors Hybrid multiprocessor architectures present an unprecedented opportunity for high performance computing through the seamless integration of large number of processors and hardware accelerators. This project addresses the research challenges in the design and exploitation of hybrid multiprocessors through innovations that span across the areas of architectures, compilers, and high performance computing. A hybrid cache coherent non uniform memory access (CC NUMA) architecture is designed that clusters CPUs, hardware accelerators, and memories to preserve locality and reduce memory latency. Partitioning models are developed to enable optimal partitioning of data among CPUs and hardware accelerators. Compiler techniques are developed for detection of parallelism, its partitioning, and assignment across CPUs and hardware accelerators. The project enables coexistence of data streaming (push data) and data fetching (pull data) mechanisms. The research benefits from detailed measurements using a 64 processor SGI Altix 4700 CC NUMA machine with FPGAs, the Intel FSB FPGA architecture accelerator, and Niveus 4000 workstation with NVIDIA GPUs.



The research has impact on large scale scientific computing. The hybrid multiprocessor technology is likely to be transferred to industry while the developed software (compilers, simulators and Hybrid SPLASH 2 benchmarks) will be distributed to researchers. The project also has impact on education and research. The SGI Altix machine is already being used in our graduate classes and further projects on hybrid parallel computing are introduced in architecture, parallel processing, and compiler classes. The project contributes to minority undergraduate education in Computer Science since UCR is recognized for its large undergraduate Hispanic population. RI: Medium Collaborative Research: Minimalist Mapping and Monitoring This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

This project addresses fundamental and challenging questions that are common to robotic systems that build their own maps and solve monitoring tasks. In particular, the work contributes to our general understanding of the interplay between sensing, control, and computation as people attempt to design systems that minimize costs and maximize robustness.

Powerful new abstractions, planning algorithms, and control laws that accomplish basic mapping and monitoring operations are being developed in this effort. This is expected to lead to improved technologies in numerous settings where mapping and monotoring are basic components.
Ample motivation is provided by technological challenges that involve searching, tracking, and monitoring the behavior of people, wildlife, and robots. Examples include search and rescue, security sweeps, mapping abandoned mines, scientific study of endangered species, assisted living, ground based military operations, and even analysis of shopping habits.

The work is particularly transformative because it lives outside of the traditional boundaries of algorithms, computational geometry, sensor networks, control theory, and robotics. Furthermore, national interest continues to grow in the direction of developing distributed robotic systems that combine sensing, actuation, and computation. By helping to break down traditional academic and scientific barriers, it is expected that the work will transform the way we think about robotics algorithms, the engineering design process, and the education of students across the robotics, computational geometry, and control disciplines. TC:Medium:Analyzing the Underground Economy This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Recent years have witnessed a dramatic change in the goals and modus operandi of malicious hackers. In particular, hackers have realized the potential monetary gains associated with Internet fraud. As a result, there has been an integration of sophisticated computer attacks with well established fraud mechanisms devised by organized crime, which, in turn, created a vibrant underground economy. This project will develop novel techniques and tools to analyze and understand the underground economy, with the ultimate goal of obtaining a comprehensive picture of the criminal process. More precisely, the underground economy will be analyzed and modeled from three different vantage points: First, the project will identify the actors participating in the underground economy and models their different roles. Second, the project will analyze the processes and interactions between different criminal actors. Third, the project will examine the infrastructure that is used by criminals to carry out their operations.

The results of this project are techniques and tools to gather information about the infrastructure of the underground economy, the involved actors, and their interactions. This information can then be used to model the underground economy, improving the understanding of its structure and processes. Such increased understanding can be leveraged to create new techniques and processes for disrupting underground activities. As a result, the broader impact of the research project has the potential to reduce the amount and severity of crime and fraud performed on the Internet, benefiting the community at large. In addition, the tools and techniques will support cyber crime law enforcement by enabling officers to identify malicious networks and ISPs to predict upcoming, significant attacks. III: Medium: Longview: Querying the Future Now This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Traditional database systems are designed to provide the user with a
clear picture of past states of the world as is represented in the
database. Recently, stream processing systems are introduced to
produce near real time answers for those applications applications
that require up to date information. Thus, we see a trend toward
shrinking the ?reality gap? to zero. But for some applications, even
real time is not good enough. There is often a desire to get out in
front of the present by delivering predictions of future events to
take advantage of opportunities or to avert calamity. Security
applications are a good example since they are typically interested in
preventing a breach rather than simply reporting that one has
happened. There is currently no database system that can effectively
serve as a generic platform to support such predictive applications.

This project aims to fill this gap by designing and building a
prototype database system called Longview to enable data centric
predictive analytics. Longview facilitates the use of statistical
models to analyze historical and current data and make predictions
about future data values and events. Users can plug new predictive
models into the system along with a modest amount of meta data, and
the system uses those models to efficiently evaluate predictive
queries.

Longview treats predictive models as first class citizens by
intelligently managing them in the process of data management and
query optimization. This involves automatically building models and
determining when and which model(s) to apply to answer predictive
queries. This also involves creating and using the proper physical
data structures to facilitate efficient model building, selection, and
execution. Longview handles both streaming and historical queries. In
fact, many streaming queries need efficient access to an archive of
past values, making it necessary to seamlessly combine both stream and
historical processing. Finally, Longview investigates white box
model support, in which the database leverages the operational
semantics and representation of models to improve performance.

Longview s goal is to make it much easier to build predictive
analytics applications in data intensive situations. Seamlessly
combining data and model management is key to make the process of
computing with predictions far easier to express and more efficient
than the current ad hoc application level approaches. The resulting
technology also allows for a better understanding and support for
user defined functions in database systems.

Longview is initially used for a real world sensor based tracking
application and a predictive web portal for easy experimentation with
different models and data sets. Further information on the project can
be found on the project web page:
http://database.cs.brown.edu/projects/longview/ NeTS:Medium:Internet Measurement in the Large Despite years of effort, the underlying structure of the Internet is still largely opaque to networking researchers. This opaque structure is also a barrier to ISP administrators trying to diagnose connectivity problems when a problem occurs, it can be difficult to identify the root cause if it is not in the ISP s own network. In our view, the opaque structure is primarily due to the lack of vantage points from which to measure the Internet: there are more than 100,000 uniquely routed prefixes, yet there are only a few hundred locations from which one can launch probes. This project addresses this lack of
vantage points by developing techniques, using special types of measurement packets, to simulate having a very much larger number of vantage points.

Specifically, the PI proposes to build these techniques into a measurement tool, work with several ISPs to have them adopt and deploy the tool, and also deploy it on an NSF and ISP supported testbed called PlanetLab. Using the tool, a number of fundamental questions about the Internet will be studied: How is the Internet evolving? Is the Internet structure self similar? What is the prevalence of asymmetric routing? Answers to these questions are needed to understand the impact of proposed changes to the Internet. The PI will also make the measurements publicly available.


This award is funded under the American Recovery and Reconstruction Act of 2009 (Public Law 111 5). TC: Medium: Collaborative Research: Rewriting Logic Foundations for Verification and Programming of Next Generation Trustworthy Web Based Systems The rewriting model of computation is simple, yet general and flexible. A system of any kind, for example, an algorithm, a database, a hardware
system, a programming language, a network protocol, a sensor network, or
the molecular biology dynamics of a cell, can be modeled by a set of
rewrite rules that
describe the systems behavior. The rewriting model of computation is
intrinsically concurrent, without any need for explicit concurrency constructs.

Any set of rules that apply to nonoverlapping system components can execute concurrently. For a large distributed system such as a network
or a cell, this means that at any given time thousands or millions of
concurrent transitions may be happening in parallel.

Maude is a language based on rewriting logic. The current Maude implementation provides a high performance rewrite engine, as well
as builtin search, unification, and model checking tools to support
execution and analysis of systems specified in Maude. To realize
inherent concurrency of rewriting, the proposed project will develop an
implementation of Maude called Distributed and

Concurrent Maude (DCMaude) that will exploit the concurrency available in multicore/multiprocessor machines, and support distributed computing
and systems programming. The rewriting semantics of Maude supports
seamless transition between concurrent and distributed execution of a
system: execution in one process, multiple processes/cores, or multiple
machines. In addition to built in strategies for concurrency, the design
of DCMaude will include a means for the programmer to control
concurrency at a high level of abstraction in a declarative way.

DC Maude will provide new methods and tools that can significantly improve both the design and the implementation of open distributed
systems, including web based systems and next generation networks.
Furthermore, by being directly based on rewriting logic, DCMaude
will close the gap between formal specifications and actual
implementations, making it possible to gain substantially higher
assurance about the formal requirements, including the security
properties, of such systems.

The project will be carried our jointly by Drs. Jose Meseguer (UIUC)
and Dr. Carolyn Talcott (SRI International). Both UIUC and SRI will
take joint responsibility for the DCMaude design. The DC Maude
implementation will be the primary responsibility of SRI International,
while the testing, benchmarking and development of DCMaude applications
will be the primary responsibility of UIUC. CSR: Medium: MAD Systems This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

We increasingly rely on MAD Services, systems that span Multiple Administrative Domains. Examples include cloud computing, peer to peer streaming, Internet routing, and wireless mesh routing. Unfortunately, when different parties control a system s nodes, the nodes may undermine the desired service either by malfunctioning or by selfishly maximizing their own benefits to the detriment of the broader system (e.g., a tragedy of the commons.)

The central thesis of this research is that any approach to building robust MAD services must account for both selfish behaviors and Byzantine malfunctions.

This research therefore brings together ideas from game theory and Byzantine fault tolerance to (1) develop the new theoretical framework for MAD distributed systems; (2) design techniques that rigorously address key practical issues in building MAD systems, such as heterogeneity, locality, and node churn; and (3) construct significant prototypes of dependable MAD systems that tolerate selfish and Byzantine behaviors and whose costs and performance are comparable to less robust systems deployed today.

This work seeks to have broader societal impact in at least two dimensions. First, it will develop the theory and practice of MAD systems at a time when crucial personal, business, and societal needs are increasingly met by MAD services. Second, we expect a significant educational impact from this work, as graduate and undergraduate students will participate in a research plan that will enhance skills in both practice and theory. AF: Medium: Center for Quantum Algorithms and Complexity Quantum computation has forced a dramatic change in our beliefs about the foundations of computer science, the security of cryptosystems, and the nature of quantum systems. This award focuses on several fundamental algorithmic questions that arise out of this viewpoint. The first is the design of new quantum algorithms. The challenge here is that the major paradigm for the design of quantum algorithms the hidden subgroup framework has recently been shown to have severe limitations in its applicability. The center will explore several approaches, including a new framework for the design of quantum algorithms via the quantum approximation of tensor networks, as well as recent work on the use of quantum algorithms for discovering hidden nonlinear structures.

Establishing the limits of quantum algorithms is equally important to the possibility of designing efficient classical cryptosystems that are immune to quantum cryptanalysis. Such post quantum cryptosystems could have an enormous practical impact well before the first working quantum computer is ever built. For this to happen it is necessary to better understand the quantum hardness of concrete classical cryptosystems such as the lattice based cryptosystems or the McEliese cryptosystem. A different approach would involve designing novel cryptosystems whose security is based on already established quantum hardness results in the hidden subgroup framework.

The center would also study fundamental questions in quantum complexity theory, including the complexity of quantum interactive proof systems. Arguably the most important challenge is proving the quantum analog of the celebrated PCP theorem. This would have wide implications including quantum hardness of inapproximability results, improved fault tolerance results for adiabatic quantum computing, as well as implications for theoretical condensed matter physics. Another fundamental question is the power of multi prover quantum interactive proof systems. Resolving whether or not this complexity class is NEXP as in the celebrated classical result MIP = NEXP, is expected to provide important insights into the nature of quantum entanglement. NeTS:Medium:Invigorating Empirical Network Research via Mediated Trace Analysis Scientific network research relies heavily on sound, empirical analysis of real world network traffic. It is often not possible to robustly validate a proposed mechanism, enhancement, or new service without understanding how it will interact with real networks and real users. Yet obtaining the necessary raw measurement data?in particular packet traces including payload?can prove exceedingly difficult. Simply put, the lack of public access to current, representative datasets significantly hinders the progress of our scientific field. Not having appropriate traces for a study can stall the most promising research. There have been extensive efforts by the community at large to change the status quo by providing collections of public network traces. However, the community?s major push to encourage institutions to release anonymized data has achieved only very limited successes. The risks involved with any release still outweigh the potential benefits in almost all environments. The lack of significant progress in this direction?despite extensive efforts?is an undeniable indication that the community needs a new approach. Towards this end, the PIs are developing in a systematic fashion a scheme that has been used informally numerous times over two decades of network research: rather than bringing the data to the experimenter, bring the experiment to the data. Past studie have packaged up an analysis for execution by somebody external who had the privileges to access network traffic out of our reach. These people crunched the traffic with our scripts and then manually verified that the output did not leak any sensitive information before passing it on to us. The PIs are establishing such mediated trace analysis as a standard approach to empirical network research. The aim is to formalize the process sufficiently to facilitate researchers tapping into a potentially broad pool of providers willing to mediate access to traces for research studies. Several large scale network environments have already confirmed to us that they consider this model a feasible approach, and are willing to participate. The main challenge to overcome is the burden the process imposes on trace providers and on the research ?development cycle?. The basic tenet is that it possible to greatly improve many of the tedious mediation steps by devising a systematic framework that accounts for the legitimate concerns of providers while reducing their effort to such a degree that it becomes practical for them to provide mediated trace analysis on a routine basis.

The key challenge is to automate the common steps of the mediation process without compromising the core requirement of the trace provider maintaining thorough control over the process. Starting with carefully examining the threats that arise, the PIs are devising a formal framework for trace mediations that will include a computational model specifically tailored to the process? unique requirements, along with a powerful suite of tools to provide extensive support for the different elements of the undertaking.

The mediation approach has the potential to broadly improve how scientists tackle network measurement studies?both opening up access to a far greater range of empirical data than is currently viable, and instilling a greater degree of scientific rigor into the process of conducting such research. By making empirical data available to many more teams of researchers than occurs today, this work will significantly broaden efforts within the field. TC: Medium: Collaborative Research: Next Generation Infrastructure for Trustworthy Web Applications Many services traditionally performed by stand alone programs running on desktop computers are being migrated to ?Web 2.0? applications, remote services that reside ?in the cloud? and are that accessed through a browser. This migration process offers a unique opportunity to re engineer the way that software is constructed, adding some extra capabilities that reduce the vulnerability of the global information infrastructure to problems such as viruses, cyber attacks, loss of privacy, and integrity violations.

With this goal in mind, this project designs and implements a next generation infrastructure for trustworthy web applications. It evolves the existing Web 2.0 technologies into a more trustworthy ?Web 2.Sec? version by introducing information labeling and strong information flow controls pervasively at the service provider, at the user?s end, and on all paths in between.

A key feature of the new Web 2.Sec architecture is that all application programs are executed on top of a virtual machine (VM) rather than directly on physical hardware. Hence the VM retains full control over the data at all times, allowing it to enforce information flow policies that guarantee confidentiality and integrity. Even a malicious or faulty program running on top of the Web 2.Sec VM cannot cause any action that would violate these policies.

A strong educational component involving both graduate and undergraduate students rounds off the project. SHF: Medium: Early Reliability Modeling and Prediction of Embedded Software Systems To build high quality software, it is increasingly important for engineers to reason quantitatively about critical system properties, such as performance and reliability, as early as possible in the design process. Ideally, these properties are assessed before significant time and cost are expended on implementation. However, making useful predictions in early design stages is difficult at best due to the interplay among many relevant factors, such as complex properties of software components, effects of the firmware, and conflicts between desired system attributes. In this project, we focus on design time evaluation of architectures with respect to one key attribute: reliability. The approach we pursue will enable an engineer to build a multi faceted, hierarchical model of a system and assess its reliability in an incremental, scalable fashion. Moreover, we particularly focus on the area of embedded systems. Embedded systems present a rich target of opportunity for this work as (1) they demand close interplay of software and execution substrate; (2) they are often unencumbered with legacy concerns, allowing easier introduction and exploration of new techniques such as the one we propose; (3) they often have stringent and complex requirements, yet are seldom approached from a software architectural perspective. An extensive evaluation of our research focuses on two measures of interest: tractability (intended to address scalability issues existing in real, complex systems) and sensitivity (intended to address issues of confidence in our predictions under numerous design time uncertainties). NeTS: Medium: AirLab: Distributed Infrastructure for Wireless Measurements This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Accurate wireless measurements are critical to the sustained growth of deployed wireless networks, as well as the successful development of wireless innovations. Because of the significant effort required to deploy wireless networks and obtain reliable measurements, current wireless traces are lacking both in breadth of environments and consistency of methodology. The AirLab project seeks to facilitate meaningful analysis of wireless networks by deploying a distributed wireless measurement infrastructure that produces consistent and comparable wireless traces over different deployment environments. AirLab provides core periodic measurements as well as user driven experiments, and a centralized repository for storing, accessing, and statistically analyzing wireless traces. The richness of these measurement datasets allows researchers to detect and confirm hidden trends, and derive statistically meaningful conclusions based on real world observations. All AirLab measurements are public and accessible by the research community, thereby lowering the barrier to entry for research and enabling researchers to innovate without the upfront expenses of deploying local wireless testbeds. The project integrates its research outcomes into the undergraduate and graduate education programs. It also proactively seeks to increase the number of women and underrepresented groups in the field. TC: Medium: Collaborative Research:Next Generation Infrastructure for Trustworthy Web Applications Collaborative Proposal 0905684 (UCI ? Franz) / 0905650 (UCSC ? Flanagan)

Next Generation Infrastructure for Trustworthy Web Applications


Abstract

Many services traditionally performed by stand alone programs running on desktop computers are being migrated to ?Web 2.0? applications, remote services that reside ?in the cloud? and are that accessed through a browser. This migration process offers a unique opportunity to re engineer the way that software is constructed, adding some extra capabilities that reduce the vulnerability of the global information infrastructure to problems such as viruses, cyber attacks, loss of privacy, and integrity violations.

With this goal in mind, this project designs and implements a next generation infrastructure for trustworthy web applications. It evolves the existing Web 2.0 technologies into a more trustworthy ?Web 2.Sec? version by introducing information labeling and strong information flow controls pervasively at the service provider, at the user?s end, and on all paths in between.

A key feature of the new Web 2.Sec architecture is that all application programs are executed on top of a virtual machine (VM) rather than directly on physical hardware. Hence the VM retains full control over the data at all times, allowing it to enforce information flow policies that guarantee confidentiality and integrity. Even a malicious or faulty program running on top of the Web 2.Sec VM cannot cause any action that would violate these policies.

A strong educational component involving both graduate and undergraduate students rounds off the project. RI: Large:Collaborative Research: Richer Representations for Machine Translation (REPS) Research in machine translation of human languages has made substantial progress recently, and surface patterns gleaned automatically from online bilingual texts work remarkably well for some language pairs. However, for many language pairs, the output of even the best systems is garbled, ungrammatical, and difficult to interpret. Chinese to English systems need particular improvement, despite the importance of this language pair, while English to Chinese translation, equally important for communication between individuals, is rarely studied. This project develops methods for automatically learning correspondences between Chinese and English at a semantic rather than surface level, allowing machine translation to benefit from recent work in semantic analysis of text and natural language
generation. One part of this work determines what types of semantic
analysis of source language sentences can best inform a translation system, focusing on analyzing dropped arguments, co reference links, and discourse relations between clauses. These linguistic phenomena must generally be made more explicit when translating from Chinese to English. A second part of the work integrates natural language generation into statistical machine translation, leveraging generation technology to determine sentence boundaries, ordering of constituents, and production of function words that translation systems tend to get wrong. A third part develops and compares algorithms for training and decoding machine translation models defined on semantic representations. All of this research exploits newly developed linguistic resources for semantic analysis of both Chinese and English.

The ultimate benefits of improved machine translation technology are easier access to information and easier communication between individuals. This in turn leads to increased opportunities for trade, as well as better understanding between cultures. This project s systems for both Chinese to English and English to Chinese are developed with the expectation that the approaches will be applied to other language pairs in the future. TC:Large: Collaborative Research: Trustworthy Virtual Cloud Computing This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Proposal#: 0910767
Collaborative Proposal #s: 0909980, 0910483, 0910653
Title: TC: Large: Collaborative Research: Trustworthy Virtual
Cloud Computing
PIs: Peng Ning, Xuxian Jiang, and Mladen Vouk
Abstract:

Virtual cloud computing is emerging as a promising solution to IT
management to both ease the provisioning and administration of complex
hardware and software systems and reduce the operational costs. With the industry?s continuous investment (e.g., Amazon Elastic Cloud Computing, IBM Blue Cloud), virtual cloud computing is likely to be a major component of the future IT solution, which will have significant impact on almost all sectors of society. The trustworthiness of virtual cloud computing is thus critical to the well being of all organizations or individuals that will rely on virtual cloud computing for their IT solutions.

This project envisions trustworthy virtual cloud computing and investigates fundamental research issues leading to this vision. Central to this vision is a new security architecture, which harnesses new opportunities and capabilities such as built in out of band system
access, processor and hardware support for trusted computing, and out of box examination by hypervisors. This project focuses on key research issues following this security architecture, including new security services that enhance the trustworthiness of virtual cloud computing, protection of management infrastructure against malicious workloads, and protection of hosted workloads from potentially malicious management infrastructure. The research will enable the adoption of virtual cloud computing for critical IT management in industry and government organizations. This project will involve both graduate and undergraduate students, and will produce open source software and tools, which will be made available to the public. TC: Large: Collaborative Research: 3Dsec: Trustworthy System Security through 3 D Integrated Hardware While hardware resources for computation and data storage are now abundant,
economic factors prevent specialized hardware security mechanisms from being
integrated into commodity parts. System owners are caught between the need to exploit
cheap, fast, commodity microprocessors and the need to ensure that critical security
properties hold.

This research will explore a novel way to augment commodity hardware after fabrication
to enhance secure operation. The basic approach is to add a separate silicon layer,
housing select security features, onto an existing integrated circuit. This 3 D Integration
decouples the function and economics of security policy enforcement from the
underlying computing hardware. As a result, security enhancements are manufacturing
options applicable only to those systems that require them, which resolves the
economic quandary. We plan to identify a minimal and realizable set of circuit level
security capabilities enabled by this approach, which can be judiciously controlled by
the software layers. This will significantly assist in reducing both the software complexity
often associated with security mechanisms and system vulnerabilities.
This research introduces a fundamentally new method to incorporate security
mechanisms into hardware and has the potential to significantly shift the economics of
trustworthy systems. A broader impact will result through collaborative and educational
activities. Graduate and undergraduate student research associates will transfer
knowledge to future teachers, researchers and Information Assurance professionals;
and project publications will provide direct technical transfer to the embedded systems
and hardware design communities. AF: Large: Collaborative Research: Random Processes and Randomized Algorithms Randomness has emerged as a core concept and tool in computation. From modeling phenomena to efficient algorithms to proof techniques, the applications of randomness are ubiquitous and powerful. Notable examples include: construction of important combinatorial objects such as expanders, rigorously establishing phase transitions in physical models, finding polynomial time algorithms for fundamental sampling problems and approximating #P hard counting problems, designing probabilistically checkable proofs (PCP s) and establishing the hardness of approximation, and discovering simpler and often faster algorithms for a variety of computational problems. In the course of these tremendous developments, several general purpose techniques have emerged, and random sampling has become a fundamental, universal tool across sciences, engineering and computation.

This project brings together leading researchers in randomized algorithms to solve hard problems in random sampling, to identify techniques, and to develop new analytical tools. The applications come from a range of fields, including complexity, physics, biology, operations research and mathematics. The most general and widely studied technique for sampling is simulating a Markov chain by taking a random walk on a suitable state space. The Markov Chain method and its application to sampling, counting and integration, broadly known as the Markov Chain Monte Carlo (MCMC) method, is a central theme of the project.

Intellectual Merit. The project focuses on applications of randomized algorithms and random sampling to rigorously address problems across several disciplines. Within computer science these topics include: massive data sets, where sampling is critical both for finding low dimensional representations and clustering; routing networks, where sampling has many applications from monitoring and path allocation to optimization; machine learning; and property testing. Recent interactions between computer science and other scientic disciplines have led to many new rigorous applications of sampling, as well as new insights in how to design and analyze efficient algorithms with performance guarantees; for instance, phase transitions in the underlying physical models can cause local Markov chains to be inefficient. The project explores deeper connections between physics and random sampling, including conjectured correlations between reconstruction problems and thresholds for the efficiency of local algorithms. Many related problems arise in biology, such as phylogenetic tree reconstruction and analysis of complex biological networks. In nanotechology, models of self assembly are simple Markov chains. In mathematics, the techniques used in the analysis of sampling algorithms in general and Markov chains in particular have drawn heavily on probability theory, both discrete and continuous.

Broader Impact. The college of computing at Georgia Tech is home to the new Algorithms and Randomness Center (ARC) with many faculty and students sharing this expertise. The project s activities include designing a summer school for graduate students in randomized algorithms and designing a course for training students from diverse backgrounds and hosting workshops focusing on both theoretical and applied aspects of randomized algorithms. Participation of women and under represented groups in all of these activities will be encouraged, and the workshops will include tutorials to increase accessibility. These coordinated efforts in education and research will solidify the impact of ARC and make it a premier center for algorithms, randomness and complexity. TC: Large:Collaborative Research: Combining Foundational and Lightweight Formal Methods to Build Certifiably Dependable Software This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

This research focuses on combining foundational and lightweight formal
methods to verify the safety, security, and dependability of
large scale software systems. Foundational approaches (to formal
methods) emphasize expressiveness and generality, but they have not
been successfully scaled to large scale systems. Lightweight
approaches emphasize automation and scalability, but they are less
effective on low level programs and on providing explicit proof
witness. The PIs are developing new high level and low level
lightweight methods and adapting them into various foundational
verification systems. More specifically, at the high level, the PIs
are developing lightweight dependent type systems and
separation logic shape analysis, and connecting current methods such as
conventional type systems, slicing analysis, and flow analysis; at the
low level, the PIs are designing specialized provers and decision
procedures to verify low level programs and to support certified
linking of heterogeneous components. If successful, this research will
dramatically improve the scalability of foundational verification
systems and provide new powerful technologies for building trustworthy
software systems. Combining lightweight and foundational approaches
also provides a common thread that will pull different verification
communities together and unify them into a single framework. RI: Large: Collaborative Research: Understanding Uncertainty in Rats and Robots Abstract
Humans, rats and other vertebrates, relying on their advanced nervous systems, are far superior at dealing with the uncertainties of the world than are artificial systems. Thus, a machine, whose behavior is guided by a neurobiologically inspired system, might demonstrate the flexible, autonomous behavior normally attributed to biological organisms. Biological organisms have the ability to respond quickly to an ever changing world. Because this adaptability is so critical for survival, all vertebrates have sub cortical structures, which comprise the neuromodulatory systems, to handle uncertainty and change in the environment. Attention, which is influenced by neuromodulation, plays a significant role in animal s ability to respond to such changes. Different neuromodulatory systems are thought to play important and distinct roles in attention. A collaborative approach, which compares rodent experiments with robots having simulated nervous systems, will examine these attentional systems. These experiments will lead to a better understanding of how animals cope with uncertainty in the environment, and will lead to the design of a robot capable of flexible and complex behavior. This work has the potential of being paradigm shifting technology that could find its way in many practical applications.

In an interdisciplinary approach, a robotic system, whose design is based on the vertebrate neuromodulatory system and its effect on attention, will be constructed and tested under similar experimental conditions to the rat, and then in a more practical application. This approach, which combines computational modeling and robotics with rodent behavioral and electrophysiological experiments, will lead to a better understanding of how areas of the brain allocate attentional resources and cause the organism to respond rapidly to essential events and objects. Two of these neuromodulatory systems, the cholinergic and noradrenergic, are thought to play important and distinct roles in attention. Expected uncertainty, the known degree of unreliability of predictive relationships in the environment, drives activity within the cholinergic system. Unexpected uncertainty, large changes in the environment that violate prior expectations, drives activity within the noradrenergic system. These systems modulate activity in brain areas to properly allocate the attention to stimuli in the environment necessary for adequate learning to occur and fluid behavior to be maintained. This knowledge will be used to construct a robust, intelligent robotic system whose capability to adapt to change, and behave effectively in a noisy, complex environment will rival that of a biological system. SHF:Large:Collaborative Research:TRELLYS: Community Based Design and Implementation of a The cost effective construction of functionally correct software systems remains an unmet challenge for Computer Science. Although industrial best practices for software construction (such as testing, code reviews, automatic bug finding) have low cost, they cannot provide strong guarantees about correctness. Classical verification methods, on the other hand, are not cost effective. Recently, the research community has been exploring the idea of dependent types, which extend the expressive power of programming languages to support verification. These rich types allow the programmer to express non trivial invariant properties of her data and code as a part of her program. That way, verification is incremental, localized and at source language level.

This multi institution collaborative project is for the design and implementation of a programming language with dependent types, called Trellys. Technically, Trellys is call by value functional programming language with full spectrum dependency. Overall, the project combines numerous fragmented research results into a coherent language design, by building a robust open source implementation. The design draws on diverse solutions to the technical problems that arise from extending traditional programming languages accommodate dependent types: type and effect inference, language interoperability, compilation, and concurrency. SHF: Large: Collaborative Research: PASS: Perpetually Available Software Systems Despite heroic efforts in testing, static analysis, specification, and verification, all real world software desktop applications, servers, and transportation systems deploys with defects and missing functionality, costing the US economy billions and threatening our well being. This project proposes a transformative paradigm shift to perpetually available software systems (PASS) that will make software more available and robust by directly addressing errors in deployed software. PASS innovations will (1) improve user experience by keeping real world software running longer; (2) ensure good performance; (3) assist developers in fixing errors while allowing patches to be safely deployed on running software, to avoid downtime. The project will mine error reports in open source software repositories to derive error classes and test suites. It will evaluate system effectiveness by comparing with bug reports and patches in repositories. Innovations will include (1) detection and remediation elements that target common errors, (2) semantic foundations for remediation and on line updating, and (3) integration of elements to exploit synergy among the components. The project will explore and analyze novel safe, probabilistically safe, and extended semantics remediations/updates. The project will develop both C/C++ and Java runtimes, because they are the most widely used languages and pose unique challenges. Methods will include combining dynamic, static, and remediation/update analysis and results. The project will train graduate, undergraduate, and post doctoral students, and participate in outreach to under represented groups. The tools will be made publicly available, adding to the national research infrastructure. TC: Large: Collaborative Research: AUSTIN An Initiative to Assure Software Radios have Trusted Interactions TC:Large: Collaborative Research: AUSTIN?An Initiative to Assure Software Radios have Trusted Interactions

Software and cognitive radios will greatly improve the capabilities of wireless devices to adapt their protocols and improve communication. Unfortunately, the benefits that such technology will bring are coupled with the ability to easily reprogram the protocol stack. Thus it is possible to bypass protections that have generally been locked within firmware. If security mechanisms are not developed to prevent the abuse of software radios, adversaries may exploit these programmable radios at the expense of the greater good.
Regulating software radios requires a holistic approach, as addressing threats separately will be ineffective against adversaries that can acquire, and reprogram these devices. The AUSTIN project involves a multidisciplinary team from the Wireless Information Network Laboratory (WINLAB) at Rutgers University, the Wireless@Virginia Tech University group, and the University of Massachusetts. AUSTIN will identify the threats facing software radios, and will address these threats across the various interacting elements related to cognitive radio networks. Specifically, AUSTIN will examine: (1) the theoretical underpinnings related to distributed system regulation for software radios; (2) the development of an architecture that includes trusted components and a security management plane for enhanced regulation; (3) onboard defense mechanisms that involve hardware and software based security; and (4) a algorithms that conduct policy regulation, anomaly detection/punishment, and secure accounting of resources.
Developing solutions that ensure the trustworthy operation of software radios is critical to supporting the next generation of wireless technology. AUSTIN will provide a holistic system view that will result in a deeper understanding of security for highly programmable wireless devices. TC:Large: Collaborative Research: AUSTIN An Initiative to Assure Software Radios have Trusted Interactions TC:Large: Collaborative Research: AUSTIN?An Initiative to Assure Software Radios have Trusted Interactions (CNS 0910557)

Software and cognitive radios will greatly improve the capabilities of wireless devices to adapt their protocols and improve communication. Unfortunately, the benefits that such technology will bring are coupled with the ability to easily reprogram the protocol stack. Thus it is possible to bypass protections that have generally been locked within firmware. If security mechanisms are not developed to prevent the abuse of software radios, adversaries may exploit these programmable radios at the expense of the greater good.
Regulating software radios requires a holistic approach, as addressing threats separately will be ineffective against adversaries that can acquire, and reprogram these devices. The AUSTIN project involves a multidisciplinary team from the Wireless Information Network Laboratory (WINLAB) at Rutgers University, the Wireless@Virginia Tech University group, and the University of Massachusetts. AUSTIN will identify the threats facing software radios, and will address these threats across the various interacting elements related to cognitive radio networks. Specifically, AUSTIN will examine: (1) the theoretical underpinnings related to distributed system regulation for software radios; (2) the development of an architecture that includes trusted components and a security management plane for enhanced regulation; (3) onboard defense mechanisms that involve hardware and software based security; and (4) a algorithms that conduct policy regulation, anomaly detection/punishment, and secure accounting of resources.
Developing solutions that ensure the trustworthy operation of software radios is critical to supporting the next generation of wireless technology. AUSTIN will provide a holistic system view that will result in a deeper understanding of security for highly programmable wireless devices. III: Small: Medieval Unicorn: Toward Enhanced Understanding of Virtual Manuscripts on the Grid in the Twenty First Century Computation has revolutionized the study of historic documents and is growing rapidly in terms of use and importance to interdisciplinary technology/domain groups and individual scholars and researchers. An interdiscipinary team at the University of Illinois proposes to research and develop cyber tools for exploratory visual studies and quantitative analyses of large volumes of midieval manuscipts. The project will focus on visual imagery embedded in thes manuscripts and where the analysis tools will require large amounts of computational resources and scalable algorithms. The test collection will be digitized copies of Froissart s Chronicles, a set of midieval manuscripts available for reserch and teaching over the Worldwide Universities Network grid and accessible through Virtual Vellum ( developed at the University of Sheffield, UK, and funded by the UKs Arts and Humanities and Engineering and Physical Sciences experimental e Science program.) The broader scientific impacts resulting from the proposed activities are expected to be in new methodologies, scalable algorithms. The work will also provide new exploratory frameworks that will support questions related to studying broad, difficult and complex topics such as the composition and structure (codicology) of manuscripts as cultural artifacts of the book trade in later medieval Paris and identifying the characteristic styles and iconographic signatures of particular artists. The research will contribute to a body of recent scholarship that seek to define how books were made, how they circulated, and what their cultural value was in the late medieval period. CSR:Large:Collaborative Research:Reclaiming Moore s Law through Ultra Energy Efficient Computing This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Moore?s law promises consistent increasing transistor densities for the foreseeable future. However, device scaling no longer delivers the energy gains that drove the semiconductor growth of the past several decades. This has created a design paradox: more gates can now fit on a die, but cannot actually be used due to strict power limits. In this project, we will address this energy crisis through the universal application of ?near threshold computing? (NTC), where devices operate at or near their threshold voltage to obtain 10X or higher energy efficiency improvements. To accomplish this we focus on three key challenges that to date have kept low voltage operation from widespread use: 1) 10X loss in performance, 2) 5X increase in performance variation, and 3) 5 orders of magnitude increase in functional failure. We present a synergistic approach combining methods from algorithm and architecture levels to the circuit and technology levels. We will demonstrate NTC for applications that range from sensor based platforms which critically depend on ultra low power (¡ÜmW) and reduced form factor (mm3) to unlock new applications, to high performance platforms in large data centers, which dissipate so much power that they require co location near dedicated cooling facilities. Our end goal is to reduce national energy consumption and environmental impact by providing dramatic gains in energy efficiency while also opening up new application areas in health care by providing for in situ monitoring of biological functions with minimum intervention. TC:Large:Collaborative Research:Combininig Foundational and Lightweight Formal Methods to Build Certifiably Dependable Software This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5)

This research focuses on combining foundational and lightweight formal
methods to verify the safety, security, and dependability of
large scale software systems. Foundational approaches (to formal
methods) emphasize expressiveness and generality, but they have not
been successfully scaled to large scale systems. Lightweight
approaches emphasize automation and scalability, but they are less
effective on low level programs and on providing explicit proof
witness. The PIs are developing new high level and low level
lightweight methods and adapting them into various foundational
verification systems. More specifically, at the high level, the PIs
are developing lightweight dependent type systems and
separation logic shape analysis, and connecting current methods such as
conventional type systems, slicing analysis, and flow analysis; at the
low level, the PIs are designing specialized provers and decision
procedures to verify low level programs and to support certified
linking of heterogeneous components. If successful, this research will
dramatically improve the scalability of foundational verification
systems and provide new powerful technologies for building trustworthy
software systems. Combining lightweight and foundational approaches
also provides a common thread that will pull different verification
communities together and unify them into a single framework TC: Large:Collaborative Research: AUSTIN An Initiative to Assure Software Radios have Trusted Interactions TC:Large: Collaborative Research: AUSTIN An Initiative to Assure Software Radios have Trusted Interactions

Software and cognitive radios will greatly improve the capabilities of wireless devices to adapt their protocols and improve communication. Unfortunately, the benefits that such technology will bring are coupled with the ability to easily reprogram the protocol stack. Thus it is possible to bypass protections that have generally been locked within firmware. If security mechanisms are not developed to prevent the abuse of software radios, adversaries may exploit these programmable radios at the expense of the greater good.
Regulating software radios requires a holistic approach, as addressing threats separately will be ineffective against adversaries that can acquire, and reprogram these devices. The AUSTIN project involves a multidisciplinary team from the Wireless Information Network Laboratory (WINLAB) at Rutgers University, the Wireless@Virginia Tech University group, and the University of Massachusetts. AUSTIN will identify the threats facing software radios, and will address these threats across the various interacting elements related to cognitive radio networks. Specifically, AUSTIN will examine: (1) the theoretical underpinnings related to distributed system regulation for software radios; (2) the development of an architecture that includes trusted components and a security management plane for enhanced regulation; (3) onboard defense mechanisms that involve hardware and software based security; and (4) a algorithms that conduct policy regulation, anomaly detection/punishment, and secure accounting of resources.
Developing solutions that ensure the trustworthy operation of software radios is critical to supporting the next generation of wireless technology. AUSTIN will provide a holistic system view that will result in a deeper understanding of security for highly programmable wireless devices. CIF: Large: Sensing Sensors: Compressed sampling with Co design of Hardware and Algorithms Across Multiple Layers in Wireless Sensor Networks This research program will develop and demonstrate theory and techniques for efficient collection, transmission and processing of information in wirelessly connected sensor networks. This fundamental research is aimed towards development of a revolutionary wireless sensor node, optimized for infrastructure monitoring, and characterized by ultra low power consumption. Power consumption, installation complexity and installation cost are significant bottlenecks to the widespread deployment of wireless infrastructure monitoring that will be addressed by this research. This research program improves energy efficiency and battery lifetime through the use of compressed sampling in sensing, physical communication and network communication and through the co design of hardware and algorithms.

A theme of the research is the development of compressive sampling techniques. Compressive sampling techniques will be developed to reduce the number of samples that need to taken by sensors, to optimize the placement of sensing nodes, and to minimize the amount of data transmitted by the sensing nodes. This research will develop new theoretical approaches to signal sampling, to ultra wideband (UWB) wireless communication of acquired information, and to sensor network architecture and protocols. Development of new theory and algorithms will go hand in hand with the development of new circuit techniques. New integrated circuit techniques for analog to digital conversion and wireless communication will be developed.

A Center for Structural Health Monitoring through Compressed Sensing will be formed to broaden the impact of this research. The scope of the education component of this program targets grade school, undergraduate and graduate students. An interdisciplinary curriculum of courses in Math, Civil Engineering and Electrical Engineering will be created. The center will also develop an extensive outreach education program. CSR: Large: Collaborative Research: Multi core Applications Modeling Infrastructure (MAMI) This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The burgeoning revolution in high end computer architecture has far reaching implications for the software infrastructure of tools for performance measurement, modeling, and optimization, which has been indispensable to improved productivity in computational science over the past decade. The heart of the problem is that new multicore processors are the foundation of next generation systems, ranging from workgroup clusters to petascale supercomputers. The main motivation by chip manufacturers for the movement to multicore processors is better performance per watt than the traditional single core processor. Hence, multicore processors are not equivalent to multiple CPUs that traditional tools addressed. While significant work is underway on understanding performance tradeoffs with multicores, much of this work is ad hoc and needs a unifying framework to which the community can contribute in a systematic manner. Furthermore, little work has been done on understanding performance power tradeoffs in supercomputer systems for large scale applications. It is important to understand performance and performance power tradeoffs in the context of the significant resource sharing that occurs in multicore systems.

This proposal is focused on developing the Multicore Application Modeling Infrastructure (MAMI) that will facilitate systematic measurement, modeling, and prediction of performance, power consumption and performance power tradeoffs in multicore systems. In addition to developing MAMI, the proposed work will use MAMI to model, analyze and optimize performance and power consumption of key benchmarks and applications on multicore systems. TC: Large: Trustworthy Information Systems for Healthcare (TISH) This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Technology infrastructure in the healthcare realm requires secure and effective systems to meet two of its most significant challenges of the 21st century: improving the quality of care and controlling costs. Yet developing, deploying and using information technology that is both secure and genuinely effective in the complex clinical, organizational and economic environment of healthcare is a significant challenge. This project s multidisciplinary approach will develop and analyze information sharing technology that ensures security and privacy while meeting the pragmatic needs of patients, clinical staff, and healthcare organizations to deliver efficient, high quality care.

Dartmouth s Trustworthy Information Systems for Healthcare (TISH) Program addresses fundamental challenges in current and emerging areas of information security in healthcare: protecting the security of clinical information, while ensuring that clinicians can access information they need, when and where they need it, and securing the collection of sensor data through personal sensor devices (including both physiological and activity data) to enable monitoring of patient outcomes while giving patients usable control over their privacy.

To be effective, such technologies must consider the economic, organizational and sociological dynamics that are critical to creating and implementing IT that is secure as well as usable and effective. Thus, the researchers will consider: usability, and its implications for secure information sharing throughout the organizational environment of medical care; privacy concerns, examining how key stakeholders (patients, clinicians, and other providers) understand and evaluate the trade offs between information sharing, usability, security, and privacy: economic risks, identifying the economically motivated threats to security and privacy in healthcare information systems and the incentives for adopting security technology; and security challenges related to collection, processing, and medical use of data from sensors worn by outpatients

This research will result scientific advances in several fields. The team will develop new secure and efficient protocols that allow remote health monitoring through a mobile phone and wearable wireless medical sensors; design new machine learning methods, especially active learning and relational learning techniques for analyzing and summarizing sensor data in a user friendly manner; seek a deeper understanding of the economics of information security in healthcare; and explore how patients and clinicians trade off usability, security, and privacy.

TISH, a collaborative project of the Institute for Security, Technology, and Society at Dartmouth College, involves a team from Computer Science, Sociology, the Dartmouth Medical School, and the Tuck School of Business. Its researchers include computer scientists with expertise in computer security and sensor networks, sociologists with expertise in healthcare organizations and health policy, a clinician with medical informatics expertise, and a business professor with expertise in the economics of information security. The TISH team will work in collaboration with local hospitals and health systems in the Upper Valley region of Vermont and New Hampshire. TC: Large: A Formal Platform for Analyzing Internet Routing Trustworthy Reactive Routing Systems

Warren A. Hunt, Jr. and Sandip Ray
Department of Computer Sciences
University of Texas at Austin

{hunt,sandip}@cs.utexas.edu

Reactive concurrent systems, such as routers, consist of a number of
interacting processes that perform non terminating computations
while receiving and transmitting messages. The design of such
routing systems is error prone, and non determinism inherent in
concurrent communications makes it hard to detect or diagnose such
errors. Our effort will develop ACL2 based tools for ensuring
trustworthy execution of large scale reactive routing systems.

We aim to develop a scalable, mechanized infrastructure for
certifying correct and secure execution of reactive routing system
implementations through: a generic framework for modeling and
specifying systems at a number of abstraction layers; a
compositional methodology for mathematically analyzing such models;
and developing a suite of tools and techniques to mechanize and
automate such analysis within a unified logical foundation. Our
research exploits ACL2 s general purpose reasoning engine while
augmenting the tool suite with a streamlined modeling and
specification methodology. We will develop a collection of targeted
tools for verifying safety, liveness, and security properties of
such systems while staying within a single logic and proof system.

To facilitate verification of correspondence between protocol
layers, we propose to enhance ACL2 s reasoning engine with automated
verification tools based on advances in BDD and SAT based
techniques. The invention and proof of inductive invariants is one
of the most time consuming activities in reactive system
verification, and we will integrate into ACL2 a general purpose
symbolic simulation capability; this technique can symbolically
simulate system models over a large number of computation steps,
thereby often obviating the construction of single step inductive
invariants. The expected results will help automate the mechanical
verification of reactive systems such as routers and CPSs. AF: Large: Networks, Learning and Markets with Strategic Agents An active line of algorithmic research over the past decade has developed techniques for analyzing systems of self interested agents. A crucial challenge in analyzing such systems is to predict aggregate properties at large scales; this requires drawing conclusions about global phenomena in systems whose behavior is currently only well understood at the level of individual agents or pairs of agents. Deriving conclusions about macroscopic properties of systems described at a microscopic level is important in both computing and the social sciences. Market prices, for instance, arise from the microscopic interactions of individual traders. Understanding both the normal functioning of markets and their failure requires methods that can bridge the gap between these different scales of resolution.

This project uses ideas about networks and learning from algorithmic game theory to bridge the micro macro gap. The research on networks considers theories of bargaining and trade in which participants are constrained by a network structure. This includes models for the distribution of power among agents in a network, as well as models in which prices in a market arise strategically through the interaction of market making intermediaries in a network. The research develops models of market failures, particularly the kinds of cascading breakdowns of trust that played a crucial role in the global financial crisis in 2008. The project employs learning models to capture how perceived counterparty risk the ability of one s trading partner to complete a transaction spreads through a market.

The research on trust in financial markets can potentially contribute to broader policy debates about methods for restoring trust in markets. Currently there is a lack of analytical techniques that can tractably manipulate non trivial learning dynamics to uncover the resulting network level consequences, such as cascades. The research will provide tools for analyzing the determinants and evolution of trust in financial markets.

The project will also inform the development of introductory courses that cut across many disciplines, providing undergraduates from a wide range of backgrounds with a computationally grounded perspective for reasoning about the behavior and consequences of networks of interacting agents. RI: Small: A Microgripper with Concurrent Actuation and Force Sensing The goal of this work is to develop a better fundamental understanding of the actuation and force sensing required for micromanipulation (objects from 5 500 microns) by exploring the following hypothesis: it is possible to manipulate micro sized rigid and flexible objects while measuring physical properties, such as stiffness, by using the same microfinger to act simultaneously as an actuator and sensor. This hypothesis is explored both theoretically and experimentally by creating a set of smaller and smaller microgrippers. Two models, one analytical, the other numeric, of this new microfinger will be developed to predict performance as a function of size. Experiments using actual microgrippers verify the quality of the model.

The work impacts science, education, and outreach to minority populations. A compliant microgripper is an enabling technology to manipulate flexible and fragile bio objects for applications in bioengineering, microbiology and genomics. A microgripper squeezes an oocyte to determine its viability by measuring the stiffness of the cell before subsequent injection of DNA or RNA, turning tedious manual procedures into programmed, automatic sequences, thereby reducing cost. New ways of detecting diseases, such as malaria, by measuring physical properties of cells with the microgripper become possible. Broader impacts include education and outreach. Great emphasis is spent on having supported students give talks about this technology at local middle schools and high schools to entice the next generation to become engineers. These talks improve the professional capabilities of the graduate students while simultaneously demonstrating to young students the purpose of studying math and science. CSR: Small: Core Scheduling to Improve Virtualized I/O Performance on Multi Core Systems This project focuses on reducing the overhead and increasing throughput of network processing in multi core platforms. In particular, packet processing functions are proposed to be balanced across the cores so as to facilitate virtualization in next generation systems.

Low cost multi core architectures that put many CPU cores on the same chip are abundantly available in the market today. Researchers are developing a range of programming techniques for different applications to efficiently utilize the parallelism available in such multi core architectures. However, research into how to alleviate the I/O bottleneck, where protocol processing overhead dominates the CPU execution time, is sparse.

Multicore has enabled broad interest in virtualization for diverse uses including server consolidation and sharing of various resources. Studies have shown that virtualization brings significant extra overhead to network I/O. The objective of this project is to develop techniques to optimize the performance of virtualized I/O with high speed networks. In particular, the research team explores new software techniques in virtualized environments, that may reduce the network I/O overhead in multi core processors, through the following approaches:

1. Life of a Packet Analysis: this involves a measurement technique to trace the life of a packet in a virtualized environment with 10 Gigabit Ethernet. The study instruments the OS software, and should reveal potential bottleneck functions that contribute heavily to packet latency.

2. Mutithreading the protocol stack: Based on life of a packet analysis, the TCP/IP protocol stack in the guest O/S and virtual machine monitor (VMM) will be divided into multiple threads that can execute in parallel on multiple cores and cut down the latency. Core scheduling techniques are developed to allocate these threads to different cores so as to exploit the cache locality of the multi core architecture.

3. Pipeline Scheduling: Instead of splitting the protocol stack in terms of latency bottleneck, tasks are partitioned based on code size and multiple threads developed. Techniques are developed to schedule the threads appropriately so that the cache misses are reduced.

4. Combined Scheduling for Virtualized Environment: Although the parallel/pipeline techniques are developed separately from the TCP/IP stack and VMM, they are combined to create multiple threads in a virtualized environment and various code scheduling optimizations are applied to reduce latency and increase I/O throughput.

A complimentary project to the one described here has been partly supported by grants from the Intel Corporation. Hence, the research results obtained from this project may have strong potential for technology transfer. The PI has mentored several Ph.D. graduates who later developed reputations for architecture research and he has mentored four female Ph.D. graduates during the last two years, contributing to increasing the representation of women in computing in the country. Such efforts continue under this NSF project. UCR is recognized as a minority serving institution. Hence, involving undergraduate students will enable minority participation in the project. RI: Small: Qualitative Preferences: Merging Paradigms, Extending the Language, Reasoning about Incomplete Outcomes The most common approach in decision theory involves preferences expressed numerically in terms of utility functions, while optimization over different choices takes into account the probability distribution over possible states of the world. An alternative approach represents preferences in qualitative terms, and is motivated, in part, by difficulties in building good utility functions, ascertaining accurate probability distributions, and related problems.

This project is advancing qualitative decision theory by focusing on two promising formalisms for representing and reasoning about qualitative preferences: conditional preference networks (CP nets) and answer set optimization (ASO) programs. Both CP nets and ASO programs offer representations for several classes of preference problems, but each has major limitations. This project addresses these limitations by developing a formalism. ASO(CP) programs, which extend both ASO programs and CP nets by exploiting key properties of both. The project s major objectives are: to introduce formally ASO(CP) programs by integrating into ASO programs generalized conditional (ceteris paribus) preferences of CP nets; to establish expressivity of ASO(CP) programs, to study properties of preorders that can be defined by means of ASO(CP) programs, and to address relevant computational issues; to investigate a crucial problem of preference equivalence, essential for automated preference manipulation; to study an extension of ASO(CP) programs to the case of incompletely specified outcomes, essential for practical applications; and, to extend ASO(CP) programs to the first order language extended with aggregate operators.

Representing preferences qualitatively and optimizing over such preferences is a fundamental problem of qualitative decision theory. By integrating and advancing understanding of major types of common preferences that are captured by ASO programs and CP nets, this project will produce theoretical and practical advances in representation and reasoning about preferences, bringing it to the point where it can be effectively used in practical decision support systems. RI: Small: A Computational Framework for Marking Physical Objects against Counterfeiting and Tampering The project?s goal is to create a science for embedding information in physical objects whose manufacturing process is inherently imprecise. The team will reach their objective through the investigation of a specific problem of significant economic importance: thwarting counterfeiting, and the related problem of physical tamper detection. Counterfeiting is a growing economic problem that has been called the ?crime of the century? by a recent manufacturing industry report, and its cost is rapidly escalating (its yearly cost to the automotive industry alone is in the tens of billions of dollars and the loss of about 250,000 U.S. jobs for the legitimate manufacturers). In terms of scientific impact, the project has excellent potential for launching a significant new field. While the marking of digital objects is a well explored area, the creation of algorithms for placing marks in physical objects is mostly unexplored territory. In terms of industrial impact, the project also has excellent potential for profoundly improving the current ?state of the practice?, which is all too easily defeated by sophisticated counterfeiters. This is because the project?s framework assumes an adversary with powerful capabilities, such as greater manufacturing prowess than the legitimate manufacturer, and full knowledge of the algorithms used to embed marks and to read marks (i.e., no ?security through obscurity?). The project?s only assumption is that the adversary does not know a secret key used to embed the mark. The work focuses on the development of the computational algorithms necessary to resolve several difficult issues and tradeoffs about what information to embed in the object, and where/how to embed it. A solution must not increase manufacturing cost and must be usable with the current manufacturing pipeline. The approach is inherently multidisciplinary, combining information hiding, computer vision/graphics, and robust algorithms. Hence, students in the project will acquire a unique combination of skills. RI: Small: Probabilistic Networks for Automated Reasoning The ultimate aim of this investigation is to develop computer systems capable of operating autonomously in dynamic and uncertain environments. The investigation comprises theoretical and experimental studies aiming at fusing causal and counterfactual relationships on top of probabilistic information to improve tasks of diagnosis, situation assessment and decision making in data intensive applications. Specific research issues to investigate include developing algorithms and conceptual tools for diagnosis and situation assessment, using causal and counterfactual relationships, and developing novel methods of structure learning using propensity score estimation and C equivalence tests, with applications to epidemiology and molecular biology.

The theoretical part of this research touches on the core of human knowledge and scientific inquiry, and is having profound methodological ramifications in the fields of epidemiology, social science, economics, medicine and biology, where causal knowledge plays a major role. The practical part focuses on the development of new algorithms for causal and counterfactual reasoning, decision making and structure learning, with direct applications in areas that use graphical methods to encode knowledge. CSR: Small: An Information Accountability Architecture for Distributed Enterprise Systems Personally identifiable or sensitive information (PII) has become a target of attackers seeking financial gain through its misuse. With the trend toward storing and processing PII on complex and insecure systems, the need for improved protection has become a goal of enterprise policy and legislative efforts. In this project, we investigate Concatenated Dynamic Information Flow Tracking (CDIFT), an architecture for performing dynamic information flow analysis at various system levels and across multiple processes in a distributed enterprise. CDIFT will allow administrators to ?map? the enterprise business logic (applications, network, storage) and determine where information of interest is stored or transmitted. The same mechanism can also be used to enforce an information flow policy, restricting where and by whom such information can be viewed. CDIFT will complement and enhance current compliance and auditing efforts, which require considerable recurrent effort and a large number of man hours spent by administrators and auditors on understanding existing systems.

We will develop and experimentally evaluate novel techniques for conducting fine grained tracking of information of interest (as defined by the system operator or, in the future, by end users, in a flexible, context sensitive manner) toward mapping the paths that such information takes through the enterprise and providing a means for enforcing information flow and access control policies. Our hypothesis is that it is possible to create efficient fine grained information tracking and access control mechanisms that operate throughout an enterprise legacy computing infrastructure through appropriate use of hypervisors and distributed tag propagation protocols. CSR: Small: A Unified Reinforcement Learning Approach for Autoconfiguration of Virtualized Resources and Appliances URL: A Unified Reinforcement Learning Approach for Auto configuration of
Virtualized Resources and Appliances

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Cloud computing, unlocked by virtualization, is emerging as an increasingly important service oriented computing paradigm. The goal of this project is to develop a unified learning approach, namely URL, to automate the configuration processes of virtualized machines and applications running on the virtual machines and adapt the systems configuration to the dynamics of cloud. The URL approach features three innovations: First is a reinforcement learning (RL) methodology for auto configuration of virtual machines (VMs) on distributed computing resources in a real time manner. Second is a unified RL approach for auto configuration of both VMs and multi tier web appliances. It is able to adapt the VM resource budget and appliance parameter settings in a coordinated way to the cloud dynamics and the changing workload for the objective of service quality assurance. Third is a distributed, cooperative RL approach that allows the RL based learning and optimization agents running on different servers and with independent action choices to make an optimal joint configuration policy in large scale systems.

Deliverables that emerge from this project will advance discovery and understanding of autonomic management of large scale complex systems with profound technical, economic, and societal impact. In addition, this project has an integral educational component. It will raise the level of awareness of system management issues and the power of machine learning technology, and prepare the students to enter the industry with adequate understanding of the challenges and opportunities in cloud computing. CIF:SMALL: Nonlocal Sparse Representations on Graphical Models: Theory, Algorithms and Applications Abstract

Image processing is an interdisciplinary field at the intersection of science and engineering. Mathematical modeling of image signals not only supports various engineering applications in our daily lives (e.g., digital cameras, high definition TV, ultrasound diagnosis and so on) but also offers a computational approach to understand sensory coding as a strategy for information processing in visual cortex. An improved understanding of image models is likely to lead to artifact free signal processing systems that well match the perception by human vision systems. Models developed for images could also facilitate the study of self organization principles underlying other complex sensory signals such as speech and video.

This research targets at a more fundamental understanding towards image modeling via nonlocal sparse representations (NSR). Unlike wavelet bases that are dilation and translation of a signal independent wavelet function, the PI advocates the representation of a signal by ?basis functions? that are dilation and translation of the signal itself. Self similarity based signal representation is closely related to the fractal theory and can be connected with sparse representations via regression shrinkage and selection. Such fruitful connection leads to a nonlocal regularization framework on graphical models and a class of novel deterministic annealing optimization techniques. This research has applications to a wide range of image processing systems from noise suppression in medical imaging to artifact removal in JPEG/JPEG2000 compression. The longer term objective of the research is to demonstrate the intimate relationship between image processing and higher level vision tasks such as segmentation and recognition. III: Small: Transforming Long Queries This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Many information needs can be more easily expressed using longer, sentence length queries, but the inadequacies of current search engines force people to try to think up the right combination of keywords to find relevant documents. This can be very difficult and often leads to search failures. On the other hand, long queries are handled poorly by current search engines. The focus of this project is on developing retrieval algorithms and query processing techniques that will significantly improve the effectiveness of long queries. A specific emphasis is on techniques for transforming long queries into semantically equivalent queries that produce better search results. In contrast to purely linguistic approaches to paraphrasing, query transformation is done in the context of, and guided by, retrieval models. Query transformation steps such as stemming, segmentation, and expansion have been studied for many years, and we are both extending and integrating this work in a common framework. The new query processing techniques for long queries are being developed and distributed using the NSF funded Lemur toolkit from UMass/CMU, and are being evaluated using a variety of document and query collections from sources such as the web, social media sites such as forums, and TREC, with an involvement of graduate and undergraduate students. The project Web site (http://ciir.cs.umass.edu/research/longqueries) will be used to further disseminate results.

Given that search is one of the two most common activities on the web and that new modalities for search, such as voice interfaces and collaborative question answering, are increasing the importance of long queries, this research could have a very broad impact, both in the home and the office. CSR: Small: Elastic Computing An Enabling Technology for Transparent, Portable, and Adaptive Multi Core Heterogeneous Computing This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Computing architectures are increasingly parallel, most relying on multi core microprocessors. Trends are also towards increased heterogeneity, with systems combining diverse components ranging from multiple microprocessors, field programmable gate arrays (FPGAs), graphics processing units (GPUs), among others, often resulting in speedups of 10x to 100x.

A number of research thrusts have focused on the challenges associated with utilizing parallelism and heterogenity in computing architectures, including languages aimed at simplifying parallel programming. Despite numerous compiler and synthesis studies, usage of such systems has largely been limited to device experts due to significantly increased application design complexity.

To reduce design complexity, this project will investigate elastic computing, which is a framework to automatically parallelize an application across numerous heterogeneous resources while dynamically optimizing for different runtime parameters such as input size, battery life, etc. Elastic computing overcomes limitations of compilers and synthesis tools by providing the optimization framework with extra knowledge of functions, thus enabling automatic exploration of thousands of possible implementations, even ones using different algorithms. This project establishes an underlying theory for elastic computing, validated for numerous architectures and application domains, thus laying the foundation for much future research. In addition, the project will enable scientists without programming expertise to more easily develop applications for powerful multi core heterogeneous systems, thus enabling new simulations that may advance the state of science. The project will also integrate elastic computing into graduate curriculum and will involve collaboration with the South East Alliance for Graduate Education and the Professoriate (SEAGEP) to recruit underrepresented students for undergraduate research. RI: Small: An Affect Adaptive Spoken Dialogue System that Responds Based on User Model and Multiple Affective States There has been increasing interest in affective dialogue systems, motivated by the belief that in human human dialogues, participants seem to be (at least to some degree) detecting and responding to the emotions, attitudes and metacognitive states of other participants. The goal of the proposed research is to improve the state of the art in affective spoken dialogue systems along three dimensions, by drawing on the results of prior research in the wider spoken dialogue and affective system communities. First, prior research has
shown that not all users interact with a system in the same way; the proposed research hypothesizes that employing different affect adaptations for users with different domain aptitude levels will yield further performance improvement in affective spoken dialogue systems. Second, prior research has shown that users display a range of affective states and attitudes while interacting with a system; the proposed research hypothesizes that adapting to multiple user states will yield further performance improvement in affective spoken dialogue systems. Third, while prior research has shown preliminary performance gains for affect adaptation in semi automated dialogue systems, similar gains have not yet been realized in fully automated systems. The proposed research will use state of the art empirical methods to build fully automated affect detectors. It is hypothesized that both fully and semi automated versions of a dialogue systemthat either adapts to affect differently depending on user class, or that adapts to multiple user affective states, can improve performance compared to non adaptive counterparts, with semi automation generating the most improvement. The three hypotheses will be investigated in the context of an existing spoken dialogue tutoring system that adapts to the user state of uncertainty. The task domain is conceptual physics typically covered in a first year physics course (e.g., Newtons Laws, gravity, etc.). To investigate the first hypothesis, a first enhanced system version will be developed; it will use the existing uncertainty adaptation for lower aptitude users with respect to domain knowledge, and a new uncertainty adaptation will be developed and implemented to be employed for higher aptitude users. To investigate the second hypothesis, a second enhanced systemversion will be developed; it will use the existing uncertainty adaptation for all turns displaying uncertainty, and a new disengagement adaptation will be developed and implemented to be employed for all student turns displaying a second state of disengagement. A controlled experiment with the two enhanced systems will then be conducted in a Wizard of Oz (WOZ) setup, with a human Wizard detecting affect and performing speech recognition and language understanding. To investigate the third hypothesis, a second controlled experiment will be conducted, which replaces the WOZ system versions with fully automated systems.

The major intellectual contribution of this research will be to demonstrate whether significant performance gains can be achieved in both partially and fully automated affective spoken dialogue tutoring systems 1) by adapting to user uncertainty based on user aptitude levels, and 2) by adapting to multiple user states hypothesized to be of primary importance within the tutoring domain, namely uncertainty and disengagement. The research project will thus advance the state of the art in both spoken dialogue and computer tutoring technologies, while at the same time demonstrating any differing effects of affect adaptive systems under ideal versus realistic conditions. More broadly, the research and resulting technology will lead to more natural and effective spoken dialogue based systems, both for tutoring as well as for more traditional information seeking domains. In addition, improving the performance of computer tutors will expand their usefulness and thus have substantial benefits for education and society. CIF: Small: An Analytical Framework for Comprehensive Study of Intermittently Connected Mobile Ad Hoc Networks Intermittently Connected Mobile Ad hoc Networks (ICMANET) are one of the new areas in the field of wireless communication. Networks under this class are potentially deployed in challenged environments using isolated mobile devices with limited resources. They are emerging as a promising technology in applications such as in wildlife management, military surveillance, underwater networks, and vehicular networks. Recent focus on these networks has naturally generated efforts to analytically understand them. In contrast to conventional Mobile Ad hoc Networks (MANETs), links on an end to end path in ICMANETs do not exist contemporaneously. This compounds the analysis of such networks. The research provides one of the first steps for performance modeling of ICMANETs. This is very crucial in the design of practical schemes targeted to offer good performance across different mobility scenarios.

Most current analytical research employs simple mobility models such as the Poisson contact model, which are unrealistic. Departing heavily from this trend, the investigator has developed performance analysis under the very general class of stationary mobility models. This paves the way to analytically understand the effect of mobility parameters on performance. In particular, the research answers questions of the following nature: What parameters of the mobility model affect ICMANETs performance? How does one extract the necessary information? The research objective is to develop novel approaches in performance modeling of ICMANETs, drawing out key ideas from Markov chain and queuing theory. The research attempts to arrive at a novel framework which is capable of capturing key network characteristics under practical constraints such as finite bandwidth, random contact durations, and finite node buffers. Key performance measures such as throughput will be explored for various communication scenarios, routing protocols, network coding and buffer management schemes. III: Small:Collaborative Research: Coordinated Visualization for Comparative Analysis of Cross Subject, Multi Measure, Multi Dimensional Brain Imaging Data The amount of medical imaging data has been growing at an unprecedented rate in recent years due to the rapid advancement in medical imaging devices and technologies. In many medical application areas, assessment of similarity and disparity from multimodality, multi dimensional data across subjects plays a central role. Current software in exploration and visualization of a collection of multimodality, multidimensional cross subject data impedes the effective utilization and better understanding of acquired, large scale data.

The overall objective of this research is to design and develop a unique coordinated visualization framework, based on advanced geometric computing as well as data and visual abstraction, for integrating, interpreting and comparative analysis of cross subject, multidimensional, multi measure brain imaging data. This developed visualization framework extends the state of the art in both information visualization and medical visualization. It employs novel geometric feature analysis for better supporting surface matching and shape comparison, and generalizes data warehousing technology to spatially varying information deep inside multi dimensional medical images. The research outcomes are disseminated through traditional publications as well as the Internet.

This research project provides a useful multimodality imaging analytics framework which contributes to diverse application domains, such as clinical diagnosis of neurological disorders, drug efficacy analysis through quantitative image analysis, and basic neuroscience. In addition, the sharing of data and software tools has both clinical and educational values for students, physicians, researchers, and the general public. The integration of the research and education components promotes further interactions between computer science and neuroscience. AF: Small: Algorithmic and Game Theoretic Issues in Bargaining and Markets The theories of bargaining and markets, key theories within game theory and mathematical economics, suffer from a serious shortcoming other than a few isolated results, they are essentially non algorithmic. With the advent of the Internet, totally new and highly successful markets have been defined and launched by Internet companies such as Google, Yahoo!, Amazon, MSN and Ebay, and bargaining has emerged as a mechanism of choice in some Internet transactions, such as those made by priceline.com and iOffer.com. Motivated by these developments, much work has been done in the last decade on developing algorithms for markets and bargaining.

This project will extend this work along several exciting directions: attacking open problems remaining in the efficient computation of market equilibria such as Fisher s model with piecewise linear, concave utilities; extending bargaining algorithms to more general utility functions, such as piecewise linear, concave utilities; developing an algorithm for the Adwords problem that achieves 1 o(1) expected competitive ratio in the stochastic setting, and at the same time, a 1 1/e ratio in the worst case; developing models of bargaining that incorporate constraints such as timing or incentive compatibility, thus making them more suitable for use on the Internet; developing further our understanding of game theoretic properties of bargaining and markets; and developing further the primal dual paradigm for the combinatorial solution of nonlinear convex programs, in particular, in the setting of approximation algorithms.

This project will increase our understanding of the interactions and complex interdependencies of information systems, markets and social systems. It will enable and support efficient massive distributed systems. This project will provide algorithms for and insights into the computational aspects of markets and transactions on the Internet, thereby helping make their operation more efficient. Hence, this project is expected to contribute advances in science and engineering, as well as to promote economic prosperity. CIF: Small: Digital Mitigation of Spurious Tones in Fractional N PLLs CIF: Small: Digital Mitigation of Spurious Tones in Fractional N PLLs

Fractional N phase locked loops (PLLs) are critical components in most modern wireless communication systems including cellular telephones and wireless local area networks. Unfortunately, the error introduced by conventional PLLs contains period disturbances referred to as spurious tones, which only can be suppressed sufficiently for typical wireless applications with techniques that increase power consumption and cost. Furthermore, these techniques become less effective as integrated circuit (IC) technology continues to scale to smaller dimensions. Therefore, the spurious tone problem negatively affects power consumption, cost, and manufacturability of wireless communication systems, and the problem gets worse as IC technology scales with Moore?s Law.
The ÄÓ modulator in a conventional PLL is the fundamental cause of spurious tones. The spurious tones are induced when the ÄÓ modulator?s quantization noise is subjected to nonlinearity from non ideal circuit behavior in the PLL. The goal of this research is to develop a ÄÓ modulator replacement, called a successive requantizer, that avoids this problem. The successive requantizer has a different principle of operation than a ÄÓ modulator and its quantization noise is much less susceptible to nonlinearity induced spurious tones. The research tasks are 1) to further develop the theory underlying successive requantizers to improve their performance, 2) to investigate how PLLs can be optimized at the circuit level to take advantage of the reduced sensitivity to nonlinear distortion offered by successive requantizers, and 3) to develop a proof of concept fractional N PLL IC enabled by the theoretical results of the project that is compliant with a demanding wireless standard such as IEEE 802.16 and exceeds the present state of the art in terms of minimizing power consumption and circuit area. SHF: Small: An Extensible Gradual Type System via Compile Time Meta Programming Many modern programming languages fit in a category called scripting languages. These languages are especially flexible, and they allow a programmer to quickly assemble pieces of a program to solve a problem. Unfortunately, a scripting language s flexibility can also hinder the programmer s ability to develop and maintain a script when it grows into a larger program. As scripting languages have become more popular, especially with new programmers, long term development and maintenance problems affect a growing body of programs at many layer of our computing infrastructure. This project is about smoothing the path from scripts to a more rigorous style of programming by introducing type systems into scripting languages. A type system can offer up front guarantees about how a program will execute, and it can help isolate the effects of program modifications. Rather than imposing a particular type system, however, this project s goal is to explore a particular way of defining and customizing a type system while introducing it gradually into an existing program. The specific technical approach in this project builds on Lisp style macros as provided by the PLT Scheme programming language. TC: Small: Privacy safe sharing of network data via secure queries (PSEQ) This project will explore a novel direction to address the privacy/ utility tradeoff of trace sharing: secure queries on original traces under their owner?s control. The data owner publishes a query language and an online portal, allowing researchers to submit sets of queries to be run on data. Only certain operations are allowed on certain data fields, and in specific contexts. This policy is specified by the provider and enforced by the language interpreter. The interpreter analyzes the queries, runs those that are permitted and returns the results to the researcher. The results consist of aggregate information such as counts, histograms, distributions and not of individual packets.

Secure queries address privacy/utility tradeoff much better than sanitization. Privacy is protected by finer grain control given to data owner, which permits detection of many passive attacks and minimization of information leakage from active attacks. Future attack vectors can be handled by adding new constraints on the query language. Secure queries also show a potential to reveal more data to researchers than it was possible with sanitization. Fine grain control via query language enables processing of many fields in the application header, and even in sensitive application content, while satisfying the owner?s privacy concerns. This is likely to increase utility of public traces for application and security research.

The work will investigate research utility of network trace data, and the relationship of known and novel attacks to combinations of packet fields, operations on those fields, and contexts that pose privacy risk. Based on these findings, the team will develop a secure query language Trol, and an interpreter for this language Patrol. Trol will support common operations on traces, needed for networking research, and Patrol will prohibit queries and contexts that pose a privacy risk as specified by the provider ?s privacy policy. Both the language and policies will be extensible by data owners to accommodate future discoveries. Trol and Patrol will be deployed at USC/ISI and will run on publicly available, sanitized trace archives and on synthetically generated, full packet traces. This deployment will help to test expressiveness and privacy protection of Trol operations. The work will also publicize the work among data owners, to motivate the shift from sanitization to protection of traces via secure queries. AF: Small: Influencing and Improving Networks Formed by Strategic Agents Influencing and Improving Networks Formed by Strategic Agents

An increasing number of networks on the forefront of scientific research, and of great importance to our world, are developed, built, and maintained by a large number of independent agents, all of whom act in their own interests. Such networks with strategic agents include the Internet, peer to peer file sharing schemes, business contracts between companies, and social networks representing relationships between groups of people. While these networks cannot be fully controlled, they can often be influenced in a limited way, sometimes resulting in dramatic improvement of the global network behavior. For example, this influence can include giving a few agents incentives to behave differently, altering a small part of the network, or even providing some extra information that makes a huge difference. This project will develop methods and algorithms with provable guarantees for influencing networks of strategic agents in order to improve the overall network behavior.

This project will study the formation of various networks by strategic agents, and will especially focus on the system of customer provider and peering contracts between Autonomous Systems (AS s) in the Internet. In the process of this research, new approximation algorithm concepts will be introduced, and will yield techniques for improving the global qualities of a network, and for preventing undesirable entities from gaining undue influence over it. While interactions of self interested agents have been studied in numerous fields, a systematic study of how to influence such agents with only a limited amount of power has never been done. Besides contributing to algorithmic game theory and the study of networks, this research will open up new research directions in economics, AI, and the social sciences. RI:Small:Collaborative Research: Infinite Bayesian Networks for Hierarchical Visual Categorization Humans possess the ability to learn increasingly sophisticated representations of the world in which they live. In the visual domain, it is estimated that we are able to identify in the order of 30,000 object categories at multiple levels of granularity (e.g. toe nail, toe, leg, human body, population). Moreover, humans continuously adapt their models of the world in response to data. Can we replicate this life long learning capacity in machines?

In this project, the PIs build hierarchical representations of data streams. The model complexity adapts to new structure in data by following a nonparametric Bayesian modeling paradigm. In particular, the depth and width of our hierarchical models grow over time. Deeper layers in this hierarchy represent more abstract concepts, such as ?a beach scene? or ?chair?, while lower levels correspond to parts, such as a ?patch of sand? or ?body part?. The formation of this hierarchy is guided by fast hierarchical bottom up segmentation of the images.

To process large amounts of information, the PIs distribute computation across many CPUs /GPUs. They develop novel fast inference techniques based on variational inference, memory bounded online inference, parallel sampling, and efficient data structures.

The technology under development has a large number of potential applications ranging from organizing digital libraries and the worldwide web, building visual object recognition systems, successfully employing autonomous robots and training a ?virtual doctor? by processing worldwide information from hospitals about diseases, diagnosis and treatments.

Results are disseminated through scientific publications and publicly available software. CSR:Small:Preventing the Exploitation of Software Vulnerabilities and Execution of Malicious Software on Embedded Systems This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Software security attack prevention, which addresses threats posed by software vulnerabilities and malicious software, is important for modern computing, especially for embedded systems. Despite widespread research efforts, the increasing complexity of software and sophistication and ingenuity of software attacks have led to a constant need for innovation. Some of the shortcomings of conventional techniques are insufficient detection accuracy (false positives/negatives) and high performance penalties.

In this project, a new methodology will be investigated for detecting and preventing malicious code execution and software vulnerability exploits, with the potential to significantly improve the accuracy and efficiency beyond current techniques. It will leverage recent advances in related areas, such as virtualization and dynamic binary instrumentation, which enable efficient creation of isolated execution environments and dynamic monitoring and analysis of program execution. The key aspects of the project are safe post execution analysis to detect violation of specific security policies, derivation of a hybrid model that represents a dynamic control of the program/data flow in terms of regular expressions and data invariants, run time prevention of malicious behavior, and several software/hardware enhancements for efficiently deploying the defense framework on embedded systems.

The methodologies will be disseminated through research articles, and software tools developed will be placed on the world wide web. Undergraduates will be encouraged to carry out independent research projects on this topic. Princeton encourages applications from female and minority students through special fellowships, which will be leveraged. Several other outreach activities are also planned for promoting education among underrepresented high school students. RI: Small: Collaborative Research: Infinite Bayesian Networks for Hierarchical Visual Categorization Humans possess the ability to learn increasingly sophisticated representations of the world in which they live. In the visual domain, it is estimated that we are able to identify in the order of 30,000 object categories at multiple levels of granularity (e.g. toe nail, toe, leg, human body, population). Moreover, humans continuously adapt their models of the world in response to data. Can we replicate this life long learning capacity in machines?

In this project, the PIs build hierarchical representations of data streams. The model complexity adapts to new structure in data by following a nonparametric Bayesian modeling paradigm. In particular, the depth and width of our hierarchical models grow over time. Deeper layers in this hierarchy represent more abstract concepts, such as ?a beach scene? or ?chair?, while lower levels correspond to parts, such as a ?patch of sand? or ?body part?. The formation of this hierarchy is guided by fast hierarchical bottom up segmentation of the images.

To process large amounts of information, the PIs distribute computation across many CPUs /GPUs. They develop novel fast inference techniques based on variational inference, memory bounded online inference, parallel sampling, and efficient data structures.

The technology under development has a large number of potential applications ranging from organizing digital libraries and the worldwide web, building visual object recognition systems, successfully employing autonomous robots and training a ?virtual doctor? by processing worldwide information from hospitals about diseases, diagnosis and treatments.

Results are disseminated through scientific publications and publicly available software. CIF:Small: Primitives for Cryptography and Communications CIF: Small: Primitives for Cryptography and Communications

Principal Investigator: Andrew Klapper, University of Kentucky

Pseudorandom sequences and highly nonlinear functions are essential for
digital communications and information technology. They are used in
stream cipher cryptosystems, spread spectrum systems in cellular
telephones, GPS systems, satellite communications, error correcting codes
for digital communication, and large simulations for such applications as
weather prediction, reactor design, oil well exploration, radiation cancer
therapy, traffic flow, and pricing of financial instruments. In each case
sequences or nonlinear functions with particular properties are needed.
Yet few general constructions of high quality pseudorandom sequences
and highly nonlinear functions are known. This research involves the
development and analysis of a large supply of these tools for a variety of
applications in cryptography, coding theory, and simulations.

In 1994 Klapper and Goresky proposed feedback with carry shift registers
(FCSRs), pseudorandom generators which are easily implemented and which
rapidly generate sequences with many desirable properties. These generalize
to algebraic feedback shift registers (AFSRs). Many basic properties of FCSRs
and AFSRs have been determined and they have been used in stream ciphers
and quasi Monte Carlo. This project addresses issues concerning FCSR and
AFSR sequences including (1) The development of new classes of highly
nonlinear functions for use in block ciphers and stream ciphers, (2) the
development of new tools for the analysis of nonlinear functions based on the
with carry paradigm, (3) the solution of the register synthesis problem
for AFSRs, (4) the identification of new classes of AFSR sequences with good
randomness properties, and (5) the extension of various ideas and methods in cryptography to vector valued functions and sequence generators. TC: Small: Exploiting Software Elasticity for Automatic Software Self Healing Software failures in server applications are a significant problem for preserving system availability. In the absence of perfect software, this research focuses on tolerating and recovering from errors by exploiting software elasticity: the ability of regular code to recover from certain failures when low level faults are masked by the operating system or appropriate instrumentation. Software elasticity is exploited by introducing rescue points, locations in application code for handling programmer anticipated failures, which are automatically repurposed and tested for safely enabling fault recovery from a larger class of unanticipated faults. Rescue points recover software from unknown faults while maintaining system integrity and availability by mimicking system behavior under known error conditions. They are identified using fuzzing, created using a checkpoint restart mechanism, and tested then injected into production code using binary patching. This approach masks failures to permit continued program execution while minimizing undesirable side effects, enabling application recovery and software self healing. III: Small Collaborative: Efficient Bayesian Model Computation for Large and High Dimensional Data Sets This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5.

This grant supports research in adapting and optimizing Markov Chain Monte Carlo methods to compute Bayesian models on large data sets resident on secondary storage, exploiting database systems techniques. The work will seek to optimize computations, preserve model accuracy and accelerate sampling techniques from large and high dimensional data sets, exploiting
different data set layouts and indexing data structures. The team will develop weighted sampling methods that can produce models of similar quality as traditional sampling methods, but which are much faster for large data sets that cannot fit on primary storage. One sub goal will study how to compress a large data set preserving its statistical properties for parametric Bayesian models, and then adapting existing methods to handle compressed data sets.

Intellectual Merit and Broader Impact

This endeavor requires developing novel computational methods that can work efficiently with large data sets and numerically intensive computations. The main technical difficulty is that it is not possible to obtain accurate samples from subsamples of a large data set. Therefore, the team will focus on accelerating sampling from the posterior distribution based on the entire data set. This problem is unusually difficult because stochastic methods require a high number of iterations (typically thousands) over the entire data set to converge. However, if the data set is compressed it becomes necessary to generalize traditional methods to use weighted points combined with higher order statistics, beyond the well known sufficient statistics for the Gaussian distribution. Developing optimizations combining primary and secondary storage is quite different from optimizing an algorithm that works only on primary storage. This research effort requires comprehensive statistical knowledge on both Bayesian models and stochastic methods, beyond traditional data mining methods. A strong database systems background in optimizing computations with large disk resident matrices is also necessary. This research will enable a faster solution of larger scale problems compared to modern statistical packages to solve stochastic models. Bayesian analysis and model management will be easier, faster and more flexible.

Broad Impact

This research will occur within the context of three separate application areas: cancer, water pollution, and medical data sets with patients having cancer and heart disease. The educational component of this grant will enhance current teaching and research on data mining. In an advanced data mining course students will apply stochastic methods to compute complex Bayesian models on hundreds of variables and millions of records. Data mining research projects will be enhanced with Bayesian models, promoting interaction between statistics and computer science.

Keywords: Bayesian model, stochastic method, database system SHF: Small: Collaborative Research: Flexible, Efficient, and Trustworthy Proof Checking for Satisfiability Modulo Theories This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).

Software bugs cost the U.S. economy over $60 billion each year. Promising bug detection technology depends on high performance logic solvers for Satisfiability Modulo Theories (SMT), which employ sophisticated algorithms to check large formulas efficiently. Sophistication has a price: the solvers themselves exhibit bugs, and are not trustworthy enough for safety critical applications. To increase confidence, some SMT solvers can emit formal proofs for valid formulas. Checking these proofs with a simple proof checker confirms the solver s results. SMT s rich logic poses challenges for standardizing a single proof format for all SMT solvers. Furthermore, proofs produced by SMT solvers can be gigabytes long, requiring an optimized proof checker.

This collaborative project is developing a verified proof checker supporting a flexible format called the Edinburgh Logical Framework with Side Conditions (LFSC). LFSC is a meta language for describing different proof systems, thus providing flexibility. Verification techniques are being applied to the proof checker itself to verify its optimizations, by writing it in a verified programming language called Guru. Support is also being added for LFSC proofs to the CVC3 solver. This research will greatly increase confidence in solver results through proofs, thus increasing the power of bug detection. SHF: Small:Design Tools and Optimization Methods for Digital Microfluidic Biochips Advances in digital microfluidics have led to the promise of biochips for applications such as point of care medical diagnostics. These devices enable the precise control of nanoliter droplets of biochemical samples and reagents. Therefore, integrated circuit (IC) technology can be used to transport and process biochemical payload in the form of nanoliter/picoliter droplets. As a result, non traditional biomedical applications and markets are opening up fundamentally new uses for ICs.

The goal of this project is to develop a design automation infrastructure for reconfigurable microfluidic biochips. It envisions an automated design flow that will transform biochip research and their use, in the same way as design automation revolutionized IC design in the 80s and 90s. Design tools and optimization methods are being developed to ensure that biochips are as versatile as the macro labs that they are intended to replace. The results from this research will enable a panel of concurrent immunoassay based diagnostic tests on an integrated microfluidic processor biochip that can be user programmed , and which can provide results in real time with picoliter sample/reagent volumes. Specific research tasks include control path synthesis and microcontroller/microfluidics integration, chip optimization for multiplexed immunoassays, microfluidic logic gates for smart decision making, and design for testability.

Miniaturized and low cost biochips will revolutionize data analysis for air quality studies and clinical diagnostics, enabling a transformation in environmental monitoring, healthcare, exposure assessment, and emergency response. This project is especially aligned with the vision of functional diversification and More than Moore , as articulated in the ITRS 2007, which highlights Medical as a System Driver for the future. The project bridges several research communities, e.g., microfluidics, electronic design automation, and biochemistry, and it provides interdisciplinary education to graduate and undergraduate students. CIF: NeTS:Small:Collaborative Research:Distributed Spectrum Leasing via Cross Layer Cooperation CIF: NeTS: Small: Collaborative research: Distributed Spectrum Leasing
via Cross Layer Cooperation

?Cognitive radio? networks, in which primary (licensed) and secondary (unlicensed) terminals coexist over the same bandwidth, are regarded as a promising solution to address spectral shortage and overcrowding. The main conventional approaches to enable such coexistence are: (i) Underlay/ overlay/ interweave strategies, which enforce strict constraints on the secondary behavior in order to avoid interference to the primary; and (ii) System wide dynamic spectrum allocation. Both frameworks have significant drawbacks for implementation of large scale distributed cognitive radio networks due to technological and theoretical limits on secondary spectrum sensing for (i) and on the stringent constraints on protocols and architectures for (ii).
To address the problems highlighted above, this research introduces and studies the novel framework of Distributed Spectrum Leasing via Cross Layer Cooperation (DiSC) as a basic mechanism to guide the design of Medium Access Control/ Data Link (MAC/DL) Physical (PHY) layer protocols in decentralized cognitive radio networks. According to this framework, dynamic ?leasing? of a transmission opportunity (e.g., a time slot) from a primary node to a secondary terminal is performed locally as driven by primary needs in terms of given Quality of Service (QoS) measures at the MAC/DL PHY layers. Specifically, DiSC enables each primary terminal to ?lease? a transmission opportunity to a local secondary terminal at MAC Protocol Data Unit (MPDU) granularity in exchange for cooperation (relaying). The project aims, on the one hand, at a theoretical understanding of the potentiality of the approach from the standpoints of network information theory and networking theory, and, on the other, at the (clean slate and back compatible) design of MAC/DL PHY protocols that effectively implements DiSC in a complex wireless environment.
Keywords: Spectrum leasing, dynamic resource allocation, cognitive radio, cooperative transmission, cross layer design, multi hop wireless networks. CIF: Small: Collaborative Research: Signal Design for Low Complexity Active Sensing Abstract

This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).

This research program is motivated by the recognition that the volume of sensor data is expected to overwhelm even the enormous performance improvements in silicon technology expressed by Moore s Law. The focus is the development of a low complexity alternative to non adaptive image formation through innovations in signal design. The objectives support closer monitoring of weather patterns and may lead to a more detailed understanding of climate change. They also impact a wider range of surveillance applications from microwave landing systems to through wall imaging. The research program is highly interdisciplinary with signal processing as the bridge between application domains and the mathematics of sequence design.
Current hardware allows transmission of wavefields that vary across space, polarization, time and frequency and which can be changed in rapid succession. However, sensing resolution is limited, not by hardware, but by the complexity of remote image formation. This research program develops new signal design principles that enable fast and reliable active sensing with minimal receiver signal processing complexity. The basic unit of transmission is a unitary matrix of phase coded waveforms indexed by array element and by pulse repetition interval, where the polarization of constituent waveforms may vary. Golay Complementary waveforms appear as entries of these matrices. Appropriate sequencing of unitary waveform matrices in time eliminates Doppler induced range sidelobes and provides resilience to multipath without compromising the simplicity of signal processing. OFDM signaling of complementary waveforms improves performance beyond conventional matched filtering by introducing nonlinear signal processing that exchanges static sidelobes for more dynamic cross terms. The development of a new mathematical framework based on group theory enables the systematic construction of new complementary sequences. TC: Small: Collaborative Research: Towards a Dynamic and Composable Model of Trust People rely on two types of trust when making everyday decisions:
vertical and horizontal trust. Vertical trust captures trust
relationships between individuals and institutions, while horizontal
trust represents the trust inferred from the observations and opinions
of other peers. Although significant benefit could be realized by
combining horizontal and vertical trust mechanisms, they have evolved
independently in computing systems.

This project focuses on developing a composable trust model capable of
tightly coupling vertical and horizontal trust in a manner that is
both amenable to formal analysis and efficiently deployable. This
research advances the state of the art in trust management through a
series of innovative results, including the design of a unified
framework for specifying composite trust policies and the design and
analysis of efficient algorithms for policy evaluation. The composite
trust management approach championed by this project also enables
policy authors to move beyond simple proof of compliance to identify
the top k preferred users satisfying security policies including
subjective assessments. The beneficiaries of this research range from
administrators of traditional computing systems who can better
incorporate previous history into their decision making processes, to
users in social networks who can more carefully manage the exposure of
their personal data. SHF:Small:Parallel ILP Based Global Routing on A Grid of Multi Cores Design of today s electronic systems would not be possible without the tools that automate the process of integrating billions of nano scale components e.g. into the brain of an iPhone. As technology advances towards mobile devices that are smaller yet more powerful, these tools need to evolve as fast as the systems that they help design in fact faster, because the nano scale components not only grow in numbers but also shrink in size, bringing along with them new challenges.

To improve existing design aid tools, a new window of opportunity has arisen due to the emergence of a more powerful yet affordable computational platform: a network of multi core computers working together as if it were one enormous machine. By leveraging this platform the proposed research investigates alternative design automation strategies which traditionally were deemed to be too time consuming.

The focus of the research is to improve an important step of the design process known as global routing, the step in which designers plan how the billions of nano scale components will be interconnected on the chip. This planning can significantly impact the severity of many issues in subsequent stages of the design cycle, yet it has to be done quickly. With the aid of large scale parallelism provided by grids, the research aims to demonstrate that the use of a computational technique called integer programming, which was previously viewed as too time consuming for global routing, can help generate significantly higher quality solutions while meeting practical runtime requirements.

Successful completion of the research contributes to faster delivery of electronic products to market with lower design cost, resulting in stronger businesses that can initiate new projects in the technology sector, and hence in the creation of more jobs. RI:Small:RUI:Intelligent Soundscape Analysis and Generation This project investigates hierarchical machine intelligence for modeling and composing complex soundscapes. We adapt methods for extracting time dependent information from text documents to the problem of extracting spectral (graphical) patterns and the probabilities that they occur or co occur in soundscapes. We are analyzing the spectral patterns that emerge from sound files of many types, including recordings of building interiors with regular foot traffic, musical files, and synthesized sound. A significant part of the research is in devising spectral features that are important for this kind of mapping/identification.

Research under this award is also investigating the use of reinforcement learning (RL) to identify time dependent landmarks from soundscape models, and we employ RL agents to compose large soundscapes from thousands of millisecond length grains of sound in a process called granular synthesis. Systems of RL agents enable us to study distributed time dependent RL agents in a complex environment, with the ability to produce aural demonstrations of the agents learning. We also expect the system to produce some compositions that are pleasing in the electronic music sense.

This research will have an impact on curricular efforts in Arts and Technology at Smith College, supporting the Computer Science Department s efforts to attract more students, especially to research. CSR: Small: Data Adaptable Reconfigurable Embedded Systems (DARES) This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Significant increases in application complexity demand processing requirements that exceed the performance achievable by current processors using software only implementations. For example, recent multimedia standards, such as JPEG2000, have significantly increased computational demands compared to previous standards. In an effort to alleviate the cost of developing software and/or hardware solutions capable of fully supporting such standards, many define several profiles ? specific settings for various configurable parameters ? that reduce the level of complexity needed to implement a specific profile. However, the number and variability of profiles even within a single domain still precludes traditional hardware implementations as a viable option for most applications.

The Data Adaptable Reconfigurable Embedded Systems (DARES) project focuses on developing hardware/software codesign and reconfigurable computing methodologies driven by data adaptability. This data adaptable approach allows designers to directly model data configurability of an application, thereby enabling a solution that can be dynamically reconfigured at runtime based on the profile of incoming data. The DARES project combines modeling techniques for capturing the data configuration space with new hardware/software codesign techniques to synthesize reconfigurable circuits and communication resources directly from the data/application model. The resulting hardware/software implementation provides the flexibility of software with the performance of hardware.

The results of the DARES project will provide a new design approach for dealing with the trend towards applications with increasing flexibility and configurability and will provide new methodologies for exploiting the reconfigurability of FPGAs beyond current approaches. TC: Small: Collaborative Research: Securing Multilingual Software Systems Most real software systems consist of modules developed in
multiple programming languages. Different languages differ in their
security assumptions and guarantees. Consequently, even if single
modules are secure in some language model and with respect to some
security policy, there is usually no uniform security guarantee on a
whole multilingual system. This project focuses on low overhead
techniques for providing security guarantees to software systems in
which type safe languages such as Java interoperate with native code.
Native code is developed in low level languages including C, C++, and
assembly languages. Although often used in software projects, native
code is notoriously insecure and is a rich source of security
vulnerabilities. The PIs are developing a two layered approach to
alleviating security threats posed by native code to type safe
languages: (1) Binary rewriting tools and their verifiers are being
incorporated into the Java Virtual Machine (JVM) for rewriting and
verifying native modules at the machine instruction level to enforce
security policies; (2) A safe dialect of C for interoperation with
Java is being designed; with the help of programmer annotations, the
safety of programs in this dialect can be statically verified. The
outcome of this project will enable popular platforms such as the JVM
and .NET and other major programming languages (e.g., Python, OCaml,
etc.) to incorporate native modules safely. The developed principles
will also be applicable to web browsers and operating systems in which
there is a need of extending them with untrusted low level modules
without comprising host security. RI: Small: Learning Strategic Behavior in Sequential Decision Tasks Many routine, real world tasks can be seen as sequential decision tasks. For instance, navigating a robot through a complex environment, driving a car in congested traffic, and routing packets in a computer network requires making a sequence of decisions that together minimize time and resources used. It would be desirable to automate these tasks, yet it is difficult because the optimal decisions are generally not known. Many existing learning methods lead to reactive behaviors that perform well in short term, but do not amount to intelligent high level behavior in the long term.

This project is developing methods for learning strategic high level behavior. Strategic methods need to (1) retain information from past states, (2) learn multimodal behavior, (3) choose between the different behaviors based on crucial detail, and (4) implement a sequential high level strategy based on those behaviors. The neuroevolution methods developed in prior work solve the first problem by evolving (through genetic algorithms) recurrent neural networks to represent the behavior. To solve the remaining problems, these methods are being extended in the proposed work with multi objective optimization, local nodes with cascaded structure, and with evolution of modules and their combinations. Preliminary results indicate that this approach is indeed feasible.

In the long term, developed technology will make it possible to build robust sequential decision systems for real world tasks. It leads to safer and more efficient vehicle, traffic, and robot control, improved process and manufacturing optimization, and more efficient computer and communication systems. It will also make the next generation of video games possible, with characters that exhibit realistic, strategic behaviors: Such technology should lead to more effective educational and training games in the future. The OpenNERO open source software platform developed in this work will be made available to the research community. III: Small: BeliefDB Adding Belief Annotations to Databases In many scientific disciplines today, a community of users is working
together to assemble, revise, and curate a shared data repository. As
the community accumulates knowledge and the database content evolves
over time, it may contain conflicting information and members can
disagree on the information it should store. Relational database
management systems (RDBMS) today can help these communities manage
their shared data, but provide limited support for managing
conflicting facts and conflicting opinions about the correctness of
the stored data.

This project develops a Belief Database System (BeliefDB) that allows
users to express belief annotations. These annotations can be positive
(agreement) or negative (disagreement), and can be of higher order
(belief annotations about other belief annotations). The approach
allows users to have a structured discussion about the database
content and annotations. A BeliefDB gives annotations a clearly
defined semantics that lets a relational database understand and
manage them efficiently.

Intellectual merits: (i) Definition of a Belief Database Model: The
project develops a formalism that extends a relational database with
belief annotations on data and on previously inserted annotations.

(ii) Design of a Belief Query Language: The project complements the
data model with a new query language that extends SQL. (iii)
Development of a canonical Belief Database Representation: The
projects develops approaches to store and manipulate belief databases
on top of a conventional RDBMS.

Broader impact: Curated databases and shared data repositories are
becoming widespread in the scientific communities. A BeliefDB
provides a new data management system that addresses the need of these
communities to manage conflicting data. If successful, the project
will be one of the pieces that will help data management technology
undergo a new paradigm shift, from managing data as content, to
supporting a community of users in collaboratively creating partly
conflicting database contents.

For further information on the project see the project web page::
http://db.cs.washington.edu/beliefDB/ SHF: AF: Small: Algebraic Methods for the Study of Logics on Trees Logics for describing properties of labeled trees, along with automata operating on such trees, have been studied extensively in connection with hardware and software verification. More recently they have been the subject of research on specification and query languages for XML documents, and this has focused attention on unranked trees those in which there is no fixed bound on the number of children a node may have. However, many fundamental questions about the expressive power of such logics remain unanswered; for instance, we do not yet possess an effective description of the properties of trees that are expressible in various versions of first order logic. This research project approaches these questions by means of a new algebraic theory of automata operating on unranked trees.

While much of the motivation for this work comes from logic, the difficult questions encountered are algebraic in nature: what is required is a deepened understanding of the ideal and decomposition structure of a new kind of algebraic invariant called forest algebras for tree automata. The project will draw on a very rich theory of the structure of finite semigroups developed in connection with the study of automata on words. One of the principal challenges entails generalizing what is already known about the structure of finite semigroups to this new and more complex setting.

This research will further the development of new mathematical tools to determine the expressive power of logics on trees. Successful completion of the project should ultimately lead to the application of these tools well beyond the problems that originally motivated the study, and has a potential practical impact in the development and analysis of languages for software and hardware verification, and of query languages for XML and related schemes for representing data. RI: Small: Foundations and Applications of Generalized Planning This project is developing automated methods of artificial intelligence (AI) for creating generalized plans that include loops and branches, can handle unknown quantities of objects, and work for large classes of problem instances. One of the key challenges is to reason about plans with loops and to do so without using automated theorem proving, which tends to be intractable. In particular, research is accomplishing the following goals: (1) develop new theoretical foundations for generalized planning; (2) develop effective abstraction mechanisms and new plan representations to support these new capabilities; (3) develop effective algorithms for plan synthesis as well as generalization of sample plans; (4) develop analysis tools to reason about the applicability, correctness and efficiency of generalized plans; (5) extend the framework to include sensing actions, conditional plans, and domain specific knowledge in the form of partially specified plans; (6) create a new set of challenging benchmark problems and perform a rigorous evaluation of the approach; and (7) increase the interaction between the AI community and other communities, particularly model checking, that study the abstraction mechanisms and theoretical foundations necessary for generalized planning. This new framework may significantly improve the scope and applicability of automated planning systems. AF: Small: Logic and Computational Complexity Logic attempts to describe the nature of sentences that have descriptions in some restricted format. Computational complexity investigates the number of steps needed on a computer to solve a given mathematical problem. Despite the seemingly disjoint nature of the topics, Logic and Computational Complexity have had a rich history of interactions in the past. Fagin s classical theorem giving a logical definition of NP is probably one of the first connections in the area. Other celebrated results in the nexus of finite logic and computational complexity include the Immerman Szelepsenyi theorem showing non deterministic logspace is closed under complementation; and Ajtai s work on certain formulae on finite structures, which shows that parity is not first order definable (which turns out to be equivalent to Furst, Saxe, and Sipser s result that parity does not have constant depth, polynomial size circuits). These results have advanced logic as well as computational complexity.

This research project is inspired by recent work by the student investigators on this project, Swastik Kopparty and Ben Rossman who have unearthed new connections between these areas leading to several new results. This project outlines novel further questions in the intersection of logic and computational complexity and outlines methods that may be employed to make progress on these questions. Some specific questions include:

Size lower bounds for uniform logarithmic depth circuits.
Explicit functions that are hard for first order logic augmented with arbitrary modular counting quantifiers (modulo composites in particular).
Choiceless algorithms for solving linear systems.
Decidability of the containment problem for conjunctive queries under multiset semantics. CSR: Small: CoreTime: Dynamic Computation Migration for Multicore System Software This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5)

This proposal develops a technique for allowing the multiple on chip caches available in multicore processors to be utilized more effectively by systems software. CoreTime is a framework to help data intensive software obtain good performance on multicore processors. Data intensive software, such as operating systems and web servers, may not be able to obtain full benefit from the increasing processing power of multicore processors, instead being bottlenecked by access to off chip Dynamic random access memory (DRAM). One example of an idea the CoreTime project is investigating is a global replacement policy among the per core caches of a multicore chip. Such a policy has the potential to significantly increase caching effectiveness, for example by preventing waste of cache space on redundant per core copies of popular data. A key principle in CoreTime s design is that software should control the policy for how cached data is spread over a multicore processor s on chip caches.

The long term importance of CoreTime is that it will increase the scalability of memory intensive software on future multicore processors by mitigating the DRAM botteneck. CoreTime software will be distributed as open source. The ideas and software will be used in MIT s undergraduate operating systems course, to give students experience with multicore processors and concurrency. Both undergraduates and graduate students will perform the research.

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5). III: Small:Collaborative Research: Bayesian Model Computation for Large and High Dimensional Data Sets This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5.

This grant supports research in adapting and optimizing Markov Chain Monte Carlo methods to compute Bayesian models on large data sets resident on secondary storage, exploiting database systems techniques. The work will seek to optimize computations, preserve model accuracy and accelerate sampling techniques from large and high dimensional data sets, exploiting
different data set layouts and indexing data structures. The team will develop weighted sampling methods that can produce models of similar quality as traditional sampling methods, but which are much faster for large data sets that cannot fit on primary storage. One sub goal will study how to compress a large data set preserving its statistical properties for parametric Bayesian models, and then adapting existing methods to handle compressed data sets.

Intellectual Merit and Broader Impact

This endeavor requires developing novel computational methods that can work efficiently with large data sets and numerically intensive computations. The main technical difficulty is that it is not possible to obtain accurate samples from subsamples of a large data set. Therefore, the team will focus on accelerating sampling from the posterior distribution based on the entire data set. This problem is unusually difficult because stochastic methods require a high number of iterations (typically thousands) over the entire data set to converge. However, if the data set is compressed it becomes necessary to generalize traditional methods to use weighted points combined with higher order statistics, beyond the well known sufficient statistics for the Gaussian distribution. Developing optimizations combining primary and secondary storage is quite different from optimizing an algorithm that works only on primary storage. This research effort requires comprehensive statistical knowledge on both Bayesian models and stochastic methods, beyond traditional data mining methods. A strong database systems background in optimizing computations with large disk resident matrices is also necessary. This research will enable a faster solution of larger scale problems compared to modern statistical packages to solve stochastic models. Bayesian analysis and model management will be easier, faster and more flexible.

Broad Impact

This research will occur within the context of three separate application areas: cancer, water pollution, and medical data sets with patients having cancer and heart disease. The educational component of this grant will enhance current teaching and research on data mining. In an advanced data mining course students will apply stochastic methods to compute complex Bayesian models on hundreds of variables and millions of records. Data mining research projects will be enhanced with Bayesian models, promoting interaction between statistics and computer science.

Keywords: Bayesian model, stochastic method, database system NetS: SMALL: Collaborative Research: Financial Dynamics of Spectrum Trading This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Presence of a well structured market is necessary for efficient and flexible use of licensed spectrum bands, and for fair pricing of spectrum usage. The goal of this project is to design radio spectrum markets that allow trading of spectral resources not only of the raw spectrum, but also of a variety of service contracts derived from the use of spectrum. Specific sub problems that will be addressed in this context include: 1) spectrum portfolio construction that optimizes risk versus return trade offs, 2) strategy design for optimal cooperation among providers, 3) price driven dynamic scheduling of subscribers, 4) optimal pricing of spectral contracts, and 5) regulatory mechanisms for effective functioning of the spectrum markets. The solutions to the above problems take advantage of similar formulations in financial engineering, and use tools from optimization, stochastic calculus, and control and game theory.

The project is expected to revolutionize spectrum trading by facilitating the design of secondary spectrum markets and spectrum regulation policies. Towards this end, This project aims at establishing a new cross disciplinary field that is at the interface of wireless networks and financial mathematics. The results will lead to more efficient use of the available spectrum by reducing spectrum wastage and fairness in pricing of wireless subscribers, and also help formulate governmental spectrum management and regulation policies. The results will be disseminated through publications in premier journals and conferences, and interactions with collaborators in the wireless industry. AF: Small: Spectral analysis, spectral algorithms, and beyond TITLE: Spectral analysis, spectral algorithms, and beyond

Spectral algorithms and spectral analysis form the basis for some of the most efficient and effective techniques in areas ranging from machine learning to scientific computing, from graphics to data mining, and from collaborative filtering to VLSI design. They also play a prominent role in combinatorial optimization, where they are used in algorithms for graph partitioning and constraint satisfaction.
Despite the apparent utility of spectral techniques and the intense effort devoted to their analysis, our ability to reason about them rigorously is still very limited. This project addresses the development of an algorithmically centered theory of spectral analysis which draws upon tools from contemporary mathematics, and is inspired by experimental evidence which has, so far, eluded a satisfactory theoretical explanation. The project also addresses the relative power of spectral methods from the viewpoint of computational complexity.
This involves the both the study of what cannot be done using only information about eigenvalues and eigenvectors, as well as what can be achieved by combining spectral analysis with classical combinatorial approaches, like computation of graph flows. SHF: Small: Data Learning Framework for Diagnosis Based Yield Optimization ABSTRACT
In the semiconductor industry, manufacturing yield, measured as the percentage of salable products produced, is a key metric that determines the financial success of a product line. Low yield translates into increased design cost, delayed time to market, and reduced productivity. When low yield occurs, tremendous engineering resources are spent to diagnose and resolve the problems.
This project proposes to develop a novel data learning framework that greatly improves the efficiency and effectiveness of the diagnosis and resolution process. The framework consists of a newly developed software infrastructure that interfaces with the existing Electronic Design Automation (EDA) and silicon test software infrastructures, through the design and silicon test data they produce. A collection of data learning software tools and methodologies that analyze said data are utilized to automatically extract knowledge for yield improvement.

The research is integrated with educational activities to develop course and tutorial materials released to the industry for broad impact, a state of the art laboratory for education, and a research program to attract undergraduate and underrepresented students. The research strives to achieve a comprehensive understanding of state of the art design and manufacturing practices including anticipated issues in the future, and to accomplish multidisciplinary studies merging knowledge from EDA, silicon test, data mining, and machine learning. Knowledge discovered through this research will provide the industry with a clear direction on where to invest resources to better cope with yield related issues in future ultra nanometer manufacturing technologies. The framework is designed to efficiently improve yield, which helps improve productivity in the semiconductor design industry. RI: Small: Exploiting Bilingual Resources to Improve Monolingual Syntactic Tools Natural language processing systems currently degrade when used outside of their training domains and languages. However, when text is analyzed along with translations into another language, the two languages provide powerful constraints on each other. For example, a syntactic construction which is ambiguous in one language may be unambiguous in another. We exploit such constraints by using multilingual models that capture the ways in which linguistic structures correspond between one language and another. These models are then used to accurately analyze both sides of parallel texts, which can in turn be used to train new, better, models for each language alone. Multilingual models are challenging because each language alone is complex, and the correspondences between languages can include deep syntactic and semantic restructurings. Focusing on syntactic parsing, we address these complexities with a hierarchy of increasingly complex models, each constraining the next. Our approach of multilingual analysis improves three technologies: resource projection, wherein tools for resource rich languages are transferred to resource poor ones, domain adaptation, wherein tools are transferred from one domain to another, and multilingual alignment, wherein correspondences between languages are extracted for use in machine translation pipelines. In addition to publishing the research results from this work, we also make freely available the multilingual modeling tools we develop. RI: Small: Performance Prediction and Validation for Object Recognition This project brings together an interdisciplinary collaboration between Engineering and Statistics for developing the science for automated object recognition. Object recognition technology is pervasive in a broad range of applications, such as safety and law enforcement, defense and security, autonomous navigation, industrial manufacturing, business process and e commerce. The proposed transformative research provides a foundational framework for predicting the performance of object recognition algorithms based on probabilities and structural characteristics of the models and data. In contrast to previous work, the proposed research considers not only the data distortion factors but also the model similarity. The performance is explicitly modeled as a function of data distortion factors (feature uncertainty, occlusion and clutter) and model factors (similarity) so that one can ultimately characterize the probability distribution of performance rather than the empirical results on a specific dataset. The research will focus in three areas: 1) Developing the Bayesian formulation of performance prediction and bounds on performance. 2) Analyzing the effects of all of the assumptions made in the performance prediction approach; evaluating the effects of mathematical tractability, complexity and the gain in performance and validating different assumptions based on real data. 3) Generating results of the approach for practical applications. The development of scientific theory, prediction and computational models for recognition will result in the development of systematic approaches to the design of recognition systems that can reliably achieve predictable results in complex real world, and contribute towards the science of computer based object recognition. CIF: Small:Compressive Projection Principal Component Analysis This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).

Principal component analysis (PCA) has long played a central role in dimensionality reduction and compression. However, the fact that PCA is a data dependent transform that is traditionally determined via a computationally expensive eigendecomposition often hinders its use in severely resource constrained settings. Hyperspectral imagery is a prime example although PCA offers excellent decorrelation and dimensionality reduction when applied spectrally to hyperspectral image volumes, the fact that many hyperspectral sensors are airborne or spaceborne devices limits wider use of PCA. In such applications, it would be greatly beneficial if PCA based dimensionality reduction and compression could be accomplished without the heavy encoder side cost entailed by traditional PCA. This research investigates a process that effectively shifts the computational burden of PCA from the resource constrained encoder to a more capable base station decoder.

The studied approach, compressive projection PCA (CPPCA), is driven by projections at the sensor onto lower dimensional subspaces chosen at random, while the CPPCA decoder, given only these random projections, recovers not only the coefficients associated with the PCA transform, but also an approximation to the PCA transform basis itself. By using encoder side random projections, CPPCA permits dimensionality reduction to be integrated directly with signal acquisition such that explicit computation of dimensionality reduction at the encoder is eliminated.
Computation and memory burdens are instead shifted to the CPPCA decoder which consists of a novel eigenvector reconstruction process based on a convex set optimization driven by Ritz vectors within the projected subspaces. Research activities are aimed at further understanding CPPCA both analytically and practically, including the exploration of CPPCA in data arising in geospatial applications, and the development of adaptations to the basic CPPCA process so as to improve performance on anomalous data. NeTS: Small: Collaborative Research:Secure and Resilient Channel Allocation in Multi Radio Wireless Networks Computing devices equipped with multiple radio interfaces and working on multiple channels are becoming predominant in wireless networks. These networks are usually Multi Interface Multi Channel Mobile Networks (MIMC MANETs). However, the study of security vulnerabilities and the research of fundamental security mechanisms in channel management of MIMC MANETs have been seriously lagging behind the rapid progress of other research.

This project studies the security of MIMC MANETs in three aspects.

1. Investigating the unique (unknown) security vulnerabilities associated with channel management in MIMC MANETs.

2. Developing MIMC enabled security mechanisms. This project redefines channel conflict, reveals the fundamental causes and consequences of channel attacks, and develops novel and attack resilient security mechanisms to secure channel management (and routing) in MIMC MANETs. New security mechanisms utilize the capability of MIMC, and include collaborative channel monitoring, channel utilization based channel conflict detection and resolution, logic based attack investigation, and cross layer security design.

3. Building MIMC security and performance evaluation toolkits: This project develops evaluation toolkits and builds experimental environments. The toolkits and experimental environments can serve as a major testbed for the whole research community to conduct future MIMC MANET research.

This project will advance the understanding of the unique security problems in MIMC MANETs. The developed techniques will greatly enhance the security of the MIMC network infrastructure and secure the mission critical applications built atop such networks. Broader impacts will result from the education, outreach, and dissemination initiatives. Educational resources from this project, including course modules and teaching laboratory designs, will be disseminated through a dedicated Website. CSR: Small: Materialized Views over Heterogeneous Structured Data Sources in a Distributed Event Stream Processing Environment Title: CSR: Small: Materialized Views over Heterogeneous Structured Data Sources in a Distributed Event Stream Processing Environment
Investigator: Suzanne W. Dietrich
Institution: Arizona State University
Proposal #: 0915325


Project Abstract: Software systems are becoming increasingly complex, requiring the coordination of heterogeneous structured data sources in a loosely coupled distributed environment with support for handling events and streaming data. Some sample systems include homeland security, criminal justice, supply chain management, health care, and consumer monitoring. Such software systems involve numerous query expressions for detecting events, monitoring conditions, handling streams, and querying data. This research analyzes the dependencies among these query expressions over structured data sources defined in different language components over relational or data centric XML to detect common subexpressions as candidates for materialized views. When views are materialized, the results of the computed view are stored so that subsequent references efficiently access the materialized view, avoiding the cost of recomputation. This performance improvement is even more critical with distributed data sources. However, the materialized view must be updated if any data source that it depends on has changed. To avoid costly recomputation, an incremental view maintenance algorithm uses the change to incrementally compute updates to the materialized view. A unique aspect of this research is the efficient maintenance of the materialized views while respecting the native format of the underlying loosely coupled, heterogeneous data sources. Using state of the art commercial and open source components, a prototype environment that supports a distributed event stream processing framework provides a research and evaluation platform for the exploration of the identification, specification, and incremental evaluation of materialized views over heterogeneous, distributed structured data. This environment also provides a shared infrastructure for undergraduate research and curriculum enhancement. NeTS:Small:Collaborative Research: An Integrated Environment Independent Approach to Topology Control in Wireless Ad Hoc Networks This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Each node in a wireless ad hoc network can choose the power at which it makes its transmissions and thus control the topology of the network. Though well studied in the research literature, the problem of topology control has been largely considered only in idealized wireless environments and in isolation as a graph theoretic abstraction. This project focuses on the design of topology control algorithms for reduced energy consumption, reduced interference and higher capacity in real wireless environments in the presence of multipath fading, link failures, high error rates and many other radio irregularities. The methodology follows two key philosophical goals: (i) an environment independent approach which makes no constraining assumptions about the wireless environment (as opposed to trying to achieve approximations of reality in the assumptions), and (ii) an integrated approach which does not merely abstract out the problem of topology control separated from routing and link scheduling but embraces these into the design at the outset. This research also explores the fundamental limits of environment independent topology control.

An immediate impact of this project is new algorithmic strategies that speeds up the actual deployment of energy efficient high performance wireless ad hoc networks with benefits to many known applications. A yet broader impact is new generalized distributed algorithms that can be employed in contexts beyond wireless ad hoc networks a variety of educational activities including course enhancements and participation in NSF RET programs for high school teachers. CSR: Small: Collaborative Research: Combining Static Analysis and Dynamic Run time Optimization for Parallel Discrete Event Simulation in Many Core Environments Proposal Title: CSR:Small:Collaborative Research: Combining Static Analysis and
Dynamic Run time Optimization for Parallel Discrete Event
Simulation in Many Core Environments
Institution: SUNY at Binghamton
Abstract Date: 07/06/09
This project investigates how a new processor paradigm (multi core architectures)
changes the way Parallel Discrete Event Simulation (PDES) is done. This topic is
important given the wide use of simulation and the emergence of multi core
architectures. PDES is likely to play an increasingly important role in discrete event
simulation as Moore?s Law is sharply curtailed and explicit parallelism becomes the
major avenue for improving performance of sequential applications. Improving PDES
performance translates to improved.
Discrete Event Simulation (DES) is widely used for performance evaluation in many
application domains. The fine grained nature of PDES causes its performance and
scalability to be limited by communication latency. The emergence of multi core
architectures and their expected evolution into manycore systems offers potential relief
to PDES and other fine grained parallel applications because the cost of communication
within a chip is dramatically lower than conventional networked communication. Absent
this dominant effect, PDES performance will be determined by issues such as load
balancing, synchronization and optimism control, and the choice and configuration of
various other algorithms and data structures of the simulator. Operation in a manycore
environment introduces new system tradeoffs that must be effectively balanced by the
system software. Primarily, the pressure on the memory system and resilience to load
fluctuations will emerge as critical issues that we address in the proposed research.
Finally, the more predictable nature of communication cost in this environment (due in
part to the more frequent synchronization possible between nearby cores) can be
exploited, especially by static analysis, for effective simulation.
As multi cores become the default microprocessor architecture, applications that are
performance constrained must evolve to use parallelism to take advantage of the
resources available on the cores. This project?s new PDES can have a significant
impact on a number of applications that rely on discrete event simulations. The PIs plan
to incorporate the research results into a graduate level course on parallel simulation
techniques and to involve undergraduate students in the project. TC: Small: Lossy Trapdoor Functions Applications, Realizations and Beyond A central goal in foundational cryptography is to find a primitive that realizes all interesting cryptographic applications, and yet has its security based on a simple assumption ideally a weak general assumption. Recently, a significant step in this direction was made with the introduction of Lossy Trapdoor Functions. A family of Lossy Trapdoor Functions (TDF) lets a user generate a publicly computable function, f, and corresponding trapdoor, t, such the user can recover x given f(x). Alternatively, the user can generate a function g such that g looses information about the input x; moreover, no computationally bounded adversary can distinguish whether it is given the description of an injective or lossy function.

Lossy Trapdoor Functions give rise to a host of cryptographic applications including: injective trapdoor functions, chosen ciphertext secure encryption, collision resistant hash functions, and oblivious transfer (OT). Furthermore, one can realize Lossy TDFs from several standard number theoretic assumptions: Decisional Diffie Hellman, the Shortest Vector (SVP) problem in lattices, and the Composite Residuosity problem. Taken all together this solved two longstanding open problems: realizing non factoring based trapdoor functions; and building chosen ciphertext secure encryption systems from lattice based assumptions.

This work will endeavor to make significant progress towards realizing the ultimate goal of building all of cryptography from simple general assumptions. The following directions will be pursued. First, the work will aim to create new constructions from both weaker number theoretic assumptions and from general assumptions. Second, the project will build trapdoors into Identity Based Cryptosystems. Constructing identity based trapdoors will enable applications such as ``deterministic encryption in the Identity Based context. Third, it will build new Non Interactive Proof Systems. The project will study the relationship to Universal Hash Proof Systems and create new Non Interactive Zero Knowledge Proof Systems.

This project will contribute to our foundational understanding of cryptography. Results will be disseminated through conferences, journals, and invited talks. In addition, funding will be used to support graduate students and build a cryptography program at UT Austin. TC: Small: Collaborative Research:Protecting Commodity Operating Systems From Vulnerable Device Drivers Device drivers constitute a large fraction of the code in commodity operating systems. Over 35,000 drivers with over 112,000 versions exist for Windows XP, while over 3.1 million lines out of 5.4 million lines of the Linux kernel is device driver code. As several recent exploits against device drivers show, drivers are rife with bugs that compromise system security.

Exploits against device drivers are dangerous because commodity operating systems execute drivers in kernel address space. A compromised driver can modify kernel data structures or execute arbitrary code with kernel privilege. Prior techniques that protect commodity OS kernels from device drivers either suffer from low performance or are limited to specific classes of vulnerabilities, such as memory errors.
Inspired by user mode driver frameworks, this project applies a three pronged approach to the problem of protecting kernel data from vulnerabilities in device drivers. First, this project will develop techniques to monitor kernel data structure updates initiated by device drivers and ensure that they do not compromise the integrity of these data structures. Second, it will develop techniques to limit driver access to kernel memory via DMA without requiring hardware support yet taking advantage of it if available. Third, it will develop new techniques for recovering from compromised drivers.

These techniques are applicable to legacy device drivers on standalone commodity operating systems and require minimal changes to the operating system. In addition, they impose negligible overheads on common case performance of device drivers and are thus practical for use even with high throughput devices. CSR:Small: Novel Techniques for Lossless Data Compression and Efficient Decompression in Heterogeneous Embedded Systems Abstract
0915376 Novel Techniques for Lossless Data Compression and Efficient Decompression in Heterogeneous Embedded Systems
PI: Prabhat Mishra

Demands for heterogeneous and complex embedded applications have soared drastically in recent years. Memory is one of the key driving factors in designing such systems since a larger memory indicates an increased chip area, more power dissipation, and higher cost. Compression techniques are promising to reduce memory requirements by reducing the program and data size. The existing approaches either perform efficient compression using complex and variable length encodings at the cost of slow decompression, or fast decompression using simple and fixed length encodings at the cost of compression efficiency. This research will develop novel tools and techniques to achieve both efficient compression and fast decompression for a wide variety of application domains and system architectures. This project will make three fundamental contributions: novel decode aware compression, optimal bitstream placement for parallel decompression, and synergistic integration of compression and encryption. A successful implementation of the proposed research will result in significant improvement in system performance and security, and drastic reduction in overall area, cost and energy requirements. Synergistic integration of compression and encryption will lead to efficient and secure embedded systems. The outcome of this research has a direct impact on everyday life. Improved compression/decompression techniques will have double impact ? low cost and low power everyday appliances for the public and improved performance and security for safety critical systems. This project will have broad educational impact by developing compression related projects in both graduate and undergraduate courses, and by involving undergraduates as well as minority students through UF Honors, University Minority Mentoring and SEAGEP programs. CIF:Small:Collaborative Research: Towards a Paradigm shift in Distributed Information Processing: Harnessing Group structure and Interaction This award is funded under the American Recovery and Reinvestment Act of 2009

(Public Law 111 5).


With a vision to fully realize the potential of next generation

communication network infrastructure based on ubiquitous sensor nodes,

this research introduces new architectures and strategies, for

distributed in network information processing, which directly harness:

the structure of the objective performance metrics for information

processing, the structure of the underlying statistical dependencies

in the information gathered by sensing nodes, the topology of the

network, and the capability for bidirectional interactive information

exchange. This research advocates a two fold paradigm shift in network

information processing: 1) a shift from the traditional goal of data

transport to the goal of function computation and 2) a shift from

unidirectional models of information flow to interactive information

flow models. This research supports the education of future scientists

and engineers by integrating research advances with curriculum

development and supports diversity by encouraging the participation of

women and under represented groups.



This research develops two fundamentally new classes of code ensembles

for interactive information processing. The first is based on

techniques from abstract algebra and random graph theory to capture

the structure of functions being computed at destinations. The second

is based on techniques from communication complexity and multiterminal

information theory to capture interactive structures of information

flow in the network. These new code classes have superseded the

performance of random code ensembles used in network information

theory since its inception. This research develops new analytical

frameworks and tools to uncover the fundamental performance limits of

interactive information processing in sensor networks. This research

facilitates the cross pollination of research fields by providing

components which build bridges between four fundamental areas, namely,

information and coding theory, abstract algebra, random graph theory,

and communication complexity theory. CSR:Small:Failure Aware Monitoring and Management of Online Availability and Performance for Dependable Computing Clusters CSR:Small:Failure Aware Monitoring and Management of Online Availability and Performance for Dependable Computing Clusters

Abstract:

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Computational clusters and clusters coalitions continue to grow in scale and in the complexity of their components and interactions. In these systems, component failures become norms instead of exceptions. Failure occurrence as well as its impact on system performance and operation costs is becoming an increasingly important concern to system designers and administrators. The success of petascale computing will depend on the ability to provide dependability at scale. Failure management and failure aware resource management are crucial techniques for understanding emergent, system wide phenomena and self managing resource burdens.

This project investigates a set of innovative techniques on failure aware monitoring and management for system level availability assurance. In this project, we will develop a framework along with mechanisms for failure aware autonomic resource management in large clusters, quantify the temporal and spatial correlations among failure occurrences for proactive failure management, and devise resource allocation and reconfiguration approaches to deal with the system availability and productivity issues caused by component failures that occur frequently in modern large and complex clusters.

Broader impacts of the project include the publication and dissemination of research results and developed software artifacts. The research enables collaborative research opportunities for students and faculty in the program, as well as undergraduate science and engineering students in New Mexico. Research based materials about dependable high performance computing will also be instilled into the undergraduate and graduate computer science and engineering curriculum. SHF:Small:Collaborative Research: Improving Code Clone Categorization During software maintenance, 50% to 90% of developer effort is spent on program comprehension activities, which are performed by developers to better understand source code. Reducing the effort spent by developers on these activities can reduce software maintenance costs. Researchers have developed techniques and tools to detect code clones (similar or identical segments of source code), because their presence can diminish program comprehensibility. However, knowledge only of the presence of clones does not allow a developer to perform maintenance tasks correctly and completely; proper performance of these tasks requires a thorough understanding of the relationships among the detected clones. Existing approaches for investigating these relationships are limited in their applicability and effectiveness.

The goal of this collaborative project is to develop an automated and rigorous analysis process for identifying and codifying the relationships among clones using their structural and semantic properties. To maximize the impact of the techniques and tools on the effectiveness and efficiency of performing maintenance tasks when clones are present, the investigators will perform a domain analysis. After initial development, the team will validate and refine the techniques and tools. The research will help developers to maintain software, reducing total software cost and improving overall software quality. CSR:Small:New Slicing Techniques for Program Parallelization This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Multicore processors have become main stream. At the same time, CPU clock rates are no longer increasing. Operating within this environment, application developers are under competitive pressure to parallelize software in order to achieve an aceptable performance level for computationally intensive features. Unfortunately, fully automated parallelization is constrained by sequential semantics and by limits of a compiler?s analytical abilities. Consequently, designing parallel programming languages for developing efficient yet reliable main stream applications continues to be a challenge. Rewriting software completely by hand is undesireable, because this kind of activity may waste years of investment in large scale software. Therefore, ideally, existing software should be automatically converted to new parallel forms.

To address this problem, new program slicing techniques are investigated in this project to enable programmers to convert existing sequential programs, with minimum hand made changes, into forms which can be safely parallelized by automatic tools. This research will result in a set of novel program analysis and transformation techniques to support the new slicing methods.

The project will have a broad impact on the US software industry?s ability to compete globally in its endeavor to retrofit existing software for the emerging hardware platforms. The tool and techniques developed in the project will be used in compiler and architecture courses offered at both graduate and undergraduate levels. The project also engages both graduate and undergraduate students in the key research activities, providing advanced technical training which will be critical to future success of a new generation of computer scientists and software engineers. SHF: Small: Collaborative Research: Design of Power and Area Efficient, Fault tolerant Network on Chip Circuits and Architectures This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The proliferation of multiple cores on the same die has given rise to communication centric systems, wherein the design of the interconnection network has become extremely important. To address the growing wire delay problems and improve performance in CMP architectures, a growing number of multi core designs have adopted a more flexible, scalable, packet switched architecture called Network on Chip (NoC). Of the several challenges facing current NoC designs, the three prominent ones are power dissipation, die area, and overall performance. In this research, we propose to develop energy efficient, area efficient, high performance, and fault tolerant NoCs by exploiting innovative technological (circuit) optimizations and architectural design space. On the technology side, we will develop and design novel circuit techniques that will achieve significant power savings, fault tolerance and considerable reduction in area requirements. On the architectural side, we will develop novel NoC designs that incorporate the proposed circuit design techniques and further improve network performance. This research is an organized effort that will combine circuit analysis, architecture study, performance evaluation and design synthesis. We will develop a comprehensive NoC design platform which will analyze the trade offs among various parameters of interest ? power, area and performance.

The success of this research is likely to have a significant impact on the design of NoC architectures for CMPs. The proposed research will tackle some of the major limitations of NoC design, namely power consumption and reliability, and will make significant advances in understanding the interplay between performance, energy, and reliability for NoC architectures. Realistic solutions to these problems will provide the ability to continue the improvements in computational performance that the information technology sector of our economy depends on. This multi disciplinary research will also play a major role in education by integrating discovery with teaching and training. AF: Small: Collaborative Research: Online Scheduling Algorithms for Networked Systems and Applications The Internet is now the world s dominant information infrastructure. Numerous requests from Internet users and their applications compete for shared resources in multiple ways. It is therefore critical to efficiently allocate limited network resources in order to provide high quality services. Improving the performance of the Internet in this manner has the potential to have extremely broad impact.

Resource management becomes even more challenging when mobile devices connecting to the Internet are considered. Designing efficient algorithms is difficult mainly due to the following factors: (1) diverse and unpredictable resource requests; (2) physical limitations on Internet links, on buffer space in network switches, on capacity of wireless channels, and on battery power in mobile devices.

This project aims to provide solutions for several fundamental algorithmic problems in networked systems and applications. Robust and insightful online algorithms will be developed for network switches forwarding prioritized packets and energy management in mobile devices. The objective is to understand the mathematical structure of these problems, to design elegant and easy to implement online algorithms, to provide rigorous analysis on their performance bounds, and to integrate these algorithms into the real systems to achieve better performance. CIF: Small: Efficient Signal Processing Algorithms for Inference of Gene Regulatory Networks This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).

Currently, one of the most important research problems encountered in molecular biology, bioinformatics, and systems biology consists in deciphering the mechanisms that lie at the basis of gene regulatory networks. The importance of gene regulatory networks is due to their fundamental role in the control and operation of the processes taking place in the living cell. Learning the structure and operation of gene regulatory networks facilitates the identification and understanding of the functions of macromolecules in cells, finding out the biological mechanisms of diseases and organ development, and developing efficient disease diagnosis and therapeutics means. The aim of this project is to build a computationally efficient signal processing framework for global understanding of the structure and functionality of gene regulatory networks.

Two major research thrusts are addressed in this project. The first research thrust develops information theoretic tools for efficient inference of causal regulations between gene expressions, and determination of global topologies for gene regulatory networks. The second research thrust develops a Bayesian information theoretic framework for inference of gene regulatory networks based on the integration of a multitude of heterogeneous data sources. A variational Bayes sampling formalism is also built to overcome the intractable computational complexity and convergence issues associated with the family of Monte Carlo techniques. This project brings important scientific, technological and educational contributions. By combining microarray data with prior biological knowledge and other data sources, the proposed computational tools have the potential of uncovering new aspects of the logic that governs the transcriptional control and interactions between genes, proteins and other macromolecules. III: Small: High Throughput Real Time Astronomy: Discrimiation, Dissemination and Decision This proposal will be awarded using funds made available by the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Intellectual Merit:
We propose to build an event driven network, a continuation of the NSF VOEventNet, applying the event driven technology to Astronomy. There are many large surveys coming online (CRTS, PTF, LOFAR, LIGO, LSST, SKA etc) that are scanning the sky repeatedly, to find what is changing and why. Observation can be caused by supernova, gamma ray burst, blazar eruption, planetary microlensing, or other exciting astrophysics. In many cases, rapid dissemination and follow up observation is the key to discovery, to get magnitudes and multi wavelength coverage.

The new VOEventNet will ingest events from multiple event surveys, professional and amateur and disseminate them for free at all levels, from child to major observatory. Trigger (selection) criteria can be built precisely and individually. When an event arrives, these triggers (boolean functions) are run to decide if an action is to happen or not. These triggers will work with heterogeneous, multi sourced data. Actions can include the moving and operating of a sensor, either human or robot guided, sending messages, fetching data to build a data portfolio, and running classification rules. Further, actions can cause the generation of new events, thus inducing a workflow. Events come in real time from an automated pipeline of a major sky survey or from a small college observatory or amateur astronomer doing follow up observations. These authors have previously registered the semantic meaning of the parameters that will be used in the actual events, so that automated systems can be effective, real time, semantically aware decision makers. The new VOEventNet will use an international standard (VOEvent) to be part of the global event infrastructure, exchanging events with other event brokers, such as NASA s GCN.

Broader Impact:
We propose here a cyber infrastructure for events that are created from real time sensors. The sensor could be many things, for example a radiation detector at a port. Excess click rate of the Geiger counter creates an event that is automatically assembled with other data, such as country of origin, and decisions made, perhaps using an additional sensor, the results triggering another action. Only important events merit human attention. This Sensor Event Action pattern occurs in many other fields of science and engineering, for example early earthquake warning, network provisioning when an internet story becomes viral, leak or pollution detection, and many others. AF: Small: Spectral Graph Theory, Point Clouds, and Linear Equation Solvers This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Two of the most important abstractions in Computer Science are graphs and point clouds. A graph abstracts relations between things: two vertices in a graph are connected by an edge if the objects associated with the vertices are related. Directed edges indicate a connection from one vertex to another. Both social networks and the web are modeled as graphs: vertices could represent people with edges between friends, or they may represent web pages with directed edges representing links. Point clouds are sets of vectors, each vector providing a list of numerical attributes. In many computer science applications, one associates a vector with each object being examined. For example, one may rate on a numerical scale different properties of a chemical, or how much a person likes movies from certain genres.

This project will unify these two abstractions by translating point clouds into graphs. Each vector becomes a vertex in a graph, with the strength of the edge connecting two vertices indicating the degree of similarity of the corresponding vectors. This translation will enable the application of numerous techniques that have been developed in graph theory to the study of point clouds.

Technical objectives of the project include the determination of the best graph to associate with a point cloud, the development of efficient algorithms for the construction of such a graph, and the development of new approaches to the analysis of graphs. In particular, a spectral analysis of directed graphs will be developed.

Both graduate students and undergraduates will be trained in research while working on this project. Educational materials developed during the course of the project will be disseminated through the internet as well as incorporated into a book under development. SHF: Small: Collaborative Research: Ultra Low Latency Optical Packet Switched Interconnects with Novel Switching Paradigm Abstract
With the advances of modern computer architectures, interconnects are playing an ever increasingly important role for providing an effective communication medium. Advanced optical switching technologies, such as optical packet switching and wavelength division multiplexing, provide a platform to exploit the huge capacity of optical fiber to meet the increasing needs. This research proposes a new switching paradigm optical cut through with electronic packet buffering, and systematically investigates the fundamental and challenging issues in the optical interconnect under this switching scheme, with the objective of designing cost effective, ultra low latency and pragmatic interconnects for future high performance computing and communications systems.

A unique feature of the proposed interconnect is that those packets that do not cause contention can pass the interconnect directly in optical form and experience minimum delay, while only those that cause contention are buffered. This research proposes to combine optical packet switching with electronic buffering, such that the interconnect will enjoy both fast switching and large buffering capacity. This research will (1) design the switching fabric and packet scheduling algorithms, (2) design practical Forward Error Control (FEC) for the interconnect, and (3) conduct extensive performance evaluations by means of simulation and emulation tools and analytical models. The outcome of this project will have a significant impact on fundamental design principles and infrastructures for the development of future high performance computing and communications systems. The PIs will integrate graduate and undergraduate students into the project and promote the participation of female and minority students. The findings will be disseminated to the research community by way of conferences, journals, and web site access. SHF: Small: Collaborative Research: Beyond Secure Processors Securing Systems Against Hardware Computer systems are increasingly located in places, where they can be physically accessed by people who are not authorized to have unrestricted control of and access to these systems. Examples of this include company laptops carried by employees, robotic vehicles, gaming consoles with copyright protection mechanisms, and remote users of servers in data centers. An emerging threat in such scenarios are hardware attacks, where snooping devices are attached to the system to directly read and/or modify data within the system. These attacks can circumvent all traditional security protections in the system, such as password checks, access permissions for data files, etc.
To address these threats, researchers and processor makers have proposed or developed various types of secure processor architectures, which encrypt and continuously verify data in the system?s memory. However, these secure processors are far from being ready for widespread use, primarily because they focus on a single processor, non mobile system that is already up and running and executing a single application.

This research project will investigate how to overcome these limitations, focusing on mechanisms for secure boot up and system configuration, secure communication between and migration of applications, secure access to peripheral devices (including the network). In essence, this project will build the intellectual framework that will be needed to make computers secure regardless of their physical location, preventing unauthorized access even if the system is captured, stolen, or actually owned by a potentially malicious entity. Other broader impacts of this project include improvements in education and workforce, by making computer hardware designers more aware of physical security and by making computer security experts more aware of implications of physical (hardware) attacks. CSR:Small:Efficient and Predictable Memory Hierarchies for High Performance Embedded Systems 0915503
CSR:Small:Efficient and Predictable Memory Hierarchies for High Performance Embedded Systems

Abstract


High end embedded systems are increasingly used for complex computation, such as advanced image processing and speech processing. Such complex tasks often involve irregular data structures such as linked data structures, or exhibit irregular access patterns even with regular data structures. Managing scratchpad memory for such tasks becomes increasingly challenging due to the difficulty of determining addresses that will be accessed in the future. For such complex applications, hardware managed storage (caches) can perform relatively well without much programming effort. However,caches do not offer predictability required to derive a tight worst case execution time (WCET) bound in real time systems, due to their dynamic behavior that is difficult to predict at compile time.

The goal of the project is to explore a new intelligent real time cache, which offers ease of storage management as well as allows programs to control caching behavior with low overheads, providing predictability that supports real time analysis. The work consists of: (1) Instruction and architecture support that provides primitives to control cache behavior with low overheads, (2)Development compile time analysis and run time support to support the new cache, and (3)Proof of concept of the proposed system.

Improved memory performance and predictability will be to make computer systems more efficient, enable more challenging problems to be solved, and improve energy efficiency of existing applications. This project will also extend outreach to undergraduate and high school students about the value of engineering to society and attract more students into the field. It will also benefit the public by disseminating research results through publications and tool releases. AF:Small:Optimization in surface embedded graphs This project aims to expand the boundaries of computational topology to include fundamental problems in combinatorial optimization, by developing efficient, practical, combinatorial algorithms to compute maximum flows, minimum cuts, and related structures in graphs embedded on topological surfaces. Preliminary results reveal intimate connections between the linear programming duality between flows and cuts, the combinatorial duality between graph embeddings, the equivalence between flows in the primal graph and shortest path distances in the dual graph, and Poincare Lefschetz duality between (relative) homology and cohomology. These connections allow maximum flows to be computed in near linear time in graphs of any fixed genus, by optimizing the relative homology class of the flow rather than directly optimizing the flow itself. However, the running time of these algorithms depends exponentially on the genus of the surface; a major goal of the project is to bring this dependence down to a small polynomial.

The project aims to advance knowledge and understanding across multiple research areas, by developing novel connections between fundamental techniques in combinatorial and algebraic topology, algorithm design, and combinatorial optimization. This research will lead to faster algorithms for several basic optimization problems, develop new applications of topological methods, and potentially settle several long standing open algorithmic questions. The project will support two PhD students at the beginning of their graduate careers. A broader goal of the research is to strengthen ties between the computer science and mathematics research communities; results will be disseminated broadly in venues visible to both communities. SHF: Small: T2T: A Framework for Amplifying Testing Resources Proposal Number: 0915526
Title: T2T: A Framework for Amplifying Testing Resources
PIs: Sebastian Elbaum and Matthew Dwyer

Abstract:

Virtually every software development company invests in the
construction and maintenance of testing resources to validate their
products. These investments, however, can be very costly.
Consequently, not all testing resources are supported equally as
companies focus their testing efforts on specific and limited types of
tests, ultimately sacrificing timely fault detection. The work proposed will address this problem by investigating strategies for amplifying the power and applicability of testing resources. The strategies will transform existing tests into new tests that add complementary testing capabilities to the validation process. The developed strategies will be unique in their treatment of tests as data. This will require the development of test representations that can be efficiently manipulated, and test transformations to realize operations that generate new and valuable tests. Test representations that are expressive enough to efficiently encode common forms of software tests will be developed, and transformations that operate both on and across different test representations will be explored. These test representations and transformations will constitute the T2T (test to test) framework and the initial step towards treating tests as data. If successful, this work will help software development companies lower product costs and enhance dependability by amplifying their existing testing resources. AF:Small:Geometric Embedding and Covering: Sequential and Distributed Approximation Algorithms Over the last decade, wireless networks of various kinds, including cellular networks, wireless LANs, sensor networks, community networks, etc. have become ubiquitous. This award focuses on algorithmic problems motivated by the design of protocols and applications for some of these networks. One feature of these networks, that the proposal attempts to take advantage of, is that network nodes typically reside in the plane or in 3 dimensional space and furthermore communication and sensing ranges of these nodes may also be modeled geometrically (for example, as disks or spheres in Euclidean space). As a result the award focuses on optimization problems in the geometric context and the goal is to design algorithms for these problems that can eventually be implemented efficiently on the network nodes. One class of problems considered are geometric embedding problems in which network nodes seek to discover their locations based only on information about which other nodes are within communication range. Solutions to these problems have the potential to improve routing protocols on these networks. Another class of problems PIs consider are geometric coverage problems whose aim is to optimally place sensor nodes with given sensing abilities so as to cover certain target regions. Such coverage problems arise in a variety of sensor network applications such as monitoring bridges, vineyards, and factory floors. CSR:Small: Online System Anomaly Prediction and Diagnosis for Large Scale Hosting Infrastructures Large scale hosting infrastructures have become important platforms for many real world systems such as cloud computing, enterprise data centers, massive data analytics, and web hosting services. Unfortunately, today s large scale hosting infrastructures are still vulnerable to various system anomalies such as performance bottlenecks, resource hotspots, service level objective (SLO) violations, and various software/hardware failures.

The goal of this project is to assess the viability of an online predictive anomaly management solution for large scale hosting infrastructures. We will develop novel techniques for 1) performing light weight online system anomaly prediction; 2) providing self evolving anomaly prediction models to achieve high quality prediction for real world dynamic systems; and 3) performing speculative, ``hot system anomaly diagnosis that search possible anomaly causes and suggest corrective actions while the system approaches the anomaly state. Our research will carry out evaluation by conducting experiments and case studies with our industrial partners on realistic platforms.

Students supported by this project will gain experience with development and testing of robust real world hosting infrastructures through interactions with our industrial partners, through internships and onsite experimentation. This work will advance diversity by involving students from under represented groups. Particularly, the prototype developed in this project will be applied to the Virtual Computing Lab (VCL) at NCSU, a platform for providing a better educational experience for K 12, community colleges, and universities.

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5). NeTS: Small: Algorithms and System Support for Monitoring of Amorphous Phenomena with Dynamic Signatures in Wireless Sensor Networks Emerging applications using wireless sensor networks for critical areas such as environmental monitoring and emergency response highlight the urgent need for more powerful algorithms for tracking amorphous events or phenomena with dynamic identities. Several such events may combine into a large whole or one event may disintegrate into several smaller ones. Current efforts in event detection and tracking have mostly assumed that either events remain distinct, never crossing or passing too close together to become indistinguishable, or if they do cross that they were identified prior and nothing new has formed. This project addresses the research challenges in designing and implementing a system that is capable of tracking events with or without well defined shapes and identities in the presence of stringent energy constraints and unpredictable network failures posed by wireless sensor networks. Specific research objectives include: design and evaluation of algorithms that detect and track any types of events including amorphous phenomena with dynamic signatures and events that possess a static shape with a crisp boundary; design and evaluation of algorithms that form and reform communication structures around events of interest; and development of an integrated system that provides interfaces to high level application tasks to execute on each identified event. Successful completion of this project will result in a rich set of tools that can be used by applications monitoring all different types of events. The tools will be made publicly available via the Internet. This project provides opportunities for recruitment of female students and undergraduate students. NeTS: Small: Cross layer Design for Seamless Mobility Support in Hybrid Wireless Mesh Networks The wireless mesh network technology has recently emerged as a promising solution to building large scale wireless Internet with quick and easy deployment. It has numerous applications, such as broadband Internet access, building automation, and intelligent transportation systems. The indispensable technology enabling large area roaming in wireless mesh networks is mobility management. Mobility management has been extensively studied in infrastructure based single hop wireless access networks. However, Internet based hybrid wireless mesh networks involve multihop wireless access from users to the Internet. In such an environment, routing and wireless channel access over multihop wireless links can produce detrimental effects on the performance of mobility management in wireless mesh networks. This research develops scalable and cost effective mobility management mechanisms for Internet based hybrid wireless mesh networks. It emphasizes the integrated design of mobility management with efficient medium access control and wireless multihop routing, which was not considered in traditional mobility management design. This research will provide innovative techniques to numerous applications of the wireless mesh network technology. It also provides an excellent opportunity for graduate and undergraduate research students. SHF: Small: Co Processors for High Performance Genome Analysis This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

This research develops novel techniques for applying the heterogeneous execution model, where a general purpose processor is accelerated by a special purpose co processor, to optimization based scientific computations. The result of this research is a library of computational building blocks that perform fundamental operations used in genome analysis, as well as a new design tool that uses this library to systematically synthesize complete co processor architectures that are optimized for the characteristics of the input dataset of interest.

Traditional development methodologies for heterogeneous computing have focused on computations that are based on data parallelized O(n) algorithms. This project demonstrates the use of heterogeneous computing for non O(n) algorithms, which have complex behavior, internal state, temporal locality, and a high ratio of computation versus communication. Adapting this class of computation to heterogeneous platforms provides high performance computing without the need for maintenance intensive and power inefficient traditional shared memory and cluster based supercomputers.

This project targets optimization based phylogeny reconstruction as a application case study. This application uses combinatorial optimization for its search for optimal phylogenetic (evolutionary) trees, as well as for its procedure for scoring candidate trees. SHF: Small: Collaborative Research: Specification Language Foundations for Modular Reasoning Methodologies This project extends the semantical foundations of object oriented (OO)
languages to cover methodologies for modular reasoning. Modular reasoning
means verifying software components assuming the specification of each
used component. Modularity is important for productivity and scalability,
but is difficult to achieve for OO software. To support modular reasoning,
researchers have proposed several methodologies that restrict programs and
their specifications. The goal of this project is to provide a theoretical
basis that supports practical techniques for justifying and using
methodologies.

This project provides guidance for the designers of programming and
specification languages, verification logics, and associated tools. The
results will improve the utility and extensibility of verification tools
a key goal of the Verified Software grand challenge. Software
developers may benefit from the integration and harmonious interoperation
of best practice methodologies. This project is potentially
transformative: it aims to enable combinations and customizations of
methodologies by tool users, scalable to real applications.

Improved OO programming methodologies may greatly improve programming
practice, especially in applications needing high assurance, reliability,
and security. This will benefit society, which increasingly depends on
computing systems built using OO components. Unification of methodologies
and streamlining of tools also facilitates the education of software
developers. SHF: Small: Exploiting Redundancy for Process Variation Resilience in Nano scale Fabrics This project addresses critical questions relevant to the successful realization of nanoelectronics based computing platforms. The promised higher density, lower power, and faster operation of nanoelectronics devices is attracting increasing interest. Two of the remaining major roadblocks to the creation of nanoscale computational structures, however, are very high levels of defects and process variations. While the small size of nanostructures and their self assembly based manufacturing provide great advantages, they also cause them to be very vulnerable to defects and parameter variations much more so than conventional CMOS. The high defect rates require a layered approach for fault tolerance and typically involve incorporating carefully targeted redundancy at multiple system levels. In addition to defects, even tiny variations in the manufacturing process can lead to very substantial variations in the actual values of key parameters, such as circuit delay. The purpose of this award is to develop a comprehensive methodology to efficiently use redundancy in nanofabrics to ameliorate the effects of both high defect levels and circuit delay variations. The methodology that will be developed will assist designers with a set of well tested approaches to provide resilience in the face of manufacturing defects and process variations. In educational terms, this project will contribute to the training of undergraduate and graduate students in the art of physical nanofabrics, nanoscale computer architecture, and circuit design for the future nano technologies. TC: Small: Collaborative Research: Mathematics of Infection Diffusion in Wireless Networks Abstract

The spread of malware has the potential to slow down or cripple wireless services. It poses a particularly inimical threat to a multitude of activities ranging across an entire spectrum from social interaction and gaming, to the flow of commerce and informational services, and, at the largest scale, to national security. Current countermeasures are mostly ad hoc and reactive in that they are used to fend off threats as they arrive or are preemptively discovered.

This project aims to develop theoretical foundations for malware control and counter measure design in wireless networks by drawing from epidemiological analogues in containment or quarantining strategies for limiting the spread of infectious diseases in human society, and game theoretic models for interactions among opponents. Optimal power control quarantining strategies that curtail and regulate the spread of contagion by exploiting the broadcast property of the wireless medium will be designed, validated analytically and experimentally, and incorporated in networking protocols.

This work will facilitate the development of new wireless paradigms where a plethora of devices need to securely communicate with each other and with other entities on the Internet. The research will not only draw from, but also contribute to, disparate disciplines such as epidemiology, game theory, optimal control and communication networks, and may eventually lead to new disciplines at the interfaces of these areas. Graduate and undergraduate students will be trained at the participating universities through supervision of doctoral and masters level dissertations and senior design projects. AF: Small: Fundamental Geometry Processing In this project new discrete geometry processing algorithms based on simple and intuitive discretizations of low order differential forms will be developed, along with the supporting theoretical foundations, and it will be shown that the proposed approach unifies and extends a number of existing mesh relaxation algorithms used for denoising, subdivision, and interactive shape deformation. In the classical theory of surfaces, a surface patch is defined by a smooth 3D valued parameterization function of two parameters, which in the language of differential forms is referred to as a 3D valued differential 0 form. The two partial derivatives of one of these 0 forms are three dimensional vector fields which define a 3D valued differential 1 form. A simple approach to surface deformations is to modify this 1 form by locally stretching and rotating its two component vector fields, and then solve for a parameterization function whose partial derivatives match the component vector fields of the modified 1 form. The discrete analog of this approach for deformations of graph embeddings and polygon meshes will be developed. The first fundamental form measures distances and angles on a smooth surface, and the second fundamental form measures how the surface normal varies, i.e., curvature. The two fundamental forms are invariant to rigid body transformations of the surface, and satisfy the Gauss Codazzi Mainardi (CDM) equations. Conversely, given two second order symmetric tensor fields satisfying together the CDM equations, the Fundamental Theorem of Surface Theory asserts that: 1) there exists a surface immersed in three dimensional Euclidean space with these fields as its first and second fundamental forms; and 2) the surface is unique modulo rigid body transformations. The analog theorem for polygon meshes will be formulated and proven, including extensions to manifold meshes of arbitrary topology, meshes with border, and even non manifold meshes. New contributions to the mesh compression literature will be made by exploiting the relationship between reconstruction algorithms and connectivity preserving mesh compression schemes. TC: Small: A High Performance Abstract Machine for Network Intrusion Detection Network intrusion detection systems (NIDS) need to balance between a set of challenges difficult to simultaneously address to their full extent: the complexity of network communication; the need to operate extremely efficiently to achieve line rate performance; and dealing securely with untrusted input. Our project aims to build an efficient and secure bridge between dealing effectively with these challenges, and offering the high level abstractions required for describing a security policy. Observing that NIDS implementations share a large degree of functionality, we introduce a new middle layer into NIDS processing, consisting of two main pieces: first, an abstract machine model that is specifically tailored to the network intrusion detection domain and directly supports the field s common abstractions and idioms in its instruction set; and second, a compilation strategy for turning programs written for the abstract machine into highly optimized, natively executable code for a given target platform, with performance comparable to manually written C code. As a broader goal, our undertaking provides the security community with a novel architecture that facilitates development and reuse of building blocks commonly required for network traffic analysis. While the focus of our effort is the design and implementation of the abstract machine environment itself, we envision enabling the community to unleash its full potential by building analysis functionality on top of the platform we develop. TC: SMALL: Contracts for Precise Types This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

In programming languages research, there is a strong trend toward extremely
precise type systems, which can encode and verify extremely detailed
assertions about the behavior of programs and the structure of the data they
manipulate. However, precision is a two edged sword. Combining precise
types with the other language features can lead to complex definitions that
are difficult to explain, implement, and reason about.

The goal of this project is to tame this complexity using contracts
executable partial program specifications that are checked at run time. The
project s primary contributions will be (1) to show how an improved theory
of contracts can be used to design and implement powerful, yet tractable,
precise type systems, (2) to develop a particular precise type system
extending regular expression types with security annotations, and (3) to use
the resulting system to demonstrate a useful form of ``updatable security
views in a multi level Wiki allowing groups of users with different
clearance levels to collaboratively author structured documents.

This application is motivated by discussions with NSA researchers about the
need in the intelligence community for such tools. The project s software
deliverables will be distributed freely under an open source license and
integrated with the popular Unison file synchronizer. TC:Small: Systems Sensitive Cryptography Cryptography often fails to impact practice because it is insensitive to the requirements and reality of the systems that implement and underlie it. This research aims to change this. Issues considered include legacy (changing the complex computer systems that make up today s world is expensive and error prone, so assessing the security of existing methods can be more important than providing new ones), malware (system penetration is widespread, exposing the keys that cryptography relies on for security) and the realities of randomness (it lacks in practice the quality expected in theory, dooming many cryptographic schemes). Work to be done includes assessment of the security of SSL encryption in the face of widespread system compromise, technically known as selective opening; analysis of the PKCS#1 standard for password based key derivation; design of hedged encryption schemes that retain as much security as possible when the randomness they are fed is of poor quality; and the study and design of cryptography secure against related key attacks. The impact of this work is higher assurance for existing, deployed cryptography currently used by millions, and new cryptography that is well placed to transition into real systems. NeTS: Small: A Practical and Efficient Trading Platform for Dynamic Spectrum Distribution This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Historical static spectrum assignment has led to a critical spectrum shortage. While new prominent wireless technologies starve for spectrum, large chunks of spectrum remain idle most of the time under their current owners. With proper economic incentives, spectrum redistribution based on an open market can eliminate the artificial shortage. This project develops S TRADE, an auction driven spectrum trading platform to implement the spectrum marketplace. S TRADE differs significantly from conventional FCC style spectrum auctions that target only a few large corporate players and take months or years to conclude. Instead, S TRADE serves many small players and enables on the fly spectrum transactions. In essence, S TRADE selectively buys idle spectrum pieces from providers and sells them to a large number of buyers matching their individual demands. By effectively multiplexing spectrum supply and demand in time and space, the proposed marketplace also significantly improve spectrum utilization. The design of S TRADE focuses on achieving spectrum multiplexing/reuse to improve spectrum utilization while guaranteeing economic robustness to encourage player participation and minimize market manipulation. This project focuses on tightly integrating novel algorithms of dynamic spectrum allocation with economic mechanism design. The research outcomes deepen our understanding of the way spectrum should be distributed and the role of economics in distributing it. By integrating economics mechanism design with wireless networking, this project forms an integral part of interdisciplinary training programs at both undergraduate and graduate levels. TC: SMALL: Language Based Accountability Distributed applications that require enforcement of fundamental authorization policies play an increasingly important role in internet and telecommunications infrastructure. Traditionally, controls are imposed before shared resources are accessed to ensure that authorization policies are respected. Recently, there has been great interest in the exploration of accountability mechanisms that rely on after the fact verification. In this approach, audit logs record vital systems information and an auditor uses these logs to identify dishonest principals and assign blame when there has been a violation of security policy. Accountability is an important tool to achieve practical security that should be viewed as a first class design goal of services in federated distributed systems.

The goals of this project are to provide a theoretical basis for the design and analysis of accountability mechanisms and to use the theory to develop language based techniques for statically validating auditors and accountability appliances. This proposal investigates operational (via game based models) and logical (via game logics) foundations for accountability to provide the theoretical basis for the design and analysis of accountability mechanisms.

The project will bring our understanding of accountability closer to the level of before the fact access control mechanisms, which benefit from well understood operational models and logics and therefore support language based methods that statically validate implementations against interfaces which specify security guarantees.

Accountability supplements purely technology based approaches to security with insights derived from the interplay between people and technology. This project aims to develop new models, logics, algorithms, and theories for analyzing and reasoning about accountability based approaches to trustworthiness. AF:Small:Coarse Grained Algorithms for Soft Matter Proposal Number: 0915718
Title: AF:Small:Coarse Grained Algorithms for Soft Matter
Principal Investigator: N. R. Aluru
Institution: University of Illinois at Urbana Champaign

Abstract

Soft matter (e.g. liquids, polymers, biopolymers, etc.) plays an important role in many emerging technologies in engineering and science. The physics of soft matter at macroscopic scales has been investigated for many decades. There is now a good understanding of how to manipulate soft matter at macroscopic scales. The physics of soft matter in confined environments (referring to the behavior of soft matter in constrained spaces) can be quite different from its macroscopic counterpart and many fundamental issues still remain. Soft matter in confined environments can find applications in important technological areas such as energy, health, sensing, sequencing, separation, etc. As a result, soft matter in confined environments has now gained significant interest from the scientific community. Various computational techniques can be used to understand physical, chemical and biological properties of soft matter. However, many of the existing techniques are either too expensive or not accurate enough to perform detailed studies. The objective of this research is to develop advanced computational algorithms to enable a detailed understanding of soft matter in confined environments.

Even though quantum mechanical and atomistic molecular dynamics simulations can be used to understand soft matter in confined spaces, they are limited to small length and short time scales. Mesoscopic methods, such as Brownian dynamics, Monte Carlo, lattice Boltzmann, dissipative particle dynamics, etc., can be used to overcome the limitations of quantum and atomistic molecular dynamics simulations, but, structural accuracy is a key issue in these methods. The objective of this research is to develop novel coarse grained algorithms where inter atomic potentials, widely used in atomistic simulation of soft matter, are directly incorporated into advanced physical theories. The inter atomic potentials will be coarse grained to ensure structural consistency. The inter atomic potential based coarse grained algorithms will be applied for several challenging examples of soft matter. The accuracy of the structural prediction from coarse grained algorithms will be compared with that from atomistic simulations. It is anticipated that inter atomic potential based coarse grained algorithms will be many orders of magnitude faster than purely atomistic simulations and the development of such algorithms will not only elucidate the fundamental aspects of soft matter in confined spaces, but will also lead to rapid computational prototyping of various applications of soft matter.

The proposed research is at the cross roads of several engineering and science disciplines. As a result, the development of inter atomic potential based coarse grained algorithms for soft matter will impact several disciplines and application areas. Some of the application areas that could benefit from this fundamental research are energy, sensing, health, sequencing, separation, etc. The main efforts of this project will result in the education of students and postdoctoral associates in the highly interdisciplinary area of soft matter. The research results from this project will be broadly disseminated via journal and conference publications, presentations at meetings and workshops, software, courses taught by the PI in the Department of Mechanical Science and Engineering and summer schools offered at University of Illinois. CIF: Small: Scalable Multimedia with Unequal Error Protection Scalable Multimedia with Unequal Error Protection
Pamela C. Cosman Laurence B. Milstein
Abstract
In today?s communications environment, it is important to have rich media content (images, audio, video) that can scale up or down depending on availability of resources. In quality scalable multimedia, some portions of a bit stream contain information that allows a moderate quality reconstruction of the image or video, and additional portions of the bit stream allow the source to be reconstructed at progressively higher quality. We consider the transmission of scalable multimedia data (image and video) through variable types of channels, with a focus on providing different levels of unequal error protection (UEP) appropriate for different levels of information importance and suitable for the channel conditions.
There are many techniques for providing protection against errors, including forward error protection (FEC), hierarchical modulation, and leaky and partial prediction in video coding. Our research involves two new techniques for combining hierarchical modulation with either image or video to produce enhanced performance. We consider a MIMO based technique, in which MIMO space time coding is used to increase reliability for the most important information in the scalable image or video data, whereas MIMO spatial multiplexing is used to increase data rate for the less important information. This is combined and optimized with existing techniques where unequal error protection is achieved by transmitting different power levels on multiple antennas. The hierarchical approaches for UEP, as well as the MIMO techniques for UEP, are considered in conjunction with FEC and with leaky/partial prediction mechanisms for scalable video. We also consider UEP for cooperative communications, where a virtual MIMO array is formed out of cooperating nodes. Lastly, we investigate the effects of delay considerations in UEP. TC: Small: Layered Modeling for Design, Analysis, and Implementation of Trusted Platform Applications As trusted computing technology, and trusted platform modules (TPMs) in particular, become widespread, it is important to build a strong foundation for applications built on this new technology. This project is a structured effort to develop a model for trusted platform applications, using a layered approach that supports rigorous analysis and security proofs, and provides a basis for modular software design that directly maps to the analytic model. Project activities are organized toward five specific objectives: development of concrete mathematical models for rigorous security analysis; analysis of specific widely applicable functionality built on TPMs; development of a timing accurate, extensible TPM simulator; study of TPM use in applications; and development of a layered security framework corresponding to the formal model. The project s approach is guided by two philosophies: keep it structured and simple (the principle of economy of mechanism) and justify constructions through rigorous analysis and proof.

Despite the growing use of trusted computing technology in modern systems, there has been very little formal research regarding this new technology. This project will fill this gap and provide a basis for future work in both using trusted computing technology and in designing extensions to the current technology. In addition to publications describing the knowledge gained, software for the extensible TPM simulator and the layered trusted computing framework will be distributed freely. The project will also result in the creation of a website that will be a portal for information on trusted computing and related research, and will include the development of a course on trusted computing and trusted platforms that will prepare both undergraduates and graduate students for work in this emerging area. SHF: Small: Statistical Analysis of Software This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The project investigates statistical software analysis, which infers
relationships among program components by using statistical properties
derived from multiple program executions.

To motivate statistical techniques, it is useful to draw analogies to
static analysis methods. Static analysis is about inferring
dependencies between program components: If a value is changed in one
component, how does that affect a value in a different component?
Static analysis tends to work best for properties that are local,
meaning the pieces of the program we are trying to relate are not
separated by a great deal of other computation. The statistical analog
of dependencies is correlation. Instead of proving definitively via
static reasoning the presence or absence of dependencies, we can
observe at run time that some properties of two components have high
or low correlation. Importantly, correlation is not affected by
syntactic or even dynamic locality: if two components have a
correlation, regardless of how much time or computation passes between
the execution of one component and the execution of the other, this
correlation can be detected if the appropriate statistical question is
asked.

The initial focus is on using cross correlation, which which computes
the maximum correlation between two sequences of observations, to
formalize statistical correlation between software components that
have a direction in time. This idea gives rise to a natural graph that
captures the strength and direction of statistical influence one
component has upon another; these graphs are analogous to traditional
dependency graphs, but have unique and useful properties. SHF: AF: SMALL: Scalable Symbolic Analysis of Hybrid Systems This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).


Embedded systems, such as controllers in automotive, medical, and avionic systems, consist of a collection of interacting software modules reacting to a continuously evolving environment. The emerging theory of hybrid systems systems with tightly integrated discrete and continuous dynamics, offers a foundation for model based design of embedded systems. For analyzing hybrid systems models, there are two prominent trends: an integral component of industrial modeling environments is numerical simulation, while a number of academic tools support formal verification of safety requirements using symbolic computation of reachable states of models. The proposed research is aimed at developing symbolic analysis techniques for simulation trajectories so as to significantly improve the simulation coverage. For this purpose, a new instrumentation scheme that would allow simulation engines to output, along with the concrete simulation trajectory, the symbolic transformers, will be developed. For managing complexity of symbolic analysis using polyhedra, new approximation schemes will be explored. The proposed algorithms will be implemented and evaluated in an analysis tool built on top of the widely used Stateflow/Simulink toolkit. The research results will be integrated in Penn s new Masters program in Embedded Systems that will train students in fundamentals of embedded systems design and implementation. CIF: Small: Removing Inherent Instabilities in Communication Networks Proposal 0915784: Removing Inherent Instabilities in Communication Networks





Abstract

Since the late 60s, communication networks have experienced dramatic changes, including: growth to unforeseen scales; operation under very dynamic and adverse conditions; integration of storage and transfer of all media; ubiquitous presence in all parts of our lives; etc. Going forward, these trends will continue to increase and broaden, resulting in mounting stress on the existing networks and, thus, growing emphasis on new network designs, often referred to as the ?clean slate architectures?.

Hence, in search of better designs, it is necessary to reexamine the existing network design principles, especially those that are inherent to all networking layers, such as the retransmission based failure recovery. To this end, recent work by the investigator discovers an entirely new networking phenomenon by showing that retransmissions can cause long ( tailed) delays and instabilities even if all traffic and network characteristics are light tailed, e.g. exponential or Gaussian. This finding is especially crucial for highly congested multi hop wireless networks that are characterized by frequent failures, e.g. ad hoc and sensor networks. Since the retransmission based failure recovery is at the core of the existing networks, this new phenomenon sets the basis for many more discoveries in this domain along the vertical (protocol stack), temporal and spatial network dimensions. Furthermore, this research also investigates how widely deployed fair resource sharing mechanisms come with a price since they may be responsible for spreading the long tailed delays to the entire network. Finally, based on the critical study of the exiting protocols, the investigator pursues careful redesign of network protocols that are shown to cause or spread long delays and instabilities. The general focus is on designing algorithms that are easy to implement, adaptive, scalable and provably near optimal. RI:HCC:Small:Preference Aggregation: Bypassing Worst Case Protections Elections are a broad model for collective decision making. Since around 1990, worst case hardness notions (most particularly, NP hardness) have been widely studied as a method for protecting election systems from manipulation, bribery, and control. Such protective worst case results have by now been obtained for many problems and many election systems.

The goal of this project is to study the ways that these protective results can be bypassed for the election manipulation, bribery, and control problems. This project will seek to transform the way security of elections is viewed: to make vividly clear by actual proofs and algorithms that worst case protections can on important real word systems and situations be shredded, and thus that bypass attacks are a true threat. The project will do this through exploring the extent of worst case protections and by finding the extent to which those protections can be bypassed, via studying restrictions on and assumptions about models, domains, and distributions.

This project involves a wide range of broader impacts, including information dissemination, bringing together local researchers interested in computational social choice, training of students, and service to the community. In addition, the topic itself is of broad relevance to society. Elections are of great importance both in human settings and in a rapidly expanding range of electronic settings, and indeed the study of elections is of active interest in computer science, economics, political science, operations research, and mathematics. The core research of this project seeks to better understand when the protection offered by worst case hardness results about election systems can be bypassed, and thus is relevant within a broad range of contexts in which elections are used for collective decision making: from spam filtering to critical human elections to sports tournaments to multiagent systems. Showing which important, known worst case safe election systems are vulnerable to bypass attacks serves the interest of the citizenry, since system designers can then avoid those systems, and in the long run more broadly secure systems can be developed. NeTS:Small:Collaborative Research:Holistic and Integrated Design of Layered Optical Networks The dramatic increase in throughput demands on transport systems has propelled the development of all optical networks. These networks can provide tremendous capacity when they are designed with their own limitations in mind, such as coarse wavelength granularity and physical impairments. In this research we consider the holistic design of optical networks that include the interdependence of three network layers: the traffic grooming layer, the lightpath management layer, and the physical fiber layer.

The network is first viewed from the top down, where sub wavelength circuit requests arrive with specific quality of service requirements. Current traffic grooming approaches are altered to incorporate their dependence on the lightpath management and physical layer constraints.
The system is then examined from the bottom up, so that the quality of transmission and efficiency of resource utilization can be optimized as the higher layer protocols evolve. Total capacity is measured from an information theoretic view point and system optimization uses ideas from game theory.

The results of the research will be practical algorithms for improved capacity and survivability of future optical networks as well as providing a quantitative proof of their superiority. The enhancement of network capability will help satisfy our society?s ever increasing need for information. It encourages the development of applications that require significant bandwidth. It also stimulates cross fertilization of ideas from the two fields of networking and communications. The algorithms and software will be made publicly available via a web site. The research enhances the education of the diverse group of graduate and undergraduate students participating in it. CIF: Small: Non Linear Processing and Coding for Compressive Sensing with Applications in Imaging This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).

Recently, the new field of compressive sensing has emerged with the
promise to revolutionize digital processing broadly. The key idea is
the use of nonadaptive linear projections to acquire an efficient,
dimensionally reduced representation of a signal or image directly
using just a few measurements. Surprisingly, Nyquist rate sampling,
which has dominated how signals are acquired and processed in science
and technology since the origin of digital processing, can be leaped
over through compressive sensing theory. Such results may have a
profound impact broadly, including applications in spectroscopy,
imaging, communications, as well as consumer electronics.

In this research, the compressive sensing framework is ``extended to
make it an integral part of a discrete time all analog processing
communications system that completely skips the digital domain and
shows an excellent robustness against noise. The key idea is the use
of non linear mappings that act as analog channel encoders, processing
the samples or compressive sensing measurements proceeding from an
analog source and producing continuous amplitude samples that are
transmitted directly through the noisy channel. Thus, all the
processing in the communications system is made in the analog domain.
In addition to its theoretical interest, the potential of this
approach in practical systems is substantial. For instance, the
proposed framework is readily applicable in practical systems such as
imaging, where it presents a performance that is very close to the
theoretical limits and clearly outperforms systems based on standard
compressive sensing. Furthermore, the idea of completely avoiding the
digital domain can be applied not just in point to point
communications systems, but also in more complex communications
problems such as distributed coding in sensor networks. III: Small: Quality Assessment of Computational Protein Models The PI plans to develop a new and robust hierarchical multi resolution framework that can be used for assessement of model uncertainty, especially that associated with prediction of protein structure. The approaches also enable respresentation of the structure in ways that enable rapid searching of structure space. It is important to establish quality estimation methods for predicted structures so that they can be used wisely by knowing the limitations of the model. Protein tertiary structure prediction has made steady progress in the past decade. However, current prediction methods are still not capable of producing highly accurate models on a regular basis. Practical use of prediction methods by biologists is limited not only the accuracy of current prediction methods but also the lack of error estimation of the models they produce. Moderately accurate models are still useful for many purposes, including design of site directed mutagenesis experiments and structure based function prediction, if the possible error range is understood. Resulting quality assessment methods will also contribute to improvement of protein structure prediction methods. In addition, structure models of proteomes of model organisms will be constructed with quality assessment data and will be made available to the public through the Internet. The proposed project leverages Purdue University s efforts in interdisciplinary computational life science and engineering by training graduate students and undergraduate students of different backgrounds through interdisciplinary coursework and direct involvement in the project. SHF: Small: Analyzing and Modeling Natural Language Usage in Software to Improve Software Maintenance Tools This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).

Large scale software is difficult to maintain, modify and keep up to date. There is a dire need for automated support for software system navigation, search, and comprehension activities to assist software developers and maintainers. This project addresses this issue by tackling some foundational issues and providing innovative methods for useful automated support for software system navigation, search, and comprehension activities. The research will lead to automatic analyses of the programmer s words and their relationships, through their usage in code, which will elucidate the concepts and actions encoded in the program. By building the conceptual models from the ``sound bites using context and program structure, a more complete picture is recovered enabling transformative improvements in automated support for software maintenance and program comprehension.

The analyses are heavily driven by natural language processing, information retrieval, and machine learning. The research will advance the theory and development of software maintenance tools, which will improve the effectiveness of software maintainers, decreasing software maintenance costs and raising software quality as software evolves. In developing a new required course on software engineering in the small, the PIs will incorporate learning how to use and evaluate software maintenance tools, including those developed in this project. SHF: Small: RUI: Collaborative Research: Accelerators To Applications Supercharging the Undergraduate Computer Science Curriculum The project integrates graduate research activities in hybrid, accelerated computing applications with undergraduate computer and computational science curricula, preparing undergraduates for graduate school and industry professions with application development experience in technologies essential to emerging high performance computing and peta scale systems. Curriculum enhancements across multiple computer science and engineering courses are investigated using real research activities to identify specific improvements needed at the undergraduate level. The research focuses on the use of leading accelerator technologies (multi core CPUs, GPUs, and FPGAs) in real scientific computing challenges and translating the insights, concepts, and examples for use in undergraduate computer science and engineering instruction. The significance of pairing research investigations with curricular development affords the opportunity to bring real experiences into the undergraduate classroom. Research level investigations will help to characterize the unique inter dependency of computer architectures and high performance applications. The resources, strategies and examples created in this project are available to undergraduate programs across the country that wish to provide instruction on the next generation hardware and software environments. The project also reaches several underrepresented populations through outreach efforts at local high schools, regional HBCUs, and leverages existing REU programs. NeTS: Small: xBeam: Cross Layer Beamforming Against Denial of Service Attacks in Wireless Networks This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5.)

It is commonly understood that to achieve a desired level of security in wireless networks is a bigger challenge than in wired networks. The denial of service (DoS) attack is a prominent threat in wired networks, and is a more potent threat in the wireless domain. This project aims to explore enabling approaches for a novel beamforming framework with cross layer interactions, referred to as xBeam, as a defense against DoS attacks in wireless networks. This project examines various DoS attacks, develops xBeam algorithms, evaluates the effectiveness of xBeam in deterring DoS attacks, and validates the algorithms using a wireless test bed.

Intellectual merit: This project develops a novel defense approach against DoS attacks in wireless networks, which is based on adaptive beamforming with cross layer interactions. The approach can deter DoS attacks across different networking layers, multiple DoS attacks mounting from different spatial directions, and attacks mounted by malicious attackers who are mobile.

Broader impact: This research contributes to a substantial improvement in wireless network security and contributes to the advance of a networked mobile wireless information society. The project supports research oriented education, by providing opportunities for advanced study of wireless network security through coursework and hands on experimentation. This research also provides quality research experiences for undergraduate students in the summer, particularly students from underrepresented groups. III: Small: Collaborative Research: Suppressing Sensitive Aggregates Over Hidden Web Databases: a Novel and Urgent Challenge The objective of this project is to understand, evaluate, and contribute towards the suppression of sensitive aggregates over hidden databases. Hidden databases are widely prevalent on the Web, ranging from databases of government agencies, databases that arise in scientific and health domains, to databases that occur in the commercial world. They provide proprietary form like search interfaces that allow users to execute search queries by specifying desired attribute values of the sought after tuple(s), and the system responds by returning a few (e.g., top k) satisfying tuples sorted by a suitable ranking function.

While owners of hidden databases would like to allow individual search queries, many also want to maintain a certain level of privacy for aggregates over their hidden databases. This has implications in the commercial domain (e.g., to prevent competitors from gaining strategic advantages) as well as in homeland security related applications (e.g., to prevent potential terrorists from learning flight occupancy distributions). The PIs prior work pioneered techniques to efficiently obtain approximate aggregates over hidden databases using only a small number of search queries issued via their proprietary front end. Such powerful and versatile techniques may also be used by adversaries to obtain sensitive aggregates; thus defending against them becomes an urgent task requiring imminent attention. This project investigates techniques to suppress the sensitive aggregates while maintaining the usability of hidden databases for bona fide search users. In particular, it explores a solution space which spans all three components of a hidden database system: (1) the back end hidden database, (2) the query processing module, and (3) the front end search interface. The intellectual merit of the project is two fold: (1) problem novelty: it initiates a new direction of research in information privacy of suppressing sensitive aggregates over hidden databases, and (2) solution novelty: it investigates a variety of promising techniques across the three components. The outcomes of this research have broader impacts on the nation s higher education system and high tech industries. Parts of the project will be carried out by students of the University of Texas Arlington and George Washington University as advanced class projects or individual research projects. CIF: Small: Signal Processing for Multi user Communications under Finite Alphabet Constraints Abstract for NSF Proposal #0915846


Abstract:
Channel capacity and mutual information have been studied extensively for various types of wire line and wireless multi user communication channels. Among the vast information theoretic literature, most of the results are based on the assumption that the channel inputs are Gaussian distributed. However, Gaussian inputs can never be realized in practical systems. The inputs are usually taken from finite alphabets, which can significantly depart from Gaussian distribution. A large nonlinear gap exists between the theoretical capacity and practical achievable rate. This nonlinear gap indicates that Gaussian input assumption may not provide a realistic design guideline to practical systems. Maximizing mutual information over channels with finite alphabet inputs will benefit not only bandwidth efficiency but also bit error rate performance. However, much less work has been done for this important topic. This is mainly due to lack of closed form solution and high computational complexity.

The project investigates the direct maximization of mutual information and throughput over multi user channels with finite alphabet inputs. The computational complexity problem is tackled by developing mathematically tractable and practically accurate algorithms via employing graph based message passing techniques. Shaping matrices are introduced to the maximization of mutual information. Parameterized approaches are developed to solve optimal shaping matrices which lead to the global maxima of the mutual information. Research efforts focus on multiple access channels, broadcast channels and interference channels, which are the fundamental channel scenarios of multi user communications. The channel state information and/or channel covariance information are made available to the transmitter and receiver for the maximization of mutual information. Both frequency flat fading and frequency selective fading channels are explored. CSR:Small: TEAMS Transplanting Artificial Life Behaviors to Mobile Robots This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The TEAMS project applies digital evolution to the design of robust communication services for cooperating groups of robots. Example applications include teams of robots for disaster relief operations, assisting humans in dangerous occupations, and monitoring of critical infrastructure and public service facilities such as drinking water reservoirs. In digital evolution, a population of self replicating computer programs exists in a user defined computational environment and is subject to mutations and natural selection. Evolved algorithms
(sequences of instructions comprising the genomes of digital organisms) can be recompiled to execute directly on hardware platforms. The project explores three overlapping classes of behavior: distributed self organization of nodes; collective communication operations that tolerate adverse conditions; and mobility aided communication methods, where nodes change their locations in order to improve communication performance. Evolved behaviors are evaluated on an NSF sponsored digital evolution testbed containing a heterogeneous collection of microrobots. The ability to observe the evolutionary process in digital organisms provides a powerful method to investigate the driving forces in complex systems. The TEAMS project exploits this open ended method of exploration to find better ways to construct software for mobile robotic systems that need to interact with the physical world, and each other, in order to solve problems. On a broader scale, the project includes several innovative educational and outreach components, such as an after school program to provide economically disadvantaged K 12 students with exciting hands on experiences involving evolutionary computation, adaptive software, and robotics. RI: Small: Intelligent Compressive Multi Walker Recognition and Tracking (iSMART) through Pyroelectric Sensor Networks Although high cost, data intensive multi camera systems have been widely used for mobile human tracking and recognition, the pyroelectric infrared (PIR) sensor has a variety of advantages including dramatically low costs, chemical stability, high sensitivity to human body thermal variation, and extremely low sensory data throughput.

This project implements an Intelligent Compressive Multi Walker Recognition and Tracking (iSMART) testbed based on PIR Sensor Networks (PSN). The novelties of iSMART include three aspects: (1) Context aware region of interest (RoI) exploration to achieve an inherent tradeoff between area of sensor coverage and degree of information acquisition resolution. This research uses strict mathematical models to measure RoI context. (2) Decentralized inference / learning for in network intelligence. This project develops a belief propagation based distributed inference scheme with data to object association for continuous tracking and recognition of multiple walkers. It uses orthogonal projection based distributed learning for sensor calibration and feature model training. (3) Networked, compressive sampling structures and sensing protocols. This project extends the latest progress in compressive and multiplex sensing theories to guide the design of novel networked sensor receiver pattern geometries and decentralized sensing protocols.

The above research efforts will lead to a novel low cost, high fidelity wireless distributed sensing system for multiple walker recognition and tracking. As an alternative to video camera systems, iSMART systems can be widely deployed to automatically monitor airports, customs / harbors, and other critical national infrastructures. This project will also generate interesting hands on labs on intelligent sensor / sensor networks and class projects for both undergraduate and graduate students. TC:Small: Formal Reasoning about Concurrent Programs for Multicore and Multiprocessor Machines Formal reasoning about concurrent programs is usually done at a
high level with strong assumptions such as built in thread primitives
(e.g., locks) and simplified memory models (e.g., sequential
consistency). Existing formal techniques (including Hoare logic and
type system) have consistently ignored important issues such as
relaxed memory models, hardware interrupts, implementation of
synchronization primitives, and support for software transactional
memory and privatization. This severely limits their applicability.

This research focuses on extending and adapting existing formal
techniques so that they can also support realistic low level
concurrent programs running on modern multicore and multiprocessor
machines. The PI is developing a new operational approach for
reasoning about programs running under relaxed memory models;
designing new program logics for certifying both weak and strong
memory operations (including the memory fence and compare and swap
instructions); and showing how to scale his approach to real world
thread implementation and to machines with relaxed memory models. If
successful, this research will help improve the reliability of
concurrent software components, which form the backbone of many
critical systems in the world. It will also facilitate the
community wide effort for finding new programming models for safe and
scalable multicore computing. AF: Small : Collaborative Research: The Polynomial Method for Learning The broad goal of this line of research is to give a principled answer to the question, What sort of data is efficiently learnable, and by what algorithms? The current state of the art in machine learning is that there is an overwhelming number of possible algorithms that can be tried on a new machine learning problem, with no clear understanding of which techniques can be expected to work on which problems. Further, it is often the case that machine learning algorithms that work well in theory do not perform as well in practice, and vice versa. The PIs have outlined a plan for resolving these difficulties, finding a unification of disparate methods via the Polynomial Method, and investigating how efficient this method can be. On a more immediate level the PIs will aim for broad impact through advising and guiding graduate students and widely disseminating research results.

Specifically, the PIs will investigate the effectiveness of the Polynomial Method in machine learning theory. The PIs observe that nearly all learning algorithms, in theory and in practice, can be viewed as fitting a low degree polynomial to data. The PIs plan to systematically develop this Polynomial Method of learning by working on the following three strands of research:

1. Understand the extent to which low degree polynomials can fit different natural types of target functions, under various data distributions and noise rates. This research involves novel methods from approximation theory and analysis.

2. Develop new algorithmic methods for finding well fitting polynomials when they exist. Here the PIs will work to adapt results in geometry and probability for the purposes of identifying and eliminating irrelevant data.

3. Delimit the effectiveness of the Polynomial Method. The PIs will show new results on the computational intractability of learning intersections of linear separators, and on learning linear separators with noise. AF: Small: Fundamental Algorithms based on Convex Geometry and Spectral Methods This project addresses fundamental open problems in the theory of algorithms using tools from convex geometry, spectral analysis and complexity. The research will also provide algorithmic insights into central questions in functional analysis and probability. The problems tackled are of a basic nature, and originate from many areas, including sampling, optimization (both discrete and continuous), machine learning and data mining. With the abundance of high dimensional data in important application areas, the need for efficient tools to handle such data is pressing and this project addresses the most basic questions arising from this need. Progress on these problems is sure to unravel deep mathematical structure and yield new analytical tools. As the field of algorithms continues to expand (and extends its reach far beyond computer science), such tools will play an important role in forming a theory of algorithms. The PI currently serves as the director of the Algorithms and Randomness Center, founded in 2006 on the premise of outreach to scientists and engineers and to identify problems and ideas that could play a fundamental role in computational complexity theory. The research results form the basis of courses at both the undergraduate and graduate level with materials available online.

The project has four focus topics:
(1) Complexity of tensor optimization. Does there exist a polynomial time algorithm for computing the spectral norm of an r fold tensor?
(2) Affine invariant algorithms. Can linear programs be solved in strongly polynomial time? What is a natural affine invariant and noise tolerant version of principal components analysis?
(3) Complexity of sampling high dimensional distributions, both upper and lower bounds. What classes of nonconvex bodies can be sampled (optimized, integrated over) efficiently? Do there exist polynomial time better than exponential approximations to the shortest lattice vector?
(4) Learnability of high dimensional functions. Can the intersection of two halfspaces be PAC learned? Can the intersection of a polynomial number of halfspaces be learned under a Gaussian distribution? III: Small: Exploring Data in Multiple Clustering Views This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The primary objective of this research is to formulate a framework for a new paradigm for clustering: discovering all possible non redundant multiple clustering views from data. Typical clustering algorithms only find one clustering solution, but many real and complex data are multi faceted by nature. Data can be interpreted in many different ways. Given the same data, what is interesting to a physician will be different from what is important to an insurance agency. This research will provide new formulations, algorithms, and tools for exploratory data analysis that are widely applicable to many domains. The PI will apply the new algorithms in detection of skin lesions, including cancers/ This will involve the automatic segmentation of the dermis and epidermis junction in skin images, automated detection of machine sounds, and in developing algorithms for multiple non redundant clustering of text and natural images. Additionally, this project will provide research experiences for both undergraduate and graduate students in the classroom and in the lab. The PI will work with the Society of Women Engineers to inspire female students to pursue careers in engineering and computer science, and to ensure that under represented groups are involved in this research. SHF: Small: Virtual Probe: A Statistically Optimal Framework for Affordable Monitoring and Tuning of Large Scale Digital Integrated Circuits ID: 0915912
Title: Virtual Probe: A Statistically Optimal Framework for Affordable Monitoring and Tuning of Large Scale Digital Integrated Circuits
PI: Xin Li and Rob Rutenbar, Carnegie Mellon University

Abstract

On chip variability monitoring and post silicon tuning have emerged as a joint strategy to combat the deleterious effects of nanoscale process variations, to maintain the aggressive scaling of integrated circuits (ICs). However, the design overhead (e.g., die area, power consumption, etc.) of these new techniques is a growing problem as devices continue to shrink, and the relative magnitude of critical process fluctuations continues to grow. This project proposes to develop a novel statistical framework called virtual probe (VP) to minimize the overhead of variability monitoring and post silicon tuning. VP accurately predicts full chip spatial variation from the smallest possible set of measurement data, thereby enabling lowest cost / highest accuracy silicon testing, characterization and tuning as IC technologies move further into the nanoscale regime.

The proposed project aims to create a radically improved platform for on chip statistical monitoring and tuning of large scale digital ICs; it is expected to yield 5 10 times performance improvement for advanced ICs in a broad range of applications, from consumer electronics to aerospace controllers. In addition, the proposed mathematical framework is applicable to many other scientific and engineering problems and, hence, offers a new avenue to study and understand these. Finally, given its broad coverage over multiple research areas, the proposed project motivates close collaboration among statistician, computer scientists and circuit designers, thereby creating enormous opportunities for interdisciplinary innovations. The interdisciplinary nature of this project also offers an excellent opportunity to train the next generation of U.S. researchers in multiple science and engineering domains. AF : Small : The Theory and Practice of Hash Based Algorithms and Data Structures Hash based data structures and algorithms are currently a booming industry in the Internet, particularly for applications related to measurement, monitoring, and security. Hash tables and related structures such as Bloom filters are used billions of times a day, and new uses keep proliferating. There remain, however, large gaps between the theoretical design and analysis of these structures and their use and implementation in practice. This research aims to bridge the gap between the theory and practice of algorithms and data structures that utilize hashing, with an emphasis on networking applications. The outcomes of this research will include tools and frameworks for translating theoretical results into real world settings, better analyses and implementations of existing algorithms and data structures, and the development and analysis of new algorithms and data structures. Related educational efforts will focus on methods to make undergraduate students, graduate students, and the professional networking community more aware of the potential and power of hash based approaches, thereby expanding the reach and influence of theoretical work in the area. RI: Small: Visual Parts for Image and Video Analysis Images and video are analyzed in terms of parts. Qualitatively,
a part of an image is a region that can be extracted consistently, even if the image changes somewhat because of differences in quality, resolution, lighting, viewpoint, or small variations in the shape of the objects being depicted. In video, these regions are extruded into time, forming tubes of sorts that persist over relatively long time intervals.

Technically, parts are regions with high saliency and stability, two notions that are defined anew in this research based on mathematical tools that span from harmonic functions to new developments in computational topology and spectral graph theory.

Parts are a key handle into image and video structure, as they allow describing the visual information succinctly and in a stable manner. They lead to indices for retrieval, and provide primitives that make it possible for computer software to recognize objects and activities. The main result from this effort is a systematic comparison of advantages and limitations of the new definition with descriptors from the literature.

Applications of part based visual analysis range from image retrieval, medical and biological imaging, and video interpretation for military and intelligence scenarios, to surveillance, the annotation and editing of images and video clips, and more. Work involves graduate students and undergraduates funded through the NSF REU program. Results of this research are disseminated through scholarly publications and classes at Duke University. A benchmark of evaluation images and video is developed for open use by the research community. NeTS: Small: Collaborative Research: Supporting unstructured peer to peer social networking 0915928 NeTS: Small: Collaborative Research: Supporting unstructured peer to peer social networking

09/01/09 08/31/12

G. de Veciana, PI, U.T. Austin, and G. Kesidis, PI, Penn State

Award Abstract:

Peer to peer systems have seen continued growth, in terms of traffic volume, and as the architecture of choice to build new applications and network services ? notable benefits lie in their distributed design leading to higher reliability and flexibility. However as users/peers increasingly become content providers, and generally conduct more of their business on the network, privacy is a critical concern.

By leveraging peers? trust relationships through referral mechanisms based on underlying reputation systems, applications that deliver a new standard of privacy are being devised. Peer to peer systems that dynamically adapt, in a distributed and scalable manner, based on the outcomes of peer transactions, are being modeled and analyzed. The focus is on unstructured networks where peer membership correlations among communities of interest can be learned to improve the search performance of reputation biased random walks and limited scope flooding. Content sharing applications are being designed based that leverage this framework to incentivize cooperative behavior while enabling collaborative filtering and content pushing.

Expected results include the development analysis and testing of a new framework for privacy preserving search for large scale, unstructured, on line, peer to peer networks. Complementary incentive mechanisms resulting in improve file sharing and promoting honest referrals will be devised. The results will be disseminated through peer reviewed venues and, where possible, industry concerns, while data and simulation tools are made available on the web.

The efforts impact will lie in contributing new ways to improve privacy and promote more honest and efficient cooperation in large scale on line peer to peer systems, for content sharing, as well as a broader set of social networking applications. AF: Small: Collaborative Research: The Polynomial Method for Learning The broad goal of this line of research is to give a principled answer to the question, What sort of data is efficiently learnable, and by what algorithms? The current state of the art in machine learning is that there is an overwhelming number of possible algorithms that can be tried on a new machine learning problem, with no clear understanding of which techniques can be expected to work on which problems.

Further, it is often the case that machine learning algorithms that work well in theory do not perform as well in practice, and vice versa. The PIs have outlined a plan for resolving these difficulties, finding a unification of disparate methods via the Polynomial Method, and investigating how efficient this method can be. On a more immediate level the PIs will aim for broad impact through advising and guiding graduate students and widely disseminating research results.

Specifically, the PIs will investigate the effectiveness of the Polynomial Method in machine learning theory. The PIs observe that nearly all learning algorithms, in theory and in practice, can be viewed as fitting a low degree polynomial to data. The PIs plan to systematically develop this Polynomial Method of learning by working on the following three strands of research:

1. Understand the extent to which low degree polynomials can fit different natural types of target functions, under various data distributions and noise rates. This research involves novel methods from approximation theory and analysis.

2. Develop new algorithmic methods for finding well fitting polynomials when they exist. Here the PIs will work to adapt results in geometry and probability for the purposes of identifying and eliminating irrelevant data.

3. Delimit the effectiveness of the Polynomial Method. The PIs will show new results on the computational intractability of learning intersections of linear separators, and on learning linear separators with noise. TC: Small: Cyber Security of Electric Power Infrastructure: Risk Modeling and Mitigation Abstract: Critical infrastructures are complex physical and cyber based systems that form the lifeline of modern society, and their reliable and secure operation is of paramount importance to national security and economic vitality. In particular, the cyber system forms the backbone of a nation?s critical infrastructures, thus a major cybersecurity incident could have significant negative impacts on the reliable, efficient, and safe operations of the physical systems that rely upon it. Recent findings, as documented in government reports and literature, indicate the growing threat of cyber based attacks on our nation?s power grid and other critical infrastructures. The goal of this project is to develop a comprehensive cybersecurity framework and algorithms for securing the electric energy infrastructure, and implement a novel curriculum, which include: (i) developing an integrated risk modeling methodology that models both cyber attacks on the Supervisory Control and Data Acquisition (SCADA) system and the resulting impacts on the performance and stability of the power grid; (ii) developing risk mitigation algorithms, both in cyber and power system domains, to prevent and mitigate cyber attacks on the power grid; (iii) evaluating the risk models and algorithms through a combination of model, simulation, and testbed based evaluations using realistic system topologies and attack scenarios; (iv) implementing a novel curriculum on cybersecurity of critical infrastructure systems through graduate courses and undergraduate projects. This project?s outcome will have broader impacts in securing our nation?s power grid and other critical infrastructures against cyber attacks, and creating a skilled workforce in this critical area of national need. AF: Small: Computational Complexity from Physical Constraints The project aims to measure the computational difficulty of determining various aspects of physical systems ?quantum mechanical systems in particular. One such computational task, of interest to physical chemists, is to distinguish the minimum energy configuration of a system of nuclear spins. No efficient algorithm running on a traditional computer is known to solve this or similar problems.

Quantum computers are devices that directly harness the laws of quantum mechanics to do computation. Although not yet implemented at a large scale, they hold great promise in speeding up certain computations, including efficiently simulating actual physical systems. Yet several interesting questions about physical systems appear beyond the reach of even quantum computers. The project will investigate problems of this sort, using techniques from theoretical computer science to help determine their intrinsic quantum computational difficulty. Showing that some of these questions are difficult for a quantum computer will help guide scientists and engineers toward more practical, less ambitious approaches.

More broadly, the project will build on previous work of the Principal Investigator and colleagues to better understand the power of quantum information processing in general and how it relates to traditional information processing. AF: Small: Algorithmic and Quantitative Problems in Semi algebraic and O minimal Geometry This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The goal of the proposed research is to further our understanding of algorithmic and quantitative semi algebraic geometry, develop new techniques especially coming from algebraic topology and the theory of o minimal structures, and broaden the applications of semi algebraic geometry in other areas such as discrete and computational geometry.

Algorithmic semi algebraic geometry lies at the heart of many problems in several different areas of computer science and mathematics including discrete and computational geometry, robot motion planning, geometric modeling, computer aided design, geometric theorem proving, mathematical investigations of real algebraic varieties, molecular chemistry, constraint databases etc. A closely related subject area is quantitative real algebraic geometry. Results from quantitative real algebraic geometry are the basic ingredients of better algorithms in semi algebraic geometry and play an increasingly important role in several other areas of computer science: for instance, in bounding the geometric complexity of arrangements in computational geometry, computational learning theory, proving lower bounds in computational complexity theory, convex optimization problems, etc. As such, algorithmic and quantitative real algebraic geometry has been an extremely active area of research in recent years.


The proposed research will develop new techniques in real algebraic geometry that would lead to new and better algorithms, for computing topological invariants of semi algebraic sets in theory, as well as practice. In addition, several open problems in quantitative real algebraic geometry and closely related problems in discrete and computational geometry will be attacked with
the mathematical tools developed by the PI. All these research objectives will be integrated in a broad program of training graduate students and curriculum development. III: Small: Avoiding Contention on Multicore Machines To take full advantage of the parallelism offered by a multicore
machine, one must write parallel code. Writing parallel code is
difficult. Even when one writes correct code, there are numerous
performance pitfalls. For example, an unrecognized data hotspot could
mean that all threads effectively serialize their access to the
hotspot, and throughput is dramatically reduced.

This project aims to provide a generic framework for performing
certain kinds of concurrent operations in parallel. Infrastructure is
provided to perform those operations in a scalable way over the
available threads in a multicore machine, automatically responding to
hotspots and other performance hazards. The goal is not to squeeze
the last drop of performance out of a particular platform. Rather,
with the planned system a programmer can, without detailed knowledge
of concurrent and parallel programming, develop code that efficiently
utilizes a multicore machine.

The project involves the development of algorithms and data structures
designed for the efficient parallel execution of generic code
fragments. The primary focus is on data intensive operations as would
typically be found in an in memory database engine. Critical research
questions include how to design generic multi threaded operators that
can be applied to a range of computations, how to avoid cache
thrashing, and how to implement the framework in a way that works on a
variety of hardware platforms. Performance improvements in throughput
of an order of magnitude are expected relative to naive solutions that
suffer from contention. The project aims to achieve performance close
to that of hand tailored expert written parallel code, with far less
coding effort.

This project has immediate applications in both commercial and
public domain database systems where performance improvements would
enhance the experience of database system users, and reduce hardware
and energy requirements for a given level of performance.

Programmability improvements would allow programmers without expertise
in parallel programming to effectively use multicore machines. The
project also provides the focus for an advanced level course on
database system implementation for multicore machines. The software
infrastructure will be made available for research use by others. CSR:Small: A Control Theoretic Approach to Simultaneously Meeting Timing and Power/Thermal Constraints for Multi Core Embedded Systems This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

High performance real time embedded systems have stringent requirements for key performance properties, such as end to end timeliness and reliability, in order to operate properly. In recent years, with the continuously decreasing feature size and increasing demand for computational capabilities, today s high performance embedded systems face an increasing probability of overheating and even thermal failures. As a result, their power consumption and temperature must be explicitly controlled for improved reliability. On the other hand, multi core processors have recently become the main trend in the current processor development, due to some well known technological barriers. Consequently, future high performance embedded systems can be expected to be equipped with multi core or even many core processors. Therefore, new control algorithms need to be developed to simultaneously meet the timing and power/thermal constraints for multi core embedded systems.

This project aims to develop a holistic framework, based on recent advances in feedback control theory, to meet both timing and power/thermal constraints for high performance embedded systems with multi core processors. Our framework can make three major contributions. First, our solution can coordinate different control strategies to meet both constraints with guaranteed system stability. Second, we propose novel power/thermal control (capping) algorithms for multi core processors to achieve improved control accuracy and system performance. Third, we propose new feedback scheduling algorithms to utilize the new features available in multi core processors, such as shared L2 caches and per core dynamic voltage frequency scaling (DVFS), for improved real time performance. We also plan to investigate heterogeneous cores, memory power management, and system controllability for better power/thermal control and timeliness guarantees. CIF: Small: Optical Diffusion Tomography, with Application to in Vivo Fluorescence Resonance Energy Transfer Imaging This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).


Optical Diffusion Tomography, with Application to In Vivo Fluorescence Resonance Energy Transfer Imaging

Kevin Webb, Purdue University

Optical sensing and imaging will continue to become more important for in vivo medicine. In most cases, light can be modeled with a diffusion equation, and the reconstruction of images based on this model is the basis of optical diffusion tomography (ODT). Enhanced contrast can be achieved with targeting of a fluorophore to cancer cells, for example, and targeted anti cancer drugs can be delivered. Another molecular imaging opportunity involves fluorescence resonance energy transfer (FRET) parameters. FRET has proved to be of immense value in the study of chemical transport into cells and the underlying cause of disease, and by coupling to ODT (FRET ODT), there is the opportunity to transfer this knowledge to in vivo studies. While substantial progress has been made in various ODT modalities, the achievable resolution and the computational burden impede effective applications. More efficient imaging strategies are thus essential.

This research involves the development of a method for deep tissue imaging of FRET parameters (FRET ODT) and fast, accurate and robust methods for optical diffusion tomography. The recent demonstration by the Webb group that it is possible to image FRET parameters using heavily scattered light is being expanded into a method for imaging FRET in vivo. This involves a solution for the intramolecular FRET parameters with both rigid and flexible linkers that are incorporated as unknown sources in a diffusion equation representation for the donor fluorescence. Multigrid algorithms are being developed and applied to fluorescence imaging and FRET ODT. A model based non iterative image reconstruction method, that has proved to substantially reduce computation time in preliminary studies, is being applied to image FRET kinetic parameters. RI: Small: Modeling Cities by Integrating 3D and 2D Data This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The problem of automated 3D reconstruction and modeling of urban environments is of great interest and importance in the fields of computer vision and graphics. This is due to the fact that accurate 3D city models are paramount in the further development of a variety of fields. This project acquires vast amounts of data via the latest generation in laser scanning technology and high resolution cameras in order to achieve the goal of 3D city modeling. That data is registered in a common frame of reference based on current techniques that are improved in order to facilitate voluminous point clouds. One of the major challenges is the complexity of the acquired dataset that needs to be simplified for efficient rendering and higher level recognition algorithms. Current simplified methods produce reduced sets of triangles that lack higher order labels. This project segments and classifies the data into coarse and fine urban elements (such as facades, windows, vegetation, etc.), discovers symmetries among the elements, and fills missing data. It also integrates image based and laser based models in order to enhance the acquired geometry, and to produce a seamless visualization result. The above will also aid recognition processes, alternative visualizations and truly photorealistic representation of the 3D scene. This work is disseminated through collaborations with the geography and film and media departments of Hunter, as well as other agencies (such as museums). It has significant impact for urban planning, architecture, and archeology applications, and in systems such as street map visualization, as well as in film and construction industries. SHF: Small: Automating the Deployment of Distributed Real time and Embedded System Software using Hybrid Heuristics based Search Techniques This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Modern automobiles and flight avionics systems form complex distributed real time and embedded (DRE) systems. A high end luxury car can have over 80 Electronic Control Units (ECUs), which are small embedded processors, and multiple networks linking the processors. Furthermore, several hundred software components can be distributed across these multiple networked ECUs. Optimizing the deployment of these software components, by packing the software more tightly onto the processors, can reduce the size of the required underlying infrastructure and have numerous positive side effects, such as weight and power consumption savings.

Determining how to deploy software to hardware in DRE systems is a challenging problem due to the large number of complex constraints that must be dealt with, such as real time scheduling constraints, component placement restrictions, and fault tolerance guarantees. This research effort focuses on developing new hybrid heuristic and meta heuristic techniques for determining how to deploy software to computational nodes. The algorithms and tools will be made available in open source through the Generic Eclipse Modeling System (http://www.eclipse.org/gmt/gems), which is distributed by 45 world wide mirrors, and the ESCHER tool repository. Opportunities for outreach will be sought through existing mechanisms in place at Vanderbilt University, such as the NSF Science and Technology Center called TRUST and Vanderbilt Center for Science Outreach (CSO) to host summer research students. We will continue to support graduate students belonging to underrepresented groups. SHF: Small: User Centered Software Analysis Tools The proposed work explores techniques for making static analysis tools
more effective in the hands of users. This will lead to the
development of new static analysis algorithms, techniques, and
interfaces that provide the information users need to find, verify,
and fix defects.

The core of the proposed work will be a framework for static analysis
visualization and interaction, with several components. First, the
framework will include checklists to help users triage defect reports,
i.e., decide whether they are true or false positives. The aim is to
develop ways to instrument static analyses to automatically generate
checklists based on imprecision introduced during the analysis.
Second, the framework will include lightweight query and search
facilities to help users work with static analysis tool results. Users
need effective ways to query the knowledge base generated by a tool
when trying to understand an error report; current static error
reports tend to provide too little or too much information. Third, the
framework will include a generic visualization for program paths, a
core part of many static analysis tools defect reports. While some
tools include simple path visualization, our framework will aim for
far more effective interfaces by applying information visualization
principles. AF: Small: Approximation, Covering and Clustering in Computational Geometry Computational geometry is the branch of theoretical computer science devoted to the design, analysis, and implementation of geometric algorithms and data structures. Computational geometry has deep roots in reality: Geometric problems arise naturally in any computational field that simulates or interacts with the physical world computer graphics, robotics, geographic information systems, computer aided design, and molecular modeling, to name a few as well as in more abstract domains such as combinatorial geometry and algebraic topology.

This research focuses on fundamental problems in computational geometry. These problems include set cover, hitting set, independent set, and other related problems. These problems have numerous applications from wireless networking to optimization.

The main theme of this research is to combine ``classical Computational Geometry techniques (like cuttings, epsilon nets, etc) together with techniques used in Operation Research (Linear Programming, rounding techniques, etc).

This research aims to greatly improve our understanding of the structure of these fundamental problems. The research may lead to improved approximation algorithms for these problems.

The algorithms and insights obtained from the technical work will benefit computer science and related disciplines where geometric algorithms are widely used. This research has potential to broaden the scope of Computational Geometry by introducing new techniques into the field.

A book partially based on the research in this award will be published in the near future. This will make the developed techniques available to wide audience consisting of students and researchers from several disciplines include engineering, mathematics, and the natural and social sciences. NeTS: Small: A Game Theoretic Framework for Agile and Resilient Wireless Systems Broadcast media underlying wireless networks enable diverse devices to communicate using shared channels, but also leave them severely exposed to adversaries. Our increased reliance on wireless networks for connectivity to the cyber infrastructure, and for monitoring our physical infrastructure, has opened the door for sophisticated denial of service attacks with potentially devastating effects on our economy and homeland security.

This project pursues a new game theoretic framework for agile and resilient wireless systems. The framework models the interaction between communicating nodes and adversaries as two player games, with strategies defined by various protocol choices and cross layer attacks. A central thesis of this game theoretic paradigm is that resilient wireless systems need to be highly agile, rapidly mixing strategies to thwart adaptive attackers.

The research methodology underlying this project is threefold. The first component formulates games capturing complex interactions between communicating nodes and adversaries across basic building blocks of wireless systems. The second component concerns optimization problems and equilibria associated with these games. The final component realizes the architecture and solutions in two
real world prototypes: a Linux based platform for 802.11 networks, and a Software Defined Radio platform for studying more sophisticated mechanisms.

This project addresses the critical need of protecting our wireless infrastructure from DoS attacks. The research conducted will lead to (a) new game theoretic analysis of wireless systems quantifying vulnerabilities and worst case scenarios, and identifying promising methods to thwart such attacks; (b) highly agile cross layer protocols that are resilient to the most adaptive adversaries; (c) new tools for implementing and evaluating these protocols in real world systems. NeTS:Small:Collaborative Research: Effective control of wireless networks via topology adaptation and randomization This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Wireless Mesh Networks have emerged as a solution for providing last mile Internet access. By exploiting advanced communication technologies, they can achieve very high rates. However, effectively controlling these networks, especially in the context of advanced physical layer technologies, realistic models for channel interference, and distributed operation, remains a major challenge. Hence, the project focuses on developing effective and practical network control algorithms that make efficient use of wireless resources through joint topology adaptation, network layer routing, MAC layer scheduling, and physical layer power, channel, and rate control. The design of the algorithms leverages recent developments in the control of dynamical systems and randomized algorithms, and takes into account realistic channel models. This includes: (i) topology adaptation algorithms that take advantage of channel allocation, power control, and the controlled mobility capabilities of some of the nodes to dynamically decompose the network into sub networks in which low complexity distributed scheduling and routing algorithms are guaranteed to achieve high throughput, (ii) randomized distributed algorithms that solve the scheduling and routing problems in a computationally efficient manner using only local topological and queue size information, and (iii) evaluation of the algorithms? performance in terms of throughput, delay, and complexity. The developed algorithms will enable highly efficient operation of wireless networks. The project incorporates training of graduate and undergraduate students, outreach activities to local high school teachers, and technology transfer to industry and government laboratories. CSR: Small: Meta Analysis Directed Execution This project aims to modify the representation of binary executable files by retaining information generated during various program transformations using an XML notation. If retrievable, the information lost during transformation processes could potentially improve the hardware/runtime system.

While high level programming languages support software development, computer architecture is implemented more efficiently around low level assembly/machine programming language. This gap between high level and machine level programming languages is bridged by translations performed by a compiler. A compiler performs significant analysis and translation on the high level program code program in order to generate optimized low level code. However, the original program s structural information is lost by the time the program has been translated into a low level representation. Consequently, an existing processor s architectures cannot benefit from such structural information.

Many dynamic optimizations performed in a processor, such as branch and value prediction, and many dynamic compiler optimizations, such as dynamic loop unrolling, can be expressed in a semantically rich binary file format. This project uses an XML based binary file format to express program structure. The program metadata is expressed as XML namespace tags. A processor, consisting of a meta engine to interpret the program level structure semantic metadata, transforms the binary program in order to affect the specified dynamic optimizations before handing it over to a classical execution engine. This approach opens up many performance enhancement opportunities, controlled by the program itself. In this seed project, the execution engine is realized through a simulation environment based on Open DOS (Open Source Dynamic Optimization). A proof of concept compiler, XMLgcc, generates the metadata tagged binary files.

This project will result in a transformative view of processor and compiler design. This may spur processor development activity due to soft ization of many of the current hard features of an architecture a corresponding compiler development has significantly lower overhead. The empirical nature of computer architecture and compilers requires a platform on which architecture and compiler variations can be implemented with low cost. Such a platform is an ideal pedagogical tool for exposing such what if iterative design process to computer architecture and compiler students.

The team will also develop instructional modules based on the SeeMe platform for various computer architecture and compiler (dynamic optimization) topics for graduate classes. RI small: Simultaneous Groupwise Nonrigid Registration, Segmentation and Smoothing of 3D Shapes and Images The goal of this research is the investigation of a computational framework that allows for group wise joint segmentation, smoothing and registration of images and shapes. With recent advances in sensor technology, images (shapes and other types of data) are being generated in abundance, and there is now a need for algorithms that operate and process images in collections instead of individually. In particular, this requires segmentation, smoothing and registration, three most important image processing operations, to be formulated in new ways that emphasize the relational aspects of their inputs. In addition, image data in computer vision applications are usually sampled from low dimensional manifolds embedded in high dimensional features spaces, and an important problem is to construct versatile and expressive computational models that exploit their geometries for solutions. The proposed computational framework addresses these two issues by formulating a variational framework that unifies smoothing, segmentation and registration. Specifically, it uses hypergraphs to model the multiple geometric relations among the inputs, and the three operations are integrated in one single discrete variational framework defined over a hypergraph. The proposed framework provides a foundation for several principled joint segmentation and registration algorithms for images and shapes that can guarantee crucial properties such as compatibility, consistency, unbiasedness and symmetry. Furthermore, it also provides a new and more discriminative numerical signature for 2D and 3D shapes that can be important for many shape related vision applications such as shape recognition, shape retrieval and image based medical diagnosis. NeTS: Small: RUI: Dynamic Indoor Location Computing and Benchmarking This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The problems of indoor location computing are not yet fully solved. They still present serious challenges and involve phenomena not yet fully understood. There is indeed a considerable distance toward a fully scalable and affordable system that is readily deployable using existing technologies (such as IEEE 802.11), and is able to provide accurate location information for indoor mobile computing. This project is developing realistic radio propagation models and adaptive signal location maps particularly adapted to different building environments leading to robust and competitive techniques for location determination. The research team is investigating an innovative direction to develop a benchmark standard in order to capture and classify critical environmental and system factors and to provide performance bounds as well as reproducible test beds for indoor research. The key of the research is to quantify radio signal measurement inaccuracy under various environments. This is achieved by measuring packets during the remodeling process of the second floor of the Science and Technology building, the new construction of the Student Campus Center at USCB, and by simulating various partitions between the transmitter and the receiver. This project transforms indoor location determination systems from high cost, labor intensive, imprecise, and static technologies to an affordable, automated, accurate, and dynamic system. This project attacks theoretical and systems challenges and will enable practical deployment of the ARIADNE indoor system. Moreover, this exciting project will help attract, educate and retain talented undergraduate students and also motivate them for higher studies. RI: Small: Exploiting Geometric and Illumination Context in Indoor Scenes The research objective is to investigate methods to recover geometry and perform spatial reasoning in rooms. This project aims to recover the room space, illumination, and object layout from an image. Together, these elements capture the layout of the room walls, the location of objects in the image and the 3D space, and a lighting representation that allows illumination artifacts to be explained and rooms to be relit with inserted objects. The work takes an integrated approach, exploiting constraints within and between spatial representations. The project also aims to leverage knowledge of room geometry to better reason about surface utility, enabling advanced spatial analysis of indoor scenes.

This research unifies ideas from geometry, multiple view computer vision, shading, and statistics to recover complex spatial representations from single views. The work further aims to create tools for object insertion and removal and scene completion, allowing the average person to more easily create the photograph that she wants or an interior designer to quickly sketch a photorealistic prototype of a new concept. The recovered spatial information also enables mobile robots to find walkable paths through cluttered rooms and to understand how objects can be physically manipulated and placed, which is essential for assistive household robotics. Other anticipated applications include surveillance, security, and transportation safety. The project contributes to education through student projects, course development, and workshops and tutorials involving a broader audience. CSR: Small: An Efficient Framework for Real Time Collaborative Browsing Collaborative browsing, also known as co browsing, allows multiple users to access the same web pages in a simultaneous manner and collaboratively fulfill certain tasks. As a useful and attractive technique, co browsing has a wide range of important applications such as online training and customer supporting. Existing co browsing solutions must use either a specific collaborative platform, a modified web server, or a dedicated proxy to coordinate the browsing activities between web users. Besides, these solutions usually require users to install special software on their computers. These requirements heavily impede the wide use of collaborative browsing among Internet users. We propose to design and develop an efficient framework for Real time Collaborative Browsing. This framework is a pure browser based solution. It leverages the power of Ajax (Asynchronous JavaScript and XML) techniques and the extensibility of modern web browsers for performing co browsing. Our objective is to achieve real time collaboration among web users, without the involvement of any third party platforms, servers, or proxies. The proposed framework will enable fine grained co browsing on arbitrary web sites and web pages. The success of this project will provide significant research contributions to on line collaboration. Our research results will be disseminated to industry and academia through high quality publications and the development of prototypes. The educational aspect of this project focuses on integrating education and research activities, and enhancing the undergraduate and graduate system and networking curricula. III: Small: RIOT: Statistical Computing with Efficient, Transparent I/O Recent technological advances enable collection of massive amounts of
data in science, commerce, and society. These datasets bring us
closer than ever before to solving important problems such as decoding
human genomes and coping with climate changes. Meanwhile, the
exponential growth in data volume creates an urgent challenge. Many
existing analysis tools assume datasets fit in memory; when applied to
massive datasets, they become unacceptably slow because of excessive
disk input/output (I/O) operations.

Across application domains, much of advanced data analysis is done
with custom programming by statisticians. Progress has been hindered
by the lack of easy to use statistical computing environments that
support I/O efficient processing of large datasets. There have been
many approaches toward I/O efficiency, but none has gained traction
with statisticians because of issues ranging from efficiency to
usability. Disk based storage engines and I/O efficient function
libraries are only a partial solution, because many sources of
I/O inefficiency in programs remain at a higher, inter operation
level. Database systems seem to be a natural solution, with efficient
I/O and a declarative language (SQL) enabling high level
optimizations. However, much work in integrating databases and
statistical computing remains database centric, forcing statisticians
to learn unfamiliar languages and deal with their impedance mismatch
with host languages.

To make a practical impact on statistical computing, this project
postulates that a better approach is to make it transparent to users
how I/O efficiency is achieved. Transparency means no SQL, or any new
language to learn. Transparency means that existing code should run
without modification, and automatically gain I/O efficiency. The
project, nicknamed RIOT, aims at extending R a widely popular
open source statistical computing environment to transparently
provide efficient I/O. Achieving transparency is challenging; RIOT
does so with an end to end solution addressing issues on all fronts:
I/O efficient algorithms, pipelined execution, deferred evaluation,
I/O cost driven expression optimization, smart storage and
materialization, and seamless integration with an interpreted host
language.

RIOT integrates research and education, and continues the tradition of
involving undergraduates through REU and independent studies. As a
database researcher, the PI is committed to learning and drawing from
work from programming languages and high performance computing.
Findings from RIOT help create synergy and seed further collaboration
with these communities. To ensure practical impact on statistical
computing, RIOT has enlisted collaboration from statisticians and the
R core development team on developing, evaluating, and disseminating
RIOT.

Further information can be found at:
http://www.cs.duke.edu/dbgroup/Main/RIOT CSR: Small: Green Farms: Towards a Stable Energy Optimization Architecture for Data Centers This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
This project develops techniques for increasing energy efficiency of modern data centers with performance constraints. Both computing and cooling energy are considered. According to the U.S. Environmental Protection Agency, data centers in the United States incur an annual energy cost of approximately $4.5 billion, which is comparable to the total consumption of 5.8 million average US households. In the absence of intervention, data center consumption is expected to double in five years. Up to 80% of this projected energy expenditure is avoidable, which constitutes a prospective reduction in nationwide carbon dioxide emissions by up to 47 million metric tons (MMT) per year; an amount comparable to the annual carbon emissions footprint of all fuel combustion in a small nation. The observation motivating the energy efficiency solutions developed in this project is that the proliferation of individual energy saving mechanisms in server installations can lead to increasingly suboptimal overall energy management. The problem lies in performance composability or lack thereof; a challenge that arises because individual optimizations generally do not compose well when combined, leading to opportunities for improvement. A theory is developed to define preconditions of composability and a holistic distributed approach is investigated for coordinating energy saving decisions across multiple components that span both the computing and cooling subsystems. A graduate course on cyber physical systems is designed to cover the addressed challenges and solutions. Women and minority students are encouraged to benefit from these opportunities. TC: Small: Provably Anonymous Networking Through Secure Function Evaluation Anonymous communication in computer networks generally relies upon the filtering of traffic through a cascade of mixes. Clients select a number of proxy nodes that modify and obscure the origin of a message so as to make the detection of a relationship between source and destination extremely difficult. While this architecture is known to not be robust against a globally passive adversary, an ever growing body of literature has demonstrated that even moderately capable adversaries can link the communications between two parties using these networks. Accordingly, new techniques ensuring stronger levels of anonymity must be explored. This project will develop a new architecture for anonymity networks offering cryptographic guarantees of anonymity based on a foundation of Secure Function Evaluation. These Provably Anonymous Networks (PANs) protect the participants of communication from traffic analysis attacks by remaining unaware that the exchange of messages has occurred. However, realizing a more trustworthy architecture is not simply the result of haphazardly assembling the above components. Rather, this system must be carefully composed so as to avoid the leakage of information useful in the identification of a communication channel. The results of this project will not only be used to enhance graduate and undergraduate curriculum, but will also be used to develop a tool to assist members of the Carter Center administer observe elections without fear of eavesdropping. RI: Small: Semi Supervised Learning for Non Experts This project develops semi supervised machine learning algorithms that are practical, and at the same time guided by rigorous theory. In particular, the project is developing learning theory that quantifies when and to what extent the combination of labeled and unlabeled data is provably beneficial. Based on the theory, novel algorithms are being developed to address issues that currently hinder the wide adoption of semi supervised learning. The new algorithms will be able to guarantee that using unlabeled data is at least no worse, and often better, than supervised learning. The new algorithms will also be able to learn from unlimited amounts of supervised and unsupervised data as they arrive in real time, something humans can do but computers cannot so far.

This project has a number of broader impacts: (1) An open source software will be an enabling tool for new discoveries in science and technology, by making machine learning possible or better in situations where labeled data is scarce. Since the software specifically targets non machine learning experts, the impact is expected to be across the whole spectrum of science and technology that utilizes machine learning. (2) It advances our understanding of the learning process via new machine learning theory, which can be applied to both computers and humans. (3) The proposal contains projects ideally suited to engage students in computer science education and research. CIF:SMALL:COLLABORATIVE: Soft Logic Modeling and Design for Synthetic Biology This project is a multi institutional collaboration in which we plan to investigate a new soft logic approach for creating ?designer gene? circuits. Genetic circuits are crafted in the laboratory using genomic building blocks, and are used to control specific behaviors in engineered bacteria. A major research challenge for genetic circuits is that there is a high level of randomness in the cell?s internal environment. Belief networks provide a well defined solution for handling the effects of randomness in genetic systems. We propose to study how belief networks can be applied to control and even exploit randomness to achieve new and useful genetic behaviors. Our research results will be used to improve a software application for genetic design called iBioSim.

There are many anticipated benefits for synthetic genetic circuits, including industrial, environmental, and medical applications. For example, bacteria can theoretically be engineered to clean oil spills, kill tumors, and deliver medicines, but only if we can precisely control when and how the bacteria perform their functions. Our investigation will illuminate the unique challenges involved in controlling highly random bacterial systems and will provide the community with rigorous theory and practicable techniques that resolve these challenges. III: Small: Ivory A Hadoop Toolkit for Distributed Text Retrieval Text search is a technology that is vital for modern information based societies. Today s systems face the daunting challenge of handling quantities of text previously unimaginable. Cluster computing is the only practical solution for addressing the issue of scale. This project leverages the MapReduce framework (via the open source Hadoop implementation) to tackle issues of robustness and scalability in processing large amounts of data for information retrieval applications. More generally, the goals are to explore the relationship between processor, disk, memory, and network in large distributed computing environments, where many assumptions made in single machines no longer hold. One pertinent example is the fundamental mismatch between Hadoop and the demands of real time interactive applications. Because it was designed for throughput oriented batch processing, Hadoop currently does not provide low latency disk access necessary for real time search. A distributed in memory object caching architecture provides a potential solution to this problem.

To achieve broader impact, the results of this research will be implemented in Ivory, an open source toolkit for distributed information retrieval built from the ground up with cluster architectures in mind. The availability of this toolkit will help sustain activities in the emerging area of cloud computing . Additional information is available on the project website (http://www.umiacs.umd.edu/~jimmylin/cloud computing). NeTS:Small:Algorithmic Approaches to Optimizing Hardware Based Packet Classification Systems via Equivalent Transformation This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
Using Ternary Content Addressable Memories (TCAMs) to perform high speed packet classification has become the de facto standard in industry. Despite their high speed, TCAMs have limitations of high cost, small capacity, large power consumption, and high heat generation. The well known range expansion problem in converting range rules to ternary rules significantly exacerbates these TCAM limitations. This project addresses TCAM limitations by developing new algorithms to transform a given rule set into an equivalent rule set that requires significantly fewer TCAM entries. This allows the use of smaller, faster, and more energy efficient TCAM chips. The algorithms developed by this project significantly outperform prior art because these new algorithms perform equivalent transformation at the list level whereas prior approaches only perform compression at the individual rule level.
Expected results of this project include effective equivalent transformation algorithms and potentially transformative concepts. Research results are broadly disseminated through publications, open source software releases, freely available course modules, and industry interaction. This project benefits society by decreasing the demand of modern routers for large TCAMs, lowering router prices and energy cost, enabling the use of small and cheap TCAMs on low end routers, and extending router life time. The technologies developed in this project greatly benefit the business of router manufacturers, TCAM chip providers, and Internet service providers. To promote education and learning, this effort actively engages high school, undergraduate, and graduate students, especially students from under represented minorities. RI: Small: Collaborative Research: Word Sense and Multilingual Subjectivity Analysis Approaches to subjectivity and sentiment analysis often rely on
manually or automatically constructed lexicons. Most such lexicons are
compiled as lists of words, rather than word meanings ( senses ).
However, many words have both subjective and objective senses as well
as senses of different polarities, which is a major source of
ambiguity in subjectivity and sentiment analysis. The proposed work
addresses this gap, by investigating novel methods for subjectivity
sense labeling, and exploiting the results in sense aware subjectivity
and sentiment analysis. To achieve these goals, three research
objectives are targeted. The first is developing methods for assigning
subjectivity labels to word senses in a taxonomy. The second is
developing contextual subjectivity disambiguation techniques to
effectively make use of the word sense subjectivity annotations. The
third is applying these techniques to multiple languages, including
languages with fewer resources than English. The project will have
broader impacts in both research and education. First, it will make
subjectivity and sentiment resources and tools more widely available,
in multiple languages, to the research community, which will help
advance the state of the art in automatic subjectivity analysis, which
in turn will benefit end applications. Second, several educational
goals will be pursued: training graduate and undergraduate students in
computational linguistics; augmenting artificial intelligence courses
with projects based on the proposed research, which will offer
students hands on experience with natural language processing
research; and reaching out to women and minorities to increase their
exposure to text processing technologies and access to research
opportunities. TC: Small: Characterizing and Mitigating Device Based Attacks in Cellular Telecommunications Networks TC: Small: Characterizing and Mitigating Device Based Attacks in Cellular Telecommunications Networks

PI: Patrick Traynor (Traynor@cc.gatech.edu)
Co PI: Jonathon Giffin (giffin@cc.gatech.edu)


Mobile phones have traditionally provided a limited set of telephony operations. However, the recent and rapid introduction of complex operating systems, sophisticated user interfaces and connectivity with the Internet has transformed these systems into highly capable general purpose computing platforms. Unfortunately, the software implementing these systems often lacks even basic security protections. In this project, we will investigate the impact of an immature and vulnerable software infrastructure on both mobile devices and the cellular network core. Characterizing and responding to the threats is critical, especially given that the security models of cellular networks are designed around the assumption of highly limited user devices. We plan first to create Caegis, an in situ set of tools for ARM code analysis and over the air cellular phone fuzzing to help manufacturers improve the quality of their software and reduce its susceptibility to malware. We will then focus on basic mechanisms for Remote Repair, which will allow the network to help a device reboot to a known safe state. The development of these two tools, in combination with improved characterizations of the impact of malware, will significantly improve the security provided to this piece of critical infrastructure. Moreover, this project will help students in both independent research and a related class to gain hands on experience with security research on our Alcatel Lucent IMS network core. TC: Small: Discovering Designer Intent through Dynamic Analysis of Malware The proposed research addresses the problem of identifying and providing source code based understanding of obfuscated malcode plaguing the Internet. Obfuscated malcode presents a clear and present danger to today?s society in terms of individual privacy, security as well as to the Internet?s overall trustworthiness. Attackers continue to use obfuscation to successfully defeat attempts by defenders to prevent infection or spread of the malcode.
The goal of the research will be to develop dynamic binary analysis processes for Internet worm executables. These processes will be used to create a reverse engineering framework to create assembly code from worm machine code and in turn create associated control and data flow graphs. Generalizations of instruction sequences, known as motifs, are extracted from these graphs using new techniques based in part on program slicing and will be applied against models of worm behavior described in terms of state machine, decision tree or family tree models; partial or complete matching against these models will yield knowledge of worm interactions with its target environment, its obfuscation techniques and its means of command, control and update by the attacker.
This research will help to assess if control and data flow graphs can represent any obfuscated malcode. In case some malcode cannot be represented using these graphs, other representations will be investigated. The accuracy of the malcode representation will be evaluated since the representation will be based mainly on system call analysis. Finally, several combinations of static and dynamic analyses will be studied to assess the impact on the malcode representation details. CIF:Small:Physical Layer Optimization for Cognitive Sensor Networks Physical Layer Optimization for Cognitive Sensor Networks

The cognitive radio concept has been a revolutionary development in wireless communications systems. Cognitive, software defined radios are able to adjust link and network resources in order to optimize communications performance. However, high rate, robust communications is often just one of many possible network objectives. For example, in sensing applications, the goal is to maximize coverage, detect important events with high probability, and track objects of interest with high accuracy. These goals are often at odds with those for optimum communications; improved coverage requires more widely dispersed sensors, complicating network connectivity. High resolution sensing requires more bits of information, which in turn place a strain on network throughput. Power devoted to routing or packet forwarding reduces a sensors lifetime. Clearly, a different paradigm is needed when sensing performance is the critical factor, or perhaps most interestingly, when both communications and sensing performance must be considered in tandem. This research effort introduces Cognitive Sensing as a concept dual to that of cognitive communications, and investigates the competing objectives of sensing and communications networks. A cognitive sensor would adaptively adjust its operating parameters in response to the environment it finds itself in so as to optimize sensing performance, or perhaps a dual performance metric that includes both sensing and communications functions. Parameters relevant to sensing performance could include sensor position, speed and heading, antenna beampattern and polarization, transmit waveform type and bandwidth, imaging camera zoom, orientation, resolution, pointing angle, etc. The question of how to allocate such sensor resources is central to this effort. SHF: Small: RUI: Making Sense of Source Code: Improving Software through Information Retrieval The cost effective construction of software is increasingly important to businesses and consumers. Given software s ever increasing size and complexity, modern software construction employs significant tool support. Recent tools complement traditional static analysis tools by exploiting the natural language found within a program s text through the use of Information Retrieval (IR). Best known for its use by search engines on the Internet, IR encompasses a growing collection of techniques that apply to large repositories of natural language. New tools using IR have tackled problems previously requiring considerable human effort. However, to reap the full benefit of IR techniques, the language across all software artifacts (e.g., requirement and design documents, test plans, as well as source code) must be normalized. Normalization align the vocabulary found in source code with that of other software artifacts. In addition to improving existing tools, normalization will also encourage the development of new techniques and methodologies useful in future tools. Empirical study of successful tool improvements will aid technology transfer of the tools expected to improve programmer productivity. Beyond its technical goals, this research promotes discovery in Loyola s undergraduate curriculum through the direct involvement of undergraduate students in scientific research and by integrating research results into classroom learning. SHF: Small: A Generic Micro Architecture for Accuracy Aware Ultra Low Power Multimedia Processing Mobile devices supporting a wide variety of multimedia applications under a very stringent energy budget are a key driver of electronic systems in future. The objective of the proposed research is to explore a generic, programmable, reconfigurable, and energy efficient architectural platform for future mobile devices for real time multimedia processing. The aim is to exploit the inherent error tolerance in multimedia applications for run time energy accuracy trade off. The proposed research will analyze the interaction of ultra low power computing and error characteristics of real time multimedia processing. This research will pursue a circuit architecture algorithm co design approach to model, analyze, and demonstrate a reconfigurable hardware platform for memories, datapath, and buses to exploit the error characteristics of real time multimedia processing algorithms for ultra low power.

This generic architecture for energy efficient multimedia systems can lay the foundation of future mobile supercomputers performing wide array of applications with minimal energy. The PIs will disseminate the research results through project website, conference and journal publications. The existing interactions with leading microprocessor and mobile handset manufacturers will provide opportunities for technology transfer. The educational goal is to create next generation engineers who understand the effects of energy and accuracy on computations. The PIs will engage in recruiting students from underrepresented groups and mentor students under the Summer Undergraduate Research Experience for minorities (SURE) program at Georgia Tech. III: Small: Optimizing News and Opinion Aggregators for Diversity Observers have raised alarms about increasing political polarization of our society, with opposing groups unable to engage in civil dialogue to find common ground or solutions. Aggregators such as Digg, Reddit, and Google News rely on ratings and links to select and present subsets of the large quantity of news and opinion items generated each day. If a majority of the raters or linkers share a political viewpoint, minority viewpoints may get little representation in the results, creating an echo chamber for the majority. Even if a site selects items based on votes or links from people with diverse views, algorithms based solely on popularity may lead to a tyranny of the majority that effectively suppresses minority viewpoints. This work is the first attempt to formalize several different instances of the general concept of diversity of viewpoints and to devise algorithms that optimize for these measures. The techniques are likely to be applicable to other domains where selecting a diverse set of items is valuable, such as search engine results and audience voting on questions to ask of a conference speaker or public official. The goals of this research are to: 1) form alternative measures of diversity for result sets; 2) develop algorithms for selecting result sets that jointly optimize for diversity and popularity; 3) assess the impacts of alternative selection and presentation methods on people s willingness to use an aggregation service, their exposure to diverse opinions, and the size of their argument repertoires.

The results of the project will provide a better understanding of alternative notions of what it means for a set of items to be diverse or balanced, and the range of reactions that different people have to varying levels and presentations of diversity. Insight into people s preferences for acceptable support and challenge may also allow for the creation of news and opinion aggregators that cause people to choose to expose themselves to greater diversity, thus reducing polarization and enhancing democracy. Results, including open source software, will be distributed via the project web site: (http://si.umich.edu/balance/). III: Small: Computational Infrastructure for the Identification of Copy Number Variations from SNP Microarrays It was recently discovered that copy number variations (CNVs) in human genome are quite common, and have important implications on phenotype. Currently, the primary platforms for large scale detection and characterization of CNVs are SNP (single nucleotide polymorphism) microarrays. The current state of the art in computational identification of CNVs from microarray data relies mostly on model based approaches (e.g., Hidden Markov Models). However, such methods
require extensive training data, which may not be always available. Furthermore, since these methods use common CNVs to train their models, they are not as successful in identifying rare CNVs, which are believed to make up a substantial proportion of all CNVs in the human population. The objective
of this project is to develop optimization based algorithms and software for the identification and genotyping of CNVs, with a view to enabling fast and accurate identification of different types of CNVs (rare and common), without the requirement of training data.

The proposed framework develops a novel computational approach by explicitly formulating CNV identification as a series of optimization problems that incorporate multiple factors, including sensitivity to noise, rarity/commonality of CNVs, genotypic specificity, and parsimony.
This formulation enables development of efficient algorithms that treat identification of rare and common CNVs as different problems with different objective functions. Availability of the resulting software to the community will enable more efficient and accurate identification of CNVs in large
samples, facilitating advances in understanding the role of CNVs in a range of complex phenotypes, including HIV, autism, schizophrenia, mental retardation, and many others. Furthermore, the computational innovations introduced by this project are likely to find applications in next generation sequencing. NeTS: Small: Predictable Optimization of Opportunistic Communication This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Opportunistic communication leverages communication opportunities arising by chance to provide significant performance benefit or even enable communication where it would be impossible otherwise. This project develops algorithms, techniques, and protocols that optimize opportunistic communication to achieve good, predictable performance in wireless mesh networks and vehicular networks.

A key challenge involved is how to systematically optimize opportunistic communication to achieve good predictable wireless performance. The research addresses the challenge by making three major contributions. First, the PIs develop novel robust optimization techniques that systematically optimize opportunistic communication even in the presence of high uncertainty. Second, they design efficient models, measurement and inference techniques, and prediction algorithms to obtain the required inputs to the optimization algorithms. Third, they exploit inter flow coding opportunities arising from multi flow diversity to further enhance the efficiency of opportunistic communication. To demonstrate the effectiveness of the approaches, the PIs design, implement, and experiment in a wireless mesh network and a vehicular network testbed deployed at UT Austin.

The project produces publications in leading network conferences and journals, and software that is publicly available online. The resulting techniques and software significantly advance wireless mesh network and vehicular network technology, and benefit other wireless network environments. The research results are also integrated into undergraduate and graduate curricula as well as outreach activities. TC: Small:Automata Based String Analysis for Detecting Vulnerabilities in Web Applications Web applications contain numerous vulnerabilities that can be exploited
by attackers to gain unauthorized access to confidential information and
to manipulate sensitive data. Many of these vulnerabilities are due to
inadequate manipulation of string variables. String analysis, a technique
that captures the string values that a certain variable might hold at a
particular program point, can be used to identify such flaws. In this
project, novel and precise string analysis techniques will be developed
using an automata based approach that represents possible values of a
string variable at a program point as an automaton. Techniques that
support path sensitivity and that enable precise analysis of loops using
automata based widening operations will be developed. Basic string analysis
techniques will be extended to a composite analysis where relationships among
string variables and other types of variables can be automatically discovered
and analyzed. The precision of string analysis plays a central role for
obtaining good results with static vulnerability detection tools. The
precise string analysis techniques developed in this project will enable
analysis of programs that cannot be analyzed with existing techniques. The
results of these improved string analysis techniques will lead to novel
software security solutions and detection of novel types of vulnerabilities. RI: Small: Enhancing Nonmonotonic Declarative Knowledge Representation and Reasoning by Merging Answer Set Programming with Other Computing Paradigms Answer Set Programming (ASP) is a recent form of declarative programming that has been applied to many knowledge intensive tasks, such as product configuration, planning, diagnosis, and information integration. Like other computing paradigms, such as SAT (Satisfiability Checking) and CP (Constraint Programming), ASP provides a common basis for formalizing and solving various problems, but is distinct from others in that it focuses on knowledge representation and has proved to be useful for rapid prototyping. While the research on ASP has produced many promising results, it has also identified serious limitations.

The project aims at overcoming the limitations by merging ASP with other computing paradigms, such as satisfiability checking, first order logic and constraint programming, and exploring the synergy between them. This project is expected to provide a transformative understanding of ASP s relation to other computing paradigms, to enhance ASP s reasoning capability and broaden the areas in which it is effective. Within knowledge representation, the study will clarify the role of ASP as a major knowledge representation formalism with effective computation methods that combines various methods available in other computing paradigms. CSR: Small: Collaborative Research: Adaptive Applications and Architectures for Variation Tolerant Systems CSR: Small: Collaborative Research: Adaptive Applications and Architectures for Variation Tolerant Systems

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The scaling of integrated circuits (ICs) into the nanometer regime has thrown up new challenges for designers, foremost among which are variations in the characteristics of IC components. Variations threaten to diminish the fundamental benefits of technology scaling, such as improvements in cost per transistor, performance and power consumption. Variation aware design techniques that have been proposed thus far are being stretched to their limits, and cannot contain the incessant increase in variations. Therefore, it is important to develop new design approaches for systems that are inherently resilient to variations in the underlying components.

This project develops a framework based on adaptive applications and architectures for the design of variation tolerant application specific systems. It advances the state of the art by (i) adopting a cross layer approach at the system architecture and application layers, (ii) leveraging the inherent ?elasticity? of a wide class of applications to adapt to variations in the underlying hardware while still producing acceptable performance and maintaining end user experience, and (iii) exploring a hybrid (design time and post fabrication) design methodology, enabling more accurate and effective system adaptation in response to variations. The developed technologies will significantly extend our ability to avail of the benefits of technology scaling in the face of increasing variations.

The efforts towards broader impact include working with the semiconductor industry to validate and transfer the developed technologies, new educational material incorporated in courses on SoC design and embedded systems, and undergraduate design projects. TC: Small: Capsule: Safely Accessing Confidential Data in a Low Integrity Environment Protecting confidential information is a major concern for organizations and individuals alike, who stand to suffer huge losses if private data falls into the wrong hands. One of the primary threats to confidentiality is malicious software, which is estimated to reside on millions of computers. Current security solutions, such as firewalls, anti virus software, and intrusion detection systems, are inadequate at preventing malware infection.

The main contribution of this project will be a novel mechanism, called Storage Capsules, that will allow users to protect confidential files on personal computers and servers. Storage Capsules are encrypted containers that will allow users on a compromised machine to securely read and write files on the container using standard applications without malware being able to steal confidential file data. Research themes include designing the system, analyzing its security and addressing the potential threats, improving its efficiency, and extending it to support user defined access policies. Broader implications of this work include providing a new method for people to secure sensitive data on their personal computers and servers, even in the presence of malware. SHF: Small: Reducing the Cost of Computation in CMPs This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Computer processor industry has moved fully into the multi core era to enable continual scaling of performance, but at the cost of increased energy consumption and increased cooling costs due to higher temperatures and thermal gradients. This proposal describes three major research thrusts that address these costs in multiple ways: (1) New modeling and simulation tools: We will integrate performance, power, temperature, reliability and cooling estimation, so that the designers will be able to analyze the impact of design choices and runtime decisions over significant time spans. (2) Runtime thread scheduling policies: We will identify and demonstrate power and thermal scheduling mechanisms that maintain performance, reduce the total energy consumption, and eliminate or reduce hot spots, but also maximize processor lifetime. The policies will use the data from thermal sensors and performance counters to proactively drive the management decisions. (3) New cooling strategies: Our goal is to create thermal management and cooling control algorithms that work in tandem to reduce the overall energy consumption.

The proposed research forms the basis for discovery and learning in the areas of multi core processors, and, more generally, system design and management. Graduate and undergraduate students will be involved in various parts of the proposed research and help in connecting this work with other NSF sponsored projects. The results of research, tools and coursework materials developed will be freely and easily distributed to engineering community at large. AF:Small: Computation in Very Large Groups Mathematical group theory has wide applications to the sciences and to other branches of mathematics. Some important examples of current interest are fast matrix multiplication, search in the presence of symmetry, and symmetries to be found in physics and chemistry. Current algorithms do not always scale or are not always practical in implementation. Some levels beyond which current algorithms tend to become impractical are permutations of a million points, matrix group dimensions beyond a few tens, and coset methods (defining equations on groups) beyond 100 million cosets. We call such groups very large groups .

This project will develop a new class of algorithms for very large groups. The new algorithms will take advantage of the experience of the P.I. and his lab in previous computations and algorithms using terabytes of parallel disk storage. The feasibility of a many disk approach had previously been shown in a popular demonstration concerning Rubik s cube: Rubik s cube can be solved in 26 moves or less. Both that and more traditional problems will be used to further develop the disk based language, Roomy.

Emphasis will be given to well known problems not known to be in polynomial time (centralizer, group intersection, normalizer, etc.). These problems have seen little progress during the last decade. They are considered hard in part due to their close connection with the conjugate group action of a group on itself. In this conjugate action view, a group is seen as a permutation group acting on almost as many points as there are elements in the group itself. In this view, even moderate size groups quickly turn into very large groups under the conjugate action. Novel methods such as the biased tadpole, coupled with the power of the Roomy language, will enable a resumption of progress in this area.

The broader impact lies in the ability to harness these new algorithms and implementations in pursuit of applications outside of group theory such as those mentioned earlier. Researchers outside of group theory have long had the potential to generate groups beyond the capabilities of standard software, such as the free and open source GAP package. Extending the capabilities of GAP and other familiar tools will enable new discoveries. The further development of the Roomy platform is also an important byproduct, whose value will extend far beyond its group theory origins. III: Small: Collaborative Research: Creating and Evolving Software via Searching, Selecting and Synthesizing Relevant Source Code Software developers rely on reusing source code snippets from existing libraries or applications to develop software features on time and within budget. The reality is such that most previously implemented features are embedded in billions of lines of scattered source code. State of the art code search engines provide no guarantee that retrieved code snippets implement these features. Even if relevant code fragments are located, developers face rather complex task of selecting and moving these fragments into their applications. Finally, synthesizing new functionality by composing selected code fragments requires sophisticated reasoning about the behavior of these fragments and the consequent code. The result of this process is an overwhelming complexity, a steep learning curve, and a significant cost of building customized software.

This research program proposes an integrated model for addressing fundamental problems of searching, selecting, and synthesizing (S3) source code. The S3 model relies on integrating program analysis and information retrieval to produce transformative models to automatically search, select, and synthesize relevant source code fragments. The S3 model will directly support new methodologies for software change and automated tools that assist programmers with various development, reuse and maintenance activities. Among the broader impacts the project includes collaboration with industry to transfer technology. RI Small: Statistical Decoding Models to Improve the Performance of Motor Cortical Brain Machine Interfaces The goal of this project is to develop statistical models to accurately and efficiently decode population neuronal activity in the motor and premotor cortex. The study focuses on motor behavior as it is easily measured and strongly correlated with neuronal activity. Recent advances in motor cortical brain machine interfaces have shown that research animals and paralyzed human patients were able to perform rudimentary actions with external devices such as robotic limbs and computer cursors. Neural decoding, which provides control commands to external devices, plays a key role in such interfaces by converting brain signals (e.g., spiking rates of a population of neurons) to kinematic states (e.g., hand position, hand movement direction).

Current decoding models are often based on the strong assumption that the neural signal sequence is a stationary process. This assumption, however, does not take into account the significant dynamic variability of spiking activity over time. Moreover, these methods have either focused on decoding the entire trajectory or on the occurrence times of a few landmarks during the movement. Effective coupling of these two complementary strategies can be expected to improve the decoding performance by better exploiting the nature of the landmark defined movement. This project will develop computational methods to address these two issues. For the non stationarity, the research team will develop adaptive versions of state of the art decoding methods such as particle filters and point process filters that can capture the varying patterns in neural signals and update the model accordingly. To couple trajectory decoding and time decoding, landmark times will be identified from the neural activity, and then incorporated into the kinematic model. The team will use simultaneous recordings from multi electrode arrays in the primary motor cortex, the dorsal premotor cortex, and the ventral premotor cortex that were recorded during behavior or visuo motor tasks. Improved decoding methods are expected to have significant impacts on neural prosthetics. SHF: Small: Nanocomputing Processes and Artifacts: Fundamental Description and Physical Information Theoretic Assessment As silicon integrated circuit technology approaches its ultimate scaling and performance limits, we can expect a rapid proliferation of innovative proposals for fundamentally new information processing technologies. The quest for the first post CMOS general purpose computing machines will likely emphasize digital computation in systems constructed from nanoscale building blocks. Proposals for new nanocomputing technologies will, however, be dificult to evaluate, both because nanocircuits are dificult to build and test experimentally and because phenomena that compromise the reliable physical representation and manipulation of information in nanoscale systems will pose new and unfamiliar challenges. These considerations motivate the development of new theoretical tools for assessing the fundamental physical limits to reliable processing of classical information in nanoscale systems, limits that follow from generic space, time and power constraints imposed by the technological objective of superseding silicon technology at the end of scaling.

This project aims to advance the fundamental physical description of digital information processing in (generally noisy and faulty) nanosystems and to develop approaches, built from such a description, that can be used to evaluate the ultimate information processing capabilities of proposed nanocomputing technologies. The first prototype assessment studies will emphasize two existing proposals quantum dot cellular automata and nanowire based NASIC fabric implementations and will integrate results from physical information theoretic analyses and physical circuit models. Other explorations will aim to provide technology independent insights into issues of generic importance for nanocomputation, such as the physical costs of error correction. These investigations, taken together, will help to clarify the nature of fundamental physical limits in information processing and their practical consequences, which will become increasingly important as the quest
for new nanocomputing technologies intensifies. AF:Small:Pseudorandomness, Codes, and Distributed Computing Randomness is useful in many areas of computer science, including algorithms, Monte Carlo simulations, cryptography, and distributed computing. In practice, however, it is expensive or impossible to get truly random numbers. Therefore, computers rely on pseudorandom generators. However, scientists have reported problems with practical pseudorandom generators. Can we construct pseudorandom generators that are provably good? The PI proposes to address this question and the related question of how to extract high quality randomness from low quality random sources.

These questions have unexpected connections to error correcting codes and distributed computing, which the PI proposes to explore further. He also proposes to attack fundamental questions in these areas. In coding theory, these questions relate to his recent results on decoding the important Reed Muller codes. In distributed computing, he proposes to advance his work on network extractor protocols. These are protocols to extract high quality randomness from low quality random sources in a distributed setting. Such protocols could be very useful in cryptography. RI: Small: Universal Automated Reasoning by Knowledge Compilation A long term goal motivating this project is the development of a universal computational engine to support reasoning across different applications of intelligent systems (e.g., planning and diagnosis). The project is founded on an approach that compiles knowledge bases into a taxonomy of tractable forms, which result from imposing various conditions, such as decomposability and determinism, on Negation Normal Form (NNF). Certain queries, which are generally intractable, become tractable on the compiled NNF forms. To implement a task like planning or diagnosis, all one needs to do is compile their knowledge base to the most succinct subset of NNF that provides polytime support for the queries required by the task. The project will focus in particular on algorithms for imposing various conditions on NNF compilations, using both top down and bottom up compilation techniques, to support a larger set of tractable forms. This will lead to developing and evaluating a more powerful inference engine than traditional SAT solvers, supported by a more comprehensive set of queries and transformations on knowledge bases. It will also lead to extending the compilation approach to a larger class of AI applications. NeTS:Small:Supporting Multi Missions in Wireless Sensor Networks This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Wireless sensor networks have been applied to many military and commercial applications. However, the sensor network envisioned so far is targeted for a single mission and often designed for some particular application. As sensors become widely deployed, multiple missions, each with different requirements, may share common sensors to achieve their goals. Each mission may have its own requirements for the type of data being reported, the sampling rate, accuracy, and location of the sampling. From resource management point of view, it will be cost effective for sensor networks to support multiple missions instead of a single mission. The specific goal of this project is to support multi missions in wireless sensor networks.

The project addresses four intertwined issues: (i) various mission driven scheduling protocols which can optimize the sensor coverage will be designed, implemented, and evaluated; (ii) novel techniques will be developed to disseminate the mission switch code/command to the affected sensor nodes quickly and efficiently; (iii) mission driven sensor assignment schemes will be designed to maximize the network utility; (iv) mission specific network configurations will be investigated to meet the real time requirements of data transfer for dynamic and competing missions. This project will make significant theoretical advances in understanding and designing multi mission oriented sensor networks, and will develop comprehensive protocols to support multi missions in sensor networks. The success of this project is likely to have a broader impact on making wireless sensor networks more affordable and amenable to commercial, civilian, and military applications. The results of the project will be disseminated widely through high quality publications and presentations. The proposed research will also be integrated with the education curricula at Penn State University. AF: Small: Ground State Complexity in Quantum Many Body Systems One of the central goals of quantum information theory is to understand quantum systems from the standpoint of computational complexity. Physicists have been using computers for decades to understand various aspects of quantum systems, but these methods are typically heuristic and achieve success on only limited classes of systems. Understanding quantum systems through the lens of computational and information theoretic complexity has already lead to new powerful computational methods in physics and deeper insight into what causes these methods to fail once we step outside specific classes. Meanwhile hardness results often have important implications for quantum computation. If computing a property of a system is shown to be as difficult as computing the output of an arbitrary quantum circuit, that system becomes a candidate for a quantum computer.

This proposal focuses specifically on the complexity of ground states, the lowest energy state of a system. How hard is it to compute the ground energy of a quantum system? What are the properties of a system that give rise to provably efficient algorithms to compute the ground state? What is the structure of the ground state? Under what circumstances does the ground state have a succinct representation? The PI will pursue these questions in the context of one dimensional systems and will begin work on two dimensional systems about which much less is known. RI: Small: An Advanced Learning Paradigm: Learning Using Hidden Information Modern machine learning is limited in its ability to use diverse information during training. This project is developing algorithms in the SVM family that allow extra information to be used effectively during training, with the understanding that this extra information will not be available during actual operation. Examples of extra information include structural homologies between proteins in a system designed to predict structure from amino acid sequences; and values for a financial time series between the time where a prediction is made and the time of the value being predicted. Preliminary testing has shown that such extra information can dramatically reduce prediction error in the learned system compared with current generation machine learning methods that cannot use this extra information.

This project encompasses analytic research to establish performance bounds on our new algorithms, and to explore the relationships of this work to human learning. The project also includes experimental work, including construction of novel training and testing datasets; software implementation of the algorithms; and training, testing and analysis of experimental results. Areas of application include handwritten character recognition; 3 D protein structure prediction; non linear time series prediction, for example of financial time series; and prediction of likelihood of hospital readmittance for elderly patients. This project aims to give greater insight into the nature of learning, whether in humans or machines, and seeks to formally take into account data that is today seen as only peripheral to the learning task, and impossible for current machine learning algorithms to use.

The project will produce technical articles, a book, and teaching materials explaining this research. In addition the project will produce sharable software that implements the best version of the algorithm devised during the life of the project. NeTS:Small:Content Retrieval Networks: Network Access for Disadvantaged Regions This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).

Even in the US, many regions suffer poor Internet connectivity, due to
high prices, capacity limitations, poor infrastructure, or low
population density. Rural communities may be underserved because of
DSL distance limits and the high cost of bringing fiber optic cable to
low density regions. Even when businesses or schools can be connected
by DSL, its capacity limits may still provide poor end user experience
when many people share the connection. This research is addressing these
problems by using intelligent software to enhance network
connectivity. By aggressive data caching and off peak prefetching,
data can be stored on disk and retrieved locally instead of always
being fetched from the Internet. For environments with many users,
this system can reduce bandwidth consumption by an order of magnitude,
providing higher effective end user bandwidth, or allow cost
reductions by purchasing less wide area bandwidth. The expected
outcome of this research is an open source system that can be deployed
at Internet gateways, which will transparently improve network
performance. Possible locations include schools, small businesses, and
Internet Service Providers (ISPs). In schools, when a classroom full
of students accesses a topic, only one copy is downloaded, and
the rest get it from the cache. For news sites that update during the
day, only the changed portions get downloaded, instead of whole
articles. Prefetching will also reduce latency, by determining user
patterns and preloading content before it is needed. AF: Small: Mathematical Programming Methods in Approximation Mathematical programming techniques like linear programming and semidefinite programming have firmly established themselves as valuable tools in the approximation algorithm design toolkit. They are often used as tractable relaxations to hard combinatorial optimization problems and as design guides to obtain approximation algorithms. This project attempts to enhance our understanding of the strengths and weaknesses of mathematical programming techniques for several fundamental optimization problems and proposes to investigate general methods to strengthen our currently known mathematical programming techniques.

The broad goals of this project are the following: (a) Attempt to devise better algorithms for unique games ? a constraint satisfaction problem that is known to capture the limitations of current semidefinite programming methods used in approximation algorithms. Beating the current best algorithms will necessarily involve developing new techniques that overcome the limitations of current SDP approaches. (b) Understand the structure of strengthened relaxations obtained by lift and project procedures and develop techniques to exploit the additional information provided by these stronger relaxations to obtain better approximation algorithms. (c) Work towards closing large gaps in our understanding of the approximability of fundamental optimization problems.

Successfully achieving the project goals will entail significant advances in the state of the art for approximation algorithms. The research could potentially develop tools and techniques with broad applicability to several optimization problems. Course materials for graduate and undergraduate courses will be developed distilling research results of this project, as well as new developments in the field. TC:Small: Towards Trustworthy Intrusion Monitoring Using Wireless Sensor Networks Intrusion monitoring using networked sensors has a broad range of applications, including border security, surveillance, and monitoring of critical infrastructure such as nuclear power plants. In these hostile environments, the sensor network will itself be an attractive target that well funded attackers will attempt to undermine. The sensor network has to be protected so that no intruders can evade monitoring, a challenging issue with many aspects not adequately addressed by existing research.

This project investigates novel techniques for location privacy and jamming resistant tracking to defeat smart and resourceful attackers in intrusion monitoring applications. Location privacy techniques hide the locations of critical infrastructure such as base stations to make it hard for an adversary to locate and attack them. Jamming resistant tracking allows the system to detect and track intruders in the presence of jamming attacks. However, existing location privacy techniques do not work effectively against resourceful adversaries who can monitor all traffic at a large area, and existing jamming resistant algorithms do not scale well to large networks. These problems will be addressed in this project.

Through this project, we expect to uncover insights and develop algorithms that will apply to intrusion monitoring systems. By providing a layered, comprehensive defense system, the success of this project will have substantial impacts on both civilian and military operations where security is a major concern. This project will also help develop course materials on sensor network security and privacy. New course materials will enhance the information assurance curricula at UTA and other institutions. CSR: Small: Collaborative Research: System Support for Managing Carbon Footprints and Electricity Costs in Internet Services Society is faced today with three main energy related challenges: US dependence on foreign energy sources, world wide dependence on non renewable energy, and climate change induced at least in part by greenhouse gas emissions. The computer industry also faces an energy crisis: the nation s data centers consume a gigantic amount of energy, which translates into large greenhouse gas footprints. This research explores the implications of these trends on data center design. In particular, the work considers how to optimize data center operation in the face of Kyoto style cap and trade frameworks, related cap and pay frameworks, and unregulated scenarios where businesses optimize energy usage to save money or achieve carbon neutrality. In both cap and trade and cap and pay, caps are imposed on activities society wants to discourage. In large computer systems, excessive brown energy consumption is one activity to discourage. Through caps on brown energy, society can also promote renewable energy.
Research challenges include: (i) balancing reductions in brown energy consumption against cost, performance and service level agreement (SLA) impact, (ii) managing energy consumption in the context of variable electricity prices; and (iii) designing multiple system layers that effectively integrate and coordinate electricity and performance management, even in the face of highly volatile request distributions and electricity/carbon prices. This research has the potential for broad impact both on computer systems design, and more broadly on an increasingly carbon conscious world. III: Small: Accurate Protein Threading via Tree Decomposable Graph Modeling This protein threading project at University of Georgia develops high throughput computer programs for accurate protein tertiary structure prediction based on a new conformational graph modeling of protein amino acid tertiary interactions. The modeling approach enables the following three novel, effective components together to achieve the project goal: (1) a very efficient sequence structure alignment method that can incorporate sophisticated energy functions to improve fold recognition accuracy, (2) a simultaneous backbone prediction and side chain packing method to improve threading alignment accuracy, and (3) a semi threading method to improve the accuracy of new structure prediction. This project also enriches the interdisciplinary education programs at University of Georgia, allowing computer science students to implement protein threading programs and visualization tools, and bioinformatics and biology students to develop threading methods, test data, and evaluate and disseminate results.

For further information see the project web page at
URL: http://www.uga.edu/RNA Informatics/?p=projects CIF: Imperfection Resilient Scalable Digital Signal Processing Algorithms and Architectures Using Significance Driven Computation Present day integrated circuits are expected to deliver high quality/high performance levels under ever diminishing power budgets. Due to quadratic dependence of power on voltage, supply voltage scaling has been investigated as an effective method to reduce power. However, supply scaling increases the delays in all computation paths and can result in incorrect or incomplete computation of certain paths. Besides power dissipation, process variations also pose a major design concern with technology scaling. Supply voltage can be scaled up or logic gates can be up sized to prevent delay failures and to achieve higher parametric yield. However, such techniques come at the cost of increased power and/or die area. Meeting the contradictory requirements of high yield, low power and high quality are becoming exceedingly challenging in nanometer designs. Hence, there is a need for a scalable design methodology in which minimal output quality degradation is achieved under changing power constraints and process conditions. In addition, for a prescribed power consumption level and process, design methodology must take into account the effects of input signal noise and distortion on the fidelity of the Digital Signal Processing (DSP) computation and ensure that graceful output quality degradation is achieved under varying degrees of noise and distortion through proper algorithm and hardware design.



The research involves development of a systematic methodology for reorganizing (transforming) algorithmic level computations, data and underlying hardware in such a way that minimum performance degradation in DSP systems is achieved under reduced power supply, increased process variations and reduced input signal quality. It has been observed that for DSP applications/systems, all computations are not equally important in shaping the output response. This information is exploited by the investigators to develop suitable algorithms/architectures that provide the ?right? trade offs between output quality vs. energy consumption (supply scaling) vs. parametric yield due to process variations vs. input signal noise. To address resilience to process variations, the investigators identify the significant/not so significant components of such systems based on output sensitivities. Under such a scenario, with scaled supply voltage and/or parameter variations, if there are potential delay failures in some paths, only the less significant computations are affected. In other words, using carefully designed algorithms and architectures, the investigators provide unequal error protection (under voltage over scaling) to significant/not so significant computation elements, thereby achieving large improvements in power dissipation with graceful degradation in output signal quality. RI: Small: Spacetime Reconstruction of Dynamic Scenes from Moving Cameras The proliferation of camera enabled consumer items, like cellular phones, wearable computers, and domestic robots, has introduced moving cameras, in staggering numbers, into everyday life. These cameras record our social environment, where people engage in different activities and objects like vehicles or bicycles are in motion. State of the art structure from motion algorithms cannot reliably reconstruct these types of scenes. The overarching focus of this work is to develop the theory and practice required to robustly reconstruct a dynamic scene from one moving camera or simultaneously from several moving cameras.

To achieve this, the PI is developing a theory of imaging in dynamic scenes. A useful ?device? for analyzing dynamic scenes is to visualize them as constructs in spacetime, analogous to static structures in space. Much of the progress in multi view geometry in static scenes has centered on the development of tensors that embody the relative positions of cameras in space. The dimensional analogue is being used to define corresponding analogues for multi view geometry in dynamic scenes. A goal in this work is to derive geometric relationships within a system of independently moving cameras. To reconstruct unconstrained dynamic scenes, factorization approaches are being extended to spacetime to simultaneously reconstruct nonrigid structure from multiple moving cameras.

The algorithms that result from this research create the space for a host of new technologies in several industries such as autonomous vehicle navigation, distributed visual surveillance, aerial video monitoring and indexing, cellphone interface, urban navigation, coordination and planning for autonomous robots, and human computer interface. SHF: Small: Collaborative Research: Evidence based Reliability Assessment of Software Product Lines This proposal will be awarded using funds made available by the American Recovery and Reinvestment Act of 2009 (Public Law 111 5), and meets the requirements established in Section 2 of the White House Memorandum entitled, Ensuring Responsible Spending of Recovery Act Funds, dated March 20, 2009. I also affirm, as the cognizant Program Officer, that the proposal does not support projects described in Section 1604 of Division A of the Recovery Act.

This collaborative research will create techniques that improve the reliability of software product lines. A software product line is a family of software systems that share certain common features and differ according to a set of specified variations. Use of software product lines has grown rapidly in industry because such reuse reduces the cost of building new systems.
Reliability is important to product line developers since many product lines, such as mobile phones, industrial robots, and surgical imaging systems, require reliable operation. This project focuses on development of a rigorous framework for incremental assessment and prediction of software product line reliability (SPL iRAP). The research has three major thrusts: (1) developing reliability modeling techniques for software product lines to handle the effects of variations and ongoing changes, (2) investigating the use of reliability models for prediction across the product line based on empirical data, and (3) quantifying the benefit of the reuse on software quality. The researchers will train students, particularly women and underrepresented groups, in software reliability techniques, create new curriculum units for teaching, and partner with industrial developers of product lines to demonstrate the new techniques. RI: Small: RUI: Resource light Morphosyntactic Tagging of Morphologically Complex Languages This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).

The main goal of this project is to develop a tagging method which neither relies on target language training data nor requires bilingual dictionaries and parallel corpora. The main assumption is that a model for the target language can be approximated by language models from one or more related source languages.

Exploiting cross lingual correspondence leads to a better understanding of 1) what linguistic properties are crucial for morphosyntactic transfer; 2) how to measure language similarity at different levels: syntax, lexicon, morphology; 3) how this method applies to pairs that do not belong to the same family; 4) what determines the success of the model, and 5) how to quantify its potential for a given language pair. By exploiting cross language relationships, the size, and hence cost, of the training data are significantly reduced.

This project is a new cross fertilization between theoretical linguistics (especially typology and diachronic linguistics) and natural language processing. The practical contribution is a robust and portable system for tagging resource poor languages. With this new approach, it is be possible to rapidly deploy tools to analyze a suddenly critical language. This approach can also enhance NSF s initiatives in documenting endangered low density languages as it leverages exactly the type of knowledge that a field linguist and a native speaker could provide. Additional benefits include high quality annotated data, automatically derived multilingual lexicons, annotation schemes for new languages, new typological generalizations, and graduate and undergraduate researchers with significant experience of highly practical work on difficult and underrepresented languages. NeTS: Small: Collaborative Research: The Flexible Internetwork Stack (FINS) Framework Modern networks invalidate many of the assumptions of traditional networking. For example, mobile ad hoc networks (MANETs) invalidate the assumption that there will be stable routes in the network, throwing traditional routing techniques into disarray. Handheld computing devices further challenge assumptions about platform mobility. While the need for cross layer design to meet these new challenges has become well known, no replacement for the traditional network stack has emerged yet. Implementing experimental cross layer approaches on commodity hardware and software remains challenging.

In this project we are building a framework for modular, extensible, experimental, network stack implementation, called the FINS (Flexible Internetwork Stack) Framework. The framework allows users to leverage existing protocols (such as TCP and IP) where needed, providing implementations that provide more real time control and transparency than is available in existing implementations, while allowing users to replace or modify components as desired. Thus, the FINS Framework allows researchers ready access to the network stack in a manner previously possible only in simulation or by making painstaking operating system modifications.

The FINS Framework facilitates ready implementation of new network technologies and context aware applications thereby lowering the bar for participation in experimental networking research. The initial implementation of the framework is on handheld devices. The framework is being released via open source license, making it broadly available to the research community. A set of hands on networking course modules utilizing the FINS Framework and handheld devices is being developed, and utilized for undergraduate research at a predominantly undergraduate institution. SHF: Small: Designing QoS Aware MPI and File Systems Protocols for Emerging InfiniBand Clusters The emergence of affordable, high performance networking technology like InfiniBand is fueling the growth of high end computing (HEC) systems. However, there are no schemes in these systems to provide a minimal Quality of Service (QoS) for the execution of parallel jobs by taking network level contention into account. InfiniBand provides feature rich QoS mechanisms. The research focuses on the following novel research directions: 1) How to take advantage of InfiniBand QoS mechanisms to design a QoS aware MPI library? 2) How to take advantage of these mechanisms to design a QoS aware parallel file system? 3) How to dynamically provide QoS by monitoring network traffic and making adjustments to virtual lane arbitration at InfiniBand s switch and adapter hardware? 4) How to design and establish job priority guidelines with the proposed QoS framework? and 5) What kind of performance, scalability, efficiency and productivity benefits can be achieved by this proposed QoS framework with petascale applications? The transformative impact of the research enables next generation InfiniBand clusters and applications to be QoS aware and highly efficient in addition to delivering performance and scalability. The research has significant impact on the design, deployment, and utilization of next generation ultra scale systems with QoS support. SHF: Small: Global Manipulation in Solid State Quantum Information Processing Protocols and Implementation Quantum physics has been applied to study problems in computing complexity in recent years, which has become a new frontier in computer science. Computer hardware that is operated according to the laws of quantum mechanics can realize novel quantum protocols and bring enormous speed up for certain computationally hard problems. The key issue in implementing such hardware is in achieving highly accurate and fast control on the quantum logic elements so that they can beat the hazardous effects from the environmental noise. For solid state quantum processors, including superconducting systems and semiconductor systems, such control is usually achieved via adjustable local parameters, where careful designing of the circuit and the connections to external sources are required. In this project, a quantum global mode will be exploited to achieve efficient implementation of the quantum protocols. Here, the quantum global mode is the microwave photon mode in a nanoscale quantum resonator that has millimeter wavelength, can couple with multiple quantum logic elements simultaneously, and has demonstrated microsecond quantum coherence times. Meanwhile, the global mode will also be considered as a probe to measure quantum entanglement and quantum coherence effects in the solid state quantum processors. Two questions will be studied in this project. First, solid state quantum simulators that can emulate quantum many body systems involving arrays of solid state elements will be studied, where the global quantum mode will act as a control as well as a detector of the quantum phase transitions in the simulators. Second, a universal quantum computer of spurious two level fluctuators in the superconducting system will be studied where the global mode can provide individual control, effective coupling, and readout of the fluctuator states. Both the hardware aspect and the software aspect will be investigated. TC: Small: Towards Automating Privacy Controls for Online Social Networks For millions of Internet users today, controlling information access on Online Social Networks (OSNs) such as Facebook and LinkedIn is a difficult challenge. Privacy controls in current systems do not provide the necessary level of flexibility and usability to their users. Some systems like MySpace and LinkedIn allow users to grant all or nothing access control to their profiles. While simple to use, these controls are imprecise and can easily leak data to unintended recipients or prevent the legitimate sharing of data. In contrast, OSNs like Facebook provide extremely powerful controls that are unfortunately too complex for most users to configure. This proposal addresses the need for privacy control policies that are both powerful and simple to use. The proposed work provides simple and powerful privacy policies by using machine learning techniques to automatically infer user preferences from observed user behavior. The work also proposes privacy lenses, a generalized mechanism to debug privacy policies by viewing user information through the access controls of any specified user. These technical solutions will be implemented on the Facebook social network as a third party application. In addition, the data gathered from the deployed application will provide evidence to either validate or refute the perplexing phenomenon known as the privacy paradox, where users take little action to protect their privacy despite expressing strong concerns about online privacy.

The proposed project addresses a significant problem fundamental to protecting online information. By allowing the social network to learn what users want based on their actions, the PIs remove the complexity of managing privacy policies, thereby giving non technical Internet users a simple and intuitive way to customize their preferences. The work is novel in its use of machine learning techniques to infer user preferences, and can change the way privacy policies are constructed for a wide variety of Internet applications. By gathering user data from a large scale social network, the project will also provide significant support to improve understanding of the motivations behind users actions concerning online privacy. Finally, the proposed work will integrate sophisticated experimental networking research techniques with detailed human studies, adding an additional dimension to traditional experiments performed by social scientists. SHF: Small: Collaborative Research: Specification and Verification of Safety Critical Java Software is increasingly important in aircraft, spacecraft, cars, and medical devices, which are all safety critical. As the size and complexity of the software increases, so does the likelihood for defects with potentially catastrophic consequences. This research aims to simultaneously increase the level of assurance and raise the level of abstraction in safety critical systems, by supporting Safety Critical Java that uses C for driver code (SCJ+C). The project will implement and evaluate SCJ+C and provide modular specification techniques and verification tools for it. Few such tools exist for object oriented real time programs, and thus there has been little critical evaluation of techniques and tools for real time safety critical programming.
Modular reasoning about timing constraints is a known hard problem, due to the dependence of a method s timing on all methods it calls.

The project will leverage the Java Modeling Language (JML) as a specification tool to build a set of practical JML based tools for the timing analysis of SCJ+C programs. The research will allow the application of formal methods for certification, correctness, bug finding, and timing properties for real time critical systems. This will help productivity, by allowing programmers to develop and reason about their systems at an appropriate level of abstraction. CSR: Small: Monitoring for Error Detection in Today s High Throughput Applications CSR: Small: Monitoring for Error Detection in Today?s High Throughput Applications

Abstract: Much of our critical infrastructure is formed by distributed systems with real time requirements. Downtime of a system providing critical services in power systems, air traffic control, banking, and railways signaling could be catastrophic. The errors may come from individual software components, interactions between multiple components, or misconfiguration of these components. It is therefore imperative to build low latency detection systems that can subsequently trigger the diagnosis and recovery phases leading to systems that are robust to failures. A powerful approach for error detection is the stateful approach, in which the error detection system builds up state related to the application by aggregating multiple messages. The rules are then based on the state, thus on aggregated information rather than on instantaneous information. Though the merits of stateful detection seem to be well accepted, it is difficult to scale stateful detection with an increasing number of application components or increasing data rate. This is due to the increased processing load of tracking application state and rule matching based on the state. In this project, we address this issue through designing a runtime monitoring system focused on high throughput distributed applications. Our solution is based on intelligent sampling, probabilistic reasoning on the application state, and opportunistic monitoring of the heavy duty rules. A successful solution will allow reliable operation of high bandwidth distributed applications and those with a large number of consumers. We will also achieve broader impact through an innovative service learning program at Purdue called EPICS and a new course. III: Small: Fast Subset Scan for Anomalous Pattern Detection This work will develop new methods for fast and scalable detection of anomalous patterns (subsets of the data that are interesting or unexpected) in massive, multivariate datasets. There will be a focus on real world applications such as an emerging disease outbreak or a pattern of smuggling activity with complex, subtle, and probabilistic patterns that are difficult to spot with existing techniques. The research is based on two key insights. First, the pattern detection problem can be framed as a search over all subsets of the data, in which can be defined a measure of the anomalousness of a subset and then maximize this measure over all potentially relevant subsets.

Second, it has been discovered that, for many spatial detection methods (including Kulldor s spatial scan statistic and many recently proposed variants), one can perform an exact search which efficiently maximizes the measure of anomalousness over all subsets of the data. The research team will explore this new combinatorial optimization method, investigate how it can be extended to constrained subset scans and to more general multivariate pattern detection problems, and examine how it can be incorporated into a subset scan framework, enabling the creation a variety of fast, scalable, and useful methods for anomalous pattern detection.

Intellectual Merit
The research team will develop, implement, and evaluate a general probabilistic framework for efficient detection of anomalous patterns in both spatial and non spatial datasets. The proposed work will address these challenging and important research questions:
1)How can one define a useful measure of the anomalousness of a subset of the data, and efficiently optimize this measure over all subsets to find the most anomalous patterns?
2) What are the necessary and sufficient conditions for a set function F (S ) to satisfy the linear time subset scanning (LTSS) property, enabling exact unconstrained optimization of F (S ) over all 2 N subsets of N records while only requiring O(N ) subsets to be evaluated?
3) How can one extend fast subset scanning methods to general multivariate datasets, and incorporate search constraints such as proximity, connectivity, and self similarity?
4) How can one deal with uncertainty about the effects of an anomalous pattern by searching over subsets of input and output attributes as well as subsets of records?

Broader Impact
Development and testing will be prioritized in three areas: 1) early detection of disease outbreaks, 2) detecting illicit container shipments, and 3) identifying anomalous trends in social networks. These applications will allow the demonstration the value of these methods across a wide spectrum of domains. Through existing collaborations, the algorithms will be incorporated into deployed systems for health and crime surveillance that contribute directly to the public good. The Principle Investigator s lab has over 5 years of history offering free machine learning software, and the software implementations of all algorithms developed through this grant will be made publicly available. The bulk of the funding will go to training graduate students who will become the next generation of researchers to explore new methods for anomalous pattern detection.

Key Words: anomalous patterns; pattern detection; fast subset scan; scan statistics; optimization. SHF: Small: Programming Abstractions for Algorithmic Software Synthesis This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

In contrast to software verification, software synthesis writes programs rather than merely checks them for errors. While verification has recently reached the programmer, the success of synthesis remains in the hands of formally trained experts. To ease the adoption of synthesis, this proposal develops algorithmic synthesis, which is to deductive synthesis what model checking is to deductive verification: Rather than deducing a program with a theorem prover, algorithmic synthesis systematically finds the program in a space of candidate implementations.

A key remaining challenge is how to describe this candidate space. Each synthesizer must be programmed with insights about the domain and its implementation tricks. In deductive synthesis, the insight is conveyed by a domain theory. In algorithmic synthesis, programmers typically communicate their insight by writing a partial program that syntactically defines the candidate space. The partial program is then completed by the synthesizer. Since the program is specified partially, programmers can control the candidate space size, making algorithmic synthesis feasible while leaving tedious program details to the synthesizer.

This project investigates linguistic aspects of algorithmic synthesis, addressing three issues: (1) How to debug partial programs? Angelically non deterministic oracles will be used for gradual development of partial programs. (2) What is domain insight and how to communicate it? Programming abstractions will be developed for defining the candidate space naturally. (3) How to refine the insight with the goal of aiding the synthesizer scalability? An interactive dialogue between the programmer and the synthesizer will help the programmer refine and formalize her insight. SHF: Small: Leveraging the Interplay between Process Variation and NBTI in Nanoscale Reliable NoC Architecture Design The trend towards multi /many core design has made network on chip (NoC) a crucial hardware component of future microprocessors. With the continuous down scaling of CMOS processing technologies, reliability is becoming a primary target in NoC design. Negative Bias Temperature Instability (NBTI) is a critical reliability threat for deep sub micrometer CMOS technologies. NBTI increases the PMOS transistor threshold voltage and reduces the drive current, causing failures in logic circuits and storage structures due to timing violations or minimum voltage limitations. Meanwhile, process variation (PV) the divergence of transistor process parameters from their design specifications caused by the difficulty in controlling sub wavelength lithography and channel doping as CMOS manufacturing technology scales, results in variability in circuit performance/power and has become a major challenge in the design and fabrication of future microprocessors and NoCs. Since NBTI and PV affect both NoC delay and power, it is imperative to address these challenges at the NoC architecture design stage to ensure its efficiency as the underlying CMOS fabrication technologies continue to scale.

The goal of this project is to develop techniques for designing novel, cost effective router microarchitectures and adaptive routing schemes that mitigate NBTI and PV impact on NoCs by leveraging the interplay between the two. The scalability and sustainability of future many core processors crucially depend on the dependability of NoCs. Mechanisms that can simultaneously tolerate PV and NBTI will be investigated for enhancing the reliability of NoCs fabricated using nanoscale transistor technologies. The educational and outearch activities include recruiting graduate and undergraduate students from under represented groups for this project and integration of research and education. AF: Small: Algorithms for Active Learning of Interaction Networks The project will seek efficient algorithms for extracting the structure of interaction networks: systems consisting of finite populations of elements in which the state of each element may change as a result of interactions with a small set of other elements according to specific rules of interaction. Such networks are ubiquitous in the physical and social sciences, and include standard models such as Boolean circuits, Bayesian networks, social networks, chemical systems, gene regulation networks, and epidemiological models of the spread of disease. The research carried out will apply methods of active learning based on recent progress by the principal investigators on determining the structure of certain kinds of Boolean, analog and probabilistic circuits and social networks using experiments. TC: Small: In Cloud Security Services for Mobile Devices Modern mobile platforms, such as the Google Android and Apple iPhone, are reinventing the mobile landscape by opening up third party development and by providing sophisticated productivity, communication, and application suites. In addition, mobile devices are increasingly used to store sensitive personal information such as financial and medical data. Mobile environments face a wide range of unique security challenges. First, emerging mobile platforms have vastly different security and trust models. Second, techniques that worked for securing desktops do not transition well to mobile environments because mobile devices are highly resource constrained. Finally, mobile devices have inherently different usability patterns than traditional desktops that impact security.

This project explores a new model for mobile security based on moving the complexity of malware detection to an in cloud security service rather than performing analysis locally on each mobile device. We will investigate in cloud security services for mobile devices based on an architecture that consists of a lightweight agent that runs on mobile devices interposing on access of applications and data, and a network service that identifies malicious applications using parallel signature, behavioral, and reputation based detection engines. Our approach is structured around three objectives: (1) functionality across a wide variety of mobile platforms and security models, (2) minimal on device CPU, memory, and power resources, and (3) security that adapts to mobile usability patterns. We will work with our industry partners to facilitate the deployment of the techniques and methods developed through this effort on live operational networks. NeTS:Small: How to Exploit Social Network Theory in Designing Secure Wireless Networks This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Securing cyberspace is one of the top priorities in protecting our national infrastructure. As wireless ad hoc networks (WANETs) have been built to access or been parts of the Internet, they become the weakest links which should be secured and protected. This project investigates security issues in WANETs, such as trust establishment and management, secure connectivity, efficient secure routing and their performance, by utilizing the social network theory (including tools such as graph theory, percolation theory and hyperbolic geometry). First, the results in empirical social network studies are utilized to improve our understanding of the properties of the trust and physical network graphs, the existence and distribution of secure paths, and secure connectivity in WANETs. Second, by mimicking the social behavior of individuals, new methodologies are established to design efficient and scalable multi hop communication schemes in WANETs. Finally, the project investigates the secure efficiency of a WANET as the end to end secure throughput and delay and conduct formal analysis of secure network performance and trade off between security and throughput/delay. This project opens a new research direction in wireless networks and invents novel network design methodologies.

The proposed research has broad impacts in many aspects including securing national cyberspace, creating multidisciplinary research themes mixing sociology, mathematics and wireless networks, disseminating the research findings to multiple communities of interest through publications in journals and conferences, and training the future multidisciplinary work force for telecommunications and information industries. AF: Small: Graph Isomorphism and Quantum Random Walks by Anyons Quantum computers offer the potential to exponentially speed up the solution of certain classically hard computational problems. It is already known, for example, that quantum computers can efficiently factor numbers (something modern computers cannot do), and this fact implies that quantum computers can break the majority of cryptographic systems which protect our nation s cyber infrastructure. The quantum factoring algorithm is the main motivation behind current research into actually building a quantum computer. In this grant, the investigator proposes a new approach to efficiently solving a computational problem the graph isomorphism problem which might also admit an exponential speedup over the best classical algorithm for the problem. The graph isomorphism problem is to tell whether two given graphs (a collection of vertices with edges connecting them) can be made to look identical to each other by permuting the different vertices. The approach taken here is different from that taken by the majority of the quantum algorithms community and centers on a novel class of quantum random walks, those in which the walkers carry topological quantum numbers. This approach follows from a series of failed proposals to graph isomorphism based on random walks by hard core bosons or fermions and is motivated by the form in which these proposals fail. Finding an efficient quantum algorithm for the graph isomorphism problem would be potentially transformative and would provide a major new justification for building a large quantum computer. The approach chosen by the PI also introduces a novel quantum algorithm technique quantum random walks by anyons which has the potential to be useful a primitive outside of the graph isomorphism problem. Finally, the award will be used to support the training of graduate students who work on the boundary between computer science and physics, and thus strengthen connections across this interdisciplinary divide. TC: Small: Understanding the Roots of the Spam Problem Email Address Trafficking Existing anti spam techniques, such as spam filters and reputation systems, face growing difficulties due to spammers use of multimedia content (which is hard to filter) and botnets (which mask true spammer identity and thus handicap reputation systems). Te goal of this project is to complement these existing efforts by targeting email address distribution channels, where lists of addresses are bought and sold among unscrupulous parties.

As a first necessary step, this project?s goal is to better understand the email trafficking phenomenon. This is a unique opportunity to study email spam in conjunction with email address trafficking. The team recently built a system for Internet users that, as a side effect, is able to collect the data that would contain information necessary for this analysis. Second, they secured cooperation from the Case Western Reserve University IT organization to test deploy this system for up to 500 users and keep it in place for at least a year, which would allow them to collect a trove of data for a large scale study and analysis. This proposal will study spam from a unique perspective ? email address trafficking. While previous studies characterize spam by considering email content and senders, address trafficking represents an important aspect of the spam problem, and better understanding of address trafficking can open new effective ways to combat spam. The broader impact includes its potential for better mechanisms to combat spam, for fostering collaboration with data mining faculty within the department, and for enhancing graduate and undergraduate education by adding material on application level network security to, respectively ?Internet Applications? graduate course and to the ?Computer Networks? core undergraduate course. RI: Small: The Dynamics of Information Flow in Embodied Cognitive Systems This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

There is a growing realization within the behavioral and brain sciences that the embodiment and situatedness of intelligent agents plays an essential role in their behavior. However, it is still a significant open challenge to understand the complex interactions between an agent s nervous system, its body and its environment. This project focuses on mathematical methods for analyzing the flow of information in models of situated and embodied cognitive agents. Specifically, the investigator will (1) develop new information theoretic tools that characterize the flow of information over time throughout the system, (2) test and refine these methods on evolved model brain body environment systems, and (3) use these techniques and models to explore the unique advantages of embodiment and situatedness for a cognitive agent.

The new analysis techniques build on the notion of conditional mutual information between random variables at specific points in time. For example, conditioning the mutual information between a stimulus feature and a state variable at time t on the information that state variable contains about the stimulus feature at time t 1 allows one to compute a measure of information gain. Similar measures can be used to compute the information loss or retention. This basic approach will be extended in several different directions. Information factoring, which builds on the notion of transfer entropy, will allow the interactions between system components to be characterized in terms of directional information flow. Information backtracking will offer a refined portrait of the structure of these interactions by tracing backwards in time from particular informational configurations of the system to determine the flows that produce them. Specific information spectra will be used to explore the informational relationships between particular values taken on by components of the system, allowing the structure of their interactions to be probed at a finer level of detail. These methods will be applied to evolved models of relational categorization, referential communication, and visually guided behavior, allowing the following questions to be explored: How do embodied agents extract, store, and suppress information? How do embodied agents integrate information about multiple features? What is the relationship between informational and dynamical properties? Finally, these techniques will be used to characterize the capabilities of embodied and situated agents, including information self structuring, information offloading, and embodied information transfer. These analysis methods are expected to have applications not only to brain body environment systems, but also to other complex biological and social networks. NeTS: Small: Collaborative Research: Tng, a Next Generation Transport Services Architecture This award is funded under the American Recovery and Reinvestment
Act of 2009 (Public Law 111 5).

The Internet has traditionally combined many orthogonal functions into transport protocols, creating significant technical and administrative hurdles to transport service evolution. New or specialized transport protocols are now nearly undeployable because they cannot traverse middleboxes such as NATs, firewalls, performance enhancing proxies, which have mushroomed in the past two decades; new congestion control schemes are restricted by the requirement to compete ?fairly? against traditional TCP flows; and deploying new features such as multi homing is difficult because applications must be adapted to new naming and communication models.

Tng ( Transport next generation ) is a new but incrementally deployable transport architecture that breaks the above evolutionary impasse by modularizing the transport layer. Tng breaks transports into four explicit layers Endpoint Naming, Flow Regulation, Identity/Security, and Semantics plus a cross layer Negotiation service. By separating the network oriented functions of endpoint naming and flow regulation from application oriented transport semantics, Tng enables middleboxes in the network to enforce network policies and optimize flow performance cleanly across diverse network technologies and administrative domains, without interfering with end to end semantics. Tng s identity/security layer in turn enforces this separation
between network and appliation oriented functions, avoiding past conflicts between middleboxes and IPsec.

By developing a working prototype and analyzing its performance and adaptability across a variety of real and simulated network environments, we expect that Tng will prove an important step towards breaking long standing deadlocks between operators, new network technologies, and end users. III: Small: Generalization of the Association Analysis Framework Association analysis finds patterns that describe the relationships among the binary attributes (variables) used to characterize a set of objects. A key strength of association pattern mining is that the potentially exponential nature of the search space can often be made tractable by using support based pruning of patterns i.e., eliminating patterns supported by too few transactions. Despite the well developed theoretical foundation of association mining, this group of techniques is not widely used as a data analysis tool in many scientific domains. For example, in the domain of bioinformatics and computational biology, while the use of clustering and classification techniques is common, techniques from association analysis are rarely employed. This is because many of the patterns required in bioinformatics and other domains are not effectively captured by the traditional association analysis framework and its current extensions. Although such patterns can be found by techniques such as bi clustering and co clustering, these approaches suffer from a number of serious limitations, most notably, an inability to efficiently explore the search space without resorting to heuristic approaches that compromise the completeness of the search. To address the challenges mentioned above, the team will extend the traditional association analysis framework. They propose two novel frameworks for directly mining patterns from real valued data that, unlike biclustering and co clustering, are able to discover all patterns satisfying the given constraints and do not suffer from the loss of information caused by discretization and other data transformation approaches. They will also extend association analysis based approaches to work with data that has class labels by effectively using the available class label information for pruning the exponential search space and finding low support patterns that discriminate between the two data classes. To evaluate the results of the work, they will develop robust evaluation methodologies for evaluating the patterns obtained from the proposed frameworks. The proposed work promises to extend the power of association analysis to a wide range of new applications in health and life sciences, such as the discovery of biomarkers and functional modules from single nucleotide polymorphism and gene expression data, with potential applications in personalized medicine and the development of drugs and bio fuels. RI: Small: Acquisition and Modeling of Dense Nonrigid Shape and Motion The objective of this project is to advance the state of the art in acquiring and modeling dynamic non rigid objects. The specific examples of non rigid objects that the project is to focus on include: human faces, hands, soft tissues, cloths, and animals. The PI seeks to address the following two fundamental questions: (1) How can non contact optical methods be used to measure dense 3D surface motion without physically modifying the appearance of the surface? (2) What physical and/or biological properties can be inferred from the acquired dense 3D motion data?

The research team addresses these two questions by two simple but general ideas, namely the space time approach and data driven models. The space time approach builds upon space time stereo, and enables accurate optical measurements of 3D surface motion, as well as automatic registration of shape sequences among different dynamic objects. The data driven models are used for both material recognition and deformation EMG correlation. The project has a wide range of scientific impacts, including generating data for 2D face alignment and 3D face recognition in biometrics, generating data for 3D face emotion recognition in human computer interaction, measuring human body deformation in biomechanics, modeling soft tissues for orthopedics and computer aided surgery, and building virtual human models for entertainment and education. These scientific impacts translate into benefits to society, for example, by building more accurate biometric systems to secure our country, innovating surgery procedure to reduce health insurance cost, and creating 3D digital replicas of great teachers to make our education available anywhere, anytime, at a lower cost. RI:Small: Relational learning and inference for network models Networks are everywhere. Discovering the underlying principles of the networks has great impact on our understanding of complex systems in many scientific, engineering, and social research areas. Nowadays, the availability of network data, such as online social networks from facebook.com or protein protein interaction data, give researchers unprecedented opportunities to quantitatively study these complex systems.

In this project, the PI brings together problems, ideas and techniques from different areas including machine learning, statistics, biology and social sciences, to develop novel computational tools and statistical models for common problems in network inference and learning. The research activities include i) designing nonparametric Bayesian models to discover latent classes from relational data, ii) developing relational Bayesian models, coupled with efficient deterministic approximate inference methods, to predict missing links and node labels, and iii) examining network dynamics at different substructure levels.

The developed models, algorithms, and tools for analyzing network data are available to the public via publication and web distribution, disseminating to other machine learning researchers and helping computational biologists and socials scientists analyze massive network data that are being generated with an unprecedented fast speed. The PI incorporates the research results into the graduate level interdisciplinary courses he teaches and recruits graduate and undergraduate students to conduct research for this project. SHF: Small: Reliability Enhancement via Adaptive Checkpoingint in Wireless Grids This research aims to turn existing wireless mesh networks (WMNs) into productive and reliable computing platforms, made possible innovatively by reliability enhancement via adaptive checkpointing (REACT) to yield wireless Grids (WiGs). WiGs can expand immensely the wired Grid in support of rapidly growing cloud applications, besides serving as their original role of ubiquitous communications. It is extremely challenging and yet interesting to realize effective checkpointing in WiGs, due to their unique characteristics. This REACT project deals with three technical challenges, which together constitute the basis of our Checkpoint Manager, able to render WMNs into productive WiGs for enhancing and complementing wired Grids.

The project holds great promise to advance technical understanding and scientific frontiers of effective checkpointing in WiGs. It will also improve the research and educational activities on Grid computing and wireless systems strongly in the University of Louisiana, with the testbed established under this project deemed a valuable asset. New research findings and technologies for effective checkpointing and wireless communication performance enhancement will be incorporated into relevant courses, helping to integrate research and education for enriched teaching, training, and learning experience and to educate quality future scientists critical to the NSF mission. Underrepresented students will be recruited aggressively to participate in this project, taking advantage of the established REACT testbed and working collaboratively with funded graduate research assistants to stimulate their research interest. CIF: Small: A Space Dimension Approach for Wireless Netowrk Information Theory This project considers the physical limits of communication in a wireless network and how these limits can be achieved by communication protocols in an effective way. Specifically, it aims at determining the fundamental limits of the capacity of the network using an innovative approach based on physics, rather than postulating random propagation channel models. The project focuses on the characterization of the amount of spatial diversity that a wireless networks can provide, which is shown to be one of the central issues to determine its capacity. The spatial diversity is quantified in terms of the dimensionality of the propagating field, which is what carries the information through the network. Hence, drawing connections between information theory, functional analysis of continuous vector spaces, electromagnetic theory, and networks, the project aims at developing an information theory for wireless networks. The project considers different geometric configurations of wireless network, and determines their corresponding number of spatial degrees of freedom. It then considers both the effects of narrow band and wide band frequency transmission. Results are in terms of scaling laws, as well as capacity laws that are not asymptotic in the number of nodes in the network. CIF:Small:Visual Information Measures for Task Based Imaging Applications Many digital images are not acquired for documentary evidence or for archival use, but instead are intermediate pieces of information to be ultimately used by a human to assess a situation, make a decision, or reach a conclusion. Often, the image need not provide picture perfect quality in order for a human to successfully perform a task consider security screening of carry on baggage, for example. Many tasks can be easily performed with images that would be considered to be severely degraded in a purely aesthetic sense. To date, however, image assessment algorithms have been primarily focused on evaluating aesthetic quality. These algorithms are typically designed for relatively high quality images, and they do not perform well on highly degraded images which exhibit low aesthetic quality but remain useful for human performed tasks.

This research characterizes the suitability of an image for recognition and the perceived utility in terms of conveying information about content. The characterization is performed in terms of properties of the image, rather than in terms of the response of higher level vision to the image, and is based upon extensive subjective experiments quantifing the perceived utility of highly degraded images and identifing recognition thresholds for images maximally degraded images that still allow recognition. A utility measure is developed which takes as input the original image and the distorted image and outputs a distance quantifying the amount of degradation in the image relative to both the recognition threshold (at the low end) and to the original or a visually lossless representation of the original (at the high end). Use of this measure is demonstrated by integration this measure into imaging applications including compression and enhancement. SHF: Small: Collaborative Research: Taxonomy for the Automated Tuning of Matrix Algebra Software CCF 0917324
SHF: Small: Collaborative Research: Taxonomy for the Automated Tuning of Matrix Algebra Software
PI Jessup, Elizabeth R. University of Colorado at Boulder
CCF ? 0916474
PI Norris, Boyana University of Chicago
Abstract:
In response to the need for high performance scientific software, we propose to study ways to ease the production of optimized matrix algebra software. Each step of the code development process presently involves many choices, most requiring expertise in numerical computation, mathematical software, compilers, or computer architecture.
The process of converting matrix algebra from abstract algorithms to high quality implementations is a complex one. When leveraging existing high performance numerical libraries, the application developer must select the appropriate numerical routines and then devise ways to make these routines run efficiently on the architecture at hand. Once the numerical routine has been identified, the process of including it into a larger application can often be tedious or difficult. The tuning of the application itself then presents a myriad of options generally centered around one or more of the following three approaches: manually optimizing code fragments; using tuned libraries
for key numerical algorithms; and, less frequently, using compiler based source transformation tools for loop level optimizations. The goals of the proposed research are three fold. First, we will construct a taxonomy of available software that can be used to build highly optimized matrix algebra computations. The taxonomy will provide an organized anthology of software components and programming tools needed for that task. The taxonomy will serve as a guide to practitioners seeking to learn what is available for their programming tasks, how to use it, and how the various parts fit together. It will build upon and improve existing collections of numerical software, adding tools for the tuning of matrix algebra computations. Second, we will develop an initial set of tools that operate in conjunction with this taxonomy. In particular, we will provide an interface that takes a high level description of a matrix algebra computation and produces a customizable code template using the software in the taxonomy. The template will aid the developer at all steps of the process from the initial construction of Basic Linear Algebra Subprogram (BLAS) based codes through the full optimization of that code. Initially, the tools will accept a MATLAB prototype and produce optimized Fortran or C. Finally, we will advance the state of the art in tuning tools by improving
some of the tools included in the taxonomy, broadening their ranges of functionality in terms of problem domains and languages. NeTS: Small: Cognitive Antennas for Wireless Ad Hoc Networks Cognitive radios, which have the ability to adjust bandwidth, modulation scheme, transmit power, error coding, and other parameters, provide tremendous flexibility for adaptation to network conditions. While cognitive radios predominantly change spectral allocation or modulation characteristics, this project considers networks of cognitive radios in which the antennas at each node can be electronically reconfigured. Coupling cognitive radios with reconfigurable antennas gives network nodes an additional degree of freedom to increase link robustness, enhance interference suppression, and increase spectral capacity.

This project demonstrates how the flexibility in radiation patterns provided by electrically reconfigurable antennas, or ?cognitive antennas?, can enable a greater density of co channel communication links and thus increase network capacity. Multi sensor data fusion is being used to incorporate antenna and radio configuration with cognitive radio scene assessment and adaptation techniques, using multiple data sources. Distributed control techniques are being developed to adapt cognitive radio settings with minimum interaction between nodes, and and the stability and convergence of these algorithms (including convergence rate) are analyzed. We are also creating practical networking protocols to include cognitive antenna capabilities in QoS routing, congestion control, path failure mitigation, and resource management. Research results are being demonstrated on a hardware testbed. The main result of this project is a framework for controlling cognitive radios equipped with reconfigurable antennas.

The proposal fosters technical innovation and design among undergraduate students, and outreach activities to attract students, especially from under represented groups, to engineering. SHF:Small: EXACT: Explicit Dynamic Branch Prediction with Active Updates A computer program consists of many low level instructions that are executed by a microprocessor. The key to executing a program faster is executing more instructions in parallel. Branch instructions hinder this process since a branch must be executed before subsequent instructions can be executed. A microprocessor attempts to circumvent this constraint by predicting the outcome of the branch, enabling instructions from the predicted target to be executed speculatively and without delay. Because it is so critical to performance, branch prediction has been studied and steadily improved for decades. Microprocessor performance is projected to be flat for the foreseeable future, after decades of exponential growth. A breakthrough in branch predictor design would be transformational.

This project provides insight into why conventional branch predictors are limited. A whole new direction in branch predictor design is revealed by this understanding. Two interrelated problems are exposed: 1) conventional predictors often fail to distinguish dynamic branches for which specialized predictions are required, especially memory dependent branches, and 2) explicitly specializing predictions for these dynamic branches does not fix the problem alone, because stores to their dependent memory addresses change their future outcomes anyway. This project proposes two unprecedented principles for branch predictor design: first, explicitly identifying dynamic branches in order to provide them with specialized predictions and, second, actively updating their predictions when stores occur to their dependent memory addresses. Together, these two principles are called EXACT, stands for EXplicit dynamic branch prediction with ACTive updates.

The goal of the proposed research is to apply these two principles to design predictors that achieve leaps in branch prediction accuracy, halving or more than halving the number of mispredictions with respect to the best known predictor. Results with idealized implementations demonstrate such leaps in accuracy are possible and a first realistic implementation already achieves a significant fraction of this potential. To achieve broader impact, project participants will collaborate closely with industry partners, Intel and IBM, to translate EXACT technology into future microprocessor designs. NeTS:Small:Fundamental Methods and Heuristics for Advanced, Network Centric Content Distribution This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Increasingly, the Internet is used to distribute content on massive scale. Massive content distribution causes shortage of network capacity, increases costs of service providers, and impairs quality of user experience. This project develops advanced content distribution techniques for achieving complex cost performance objectives. An interesting class of existing content distribution techniques, known as ?swarming?, use ad hoc methods designed to help file receivers exchange pieces of the file they have already received. However, little is known about how to develop suitable swarming techniques for infrastructure networks under complex objectives.

This project is unleashing the potential of swarming as the most advanced content distribution technique. Unlike existing ad hoc approaches, the new approach uses general optimization theory in developing swarming algorithms. The novelty is to conceptualize swarming as a technique for distributing content over multiple multicast trees (MMT). Under the new approach, content distribution is formulated as a problem of optimal swarming over MMT. The solution algorithms become the best way of conducting swarming under each performance objective. Even heuristic solutions have performance guarantees.

The project will contribute a family of optimized and robust swarming techniques, suitable for future networks and applications. These techniques can mitigate network congestion or increase distribution speed, leading to eventual outcomes in increasing economic value, accelerating the rate of innovation, and improving productivity. The algorithms can be applied generally to other content intensive applications, including IPTV, or to scientific networks that routinely transfer massive data. CIF: Small: List Decoding for Algebraic Geometry Codes: Theoretical Analysis, Efficient Algorithms, Practical Implementation CIF:Small:Decoding of Algebraic Geometry Codes:
Theoretical Analysis, Efficient Algorithms, Practical Implementation

Error control coding ensures the reliability of data transmitted in a noisy environment and is therefore a critical component of communications systems. By adding a bit of redundancy to data, and using sophisticated mathematical algorithms to encode and decode, errors in the system can be reduced to an arbitrarily low threshold. This project concerns algebraic geometry (AG) codes, a large and powerful family of codes that includes Reed Solomon (RS) codes, which are the standard code used in commercial products today. The standard decoding algorithm for Reed Solomon codes is the Berlekamp Massey algorithm, which decodes up to the sphere packing bound. In the 1980 s and 1990 s, AG codes were discovered that yielded better error correction performance than RS codes, and efficient algorithms generalizing Berlekamp Massey were developed. A method for efficiently decoding beyond the sphere packing bound, called list decoding, was discovered in the 1990 s by Sudan.

This project advances the theoretical, algorithmic and applied understanding of list decoding for AG codes. There is strong evidence that Sudan s method, when used on high rate AG codes, can perform much better than current analysis predicts, so a primary focus is improving the theoretical underpinnings of list decoding to discover its maxim capabilities for AG codes. The investigators also improve the efficiency of current algorithms and tailor them to hardware implementation by combining classical Berlekamp Massey type methods and recent innovations due to several researchers. The investigators work with hardware and communication engineers in industry and in academia to identify applications where the special properties of AG codes will be particularly advantageous. CIF:Small: General Linear Time stepping Methods for Large Scale Simulations General Linear Time stepping Methods for Large Scale Simulations

Runge Kutta(RK) and linear multistep (LM) methods have been extensively used for the integration of ordinary and partial differential equations (PDEs). Both families of methods have well known limitations. Stability requirements limit the efficiency attainable by any LM method, whereas RK methods suffer from accuracy reduction in the presence of stiffness and nonhomogeneous boundary and source terms. General linear (GL) time stepping methods are generalizations of both RK and LM methods and therefore allow the development of new integration schemes with superior properties. However, GL methods have not been extensively studied in the context of time dependent PDEs, and very little has been done to make this class of methods available for practical use. The proposed research seeks to fill this gap.

This research will investigate theoretically order conditions for a class of general linear methods of practical importance. This theory will be used to develop new high order methods that circumvent the efficiency and accuracy reduction due to boundaries, sources, and stiffness. A rigorous analysis of the stiff behavior will be carried out in a singular perturbation framework, and will be extended to index one differential algebraic systems. The proposed research is the first to address strong stability preserving GL schemes for hyperbolic systems. A framework for partitioned general linear schemes will be developed to address multiphysics problems. The new GL methods will be made available to the science and engineering community at large through a general purpose software package. Their performance will be illustrated on real life, multiscale, multiphysics simulations arising in the prediction of atmospheric pollution. AF: Small: Collaborative Research: Studies in Nonuniformity, Completeness, and Reachability Computational complexity theory classifies computational problems into various complexity classes based on the amount of resources needed to solve them. This classification is done by measuring various resources such as time, space, nonuniformity, nondeterminism, and randomness. A better understanding of the relationships among these various resources shed light on the computational difficulty of the problems that are encountered in practice.

This project explores several central questions regarding nonuniformity, complete problems, and space bounded computations. This project attempts to discover improved upper bounds for problems with high circuit complexity. Regarding complete sets, non relativizing properties of complete sets will be explored. Space bounded computations will be investigated in the context of planar graph reachability problems.

This project addresses several basic questions in computational complexity theory. The results from this project will further our understanding of computational resources such as nonuniformity, nondeterminism, and space. Research results will be published in peer reviewed journals and will be presented at national and international conferences, thus enabling broad dissemination of the the results to enhance scientific understanding. New courses will be created and taught along the themes of this project, thus integrating teaching and research. The project supports various human resource development activities such as supporting and mentoring graduate students and inviting visitors. SHF: Small: Energy Efficient Memory Subsystems for the Many Core Era The energy consumed by the memory subsystem is increasing as a fraction of the overall server energy consumption. It is anticipated that upcoming low power, byte addressable, persistent memory technologies, namely Phase Change RAM (PC RAM) and Spin Transfer Torque RAM (STT RAM), are likely to play a major role in conserving server energy. These technologies are far superior to Flash and, most interestingly, may actually replace DRAM as well.

The research will study hybrid memory subsystems by combining these technologies and DRAM, as well as DRAM free memory subsystems. The outcomes of this investigation will be: (1) a body of knowledge about PC RAM and STT RAM and their potential benefits and limitations; (2) a collection of memory controller and operating system techniques for using these technologies to conserve energy, while bounding performance degradation to user defined limits; and (3) a simulation and operating system infrastructure that can be used by others in their investigations of memory and energy issues. This work can promote a new direction in memory subsystem design and energ conservation, one that can have a profound impact on the design of future servers. RI: Small: Coordinating Language Modeling, Computer Vision, and Machine Learning for Dramatic Advances in Optical Character Recognition The goal of this research is to develop new methods for improving the performance of optical character recognition (OCR) systems. In particular, the PI investigates iterative contextual modeling , an approach to OCR in which high confidence recognitions of easier document portions are used to help in developing document specific models. These models can be related to appearance for example a sample of correct words can be used to develop a model for the font in a particular document. In addition, the models can be based on language and vocabulary information. For example, after recognizing a portion of the words in a document, the general topic of the document may be detected, at which point the distribution over likely words in the document can be changed. The ability to modify character appearance distributions and language statistics and tune them specifically to the document at hand is expected to produce significant increases in the quality of OCR results. RI: Small: A Simple but General Hand This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Robot hands are usually simple, with just two or three fingers, perhaps a single actuator, and most often no sensors at all. These simple hands are also very specific in their function, such as picking up a specific part. Research on more general robot hands usually focuses on complex hands, often resembling human hands. This project is developing simple hands with general capabilities. The approach is inspired by a variety of simple hands, such as a prosthetic hook, which have proven generality when controlled by a human, yet have never demonstrated great generality when controlled by an autonomous robot. In particular the project is developing hands that can blindly capture objects among clutter, and testing these hands both in a factory automation application and in a home assistive robotics application.


Results from this study will be broadly applicable. Every advance in hand design enables new applications, so development of new principles broadly advancing the generality of hands will be useful. Specific cases are advancing the nation s manufacturing workforce productivity, and enabling the elderly to live independently. Results will be disseminated by scholarly publication of new principles, analysis, and experimental results, as well as distribution of analytical software, planning and control software, and hand designs. CIF:Small: Recursive Estimation of Randomly Modulated Processes Project Abstract: CCF 0916568

?CIF: Small: Recursive Estimation of Randomly Modulated Processes?

Many phenomena observed in science and engineering exhibit random variations over time. Internet traffic, speech signals, and biological signals, are a few such examples. Choosing a random process to model such phenomena involves a tradeoff between accuracy versus complexity. This project investigates fast computational methods for modeling random phenomena using a class of versatile random processes called randomly modulated processes. A randomly modulated process consists of two simpler random processes, one which is observable and another which modulates the observable process. By exploiting the structural properties of randomly modulated processes, the investigators are devising efficient and accurate methods to model a wide class of random phenomena. Specifically, the project develops recursive estimators to characterize Internet traffic accurately in real time. Other important applications of the estimators can be found in speech processing, nuclear medicine, biology, genetics, and finance.

The project focuses on the intertwined problems of signal and parameter estimation of randomly modulated processes. Using a transformation of measure approach, the investigators are developing recursive estimators for such processes and investigating feasible approaches for solving the associated stochastic differential equations. The research involves in depth investigation and comparison of the transformation of measure approach with respect to traditional likelihood based approaches that lead invariably to batch algorithms that can only be executed offline. The project also involves the derivation of asymptotic properties of maximum likelihood estimators for randomly modulated processes. Implementations of the recursive estimators are being applied to Internet traffic traces, with the objective of leveraging the estimators for network admission control and anomaly detection. SHF:Small:Collaborative Research:Dynamic Invariant Inference, Enhanced In just a decade, dynamic invariant inference has emerged as one of the most promising directions in program analysis, with a variety of applications. An invariant inference system observes a program during test execution and filters a large number of candidate invariants (i.e., suspected relations between program data), finally reporting only those that hold with high confidence. However, inferred invariants are not always true (they depend on the quality of a test suite), and the few really useful invariants discovered are often accompanied by many more true but trivial and irrelevant facts. This work improves the quality of discovered invariants by ensuring their consistency with facts that are known statically. For instance, even though the invariants describing the behavior of two functions f1 and f2 may be unknown, we may know that any valid input for f1 is also valid for f2. This fact can be incorporated in the inference process to eliminate inconsistent invariants. More generally, the work explores techniques for expressing, discovering, and employing such consistency constraints to improve the quality of produced invariants, from type information and other sources including static analysis and user supplied annotation.

The work will impact many aspects of software engineering, including scientific and industrial uses. Concrete benefits will be in the form of publications, usable software (released under an academic open source license), software prototypes, and educational activities and resources (enhancement of a textbook and current courses, internships for high school students). AF: Small: A Theory of Cryptography and the Physical World The traditional goal of cryptography is to design cryptographic algorithms for well defined tasks, such as public key encryption. We propose to study the following conceptually intriguing question: when can we embed a cryptographic function into a function which was not designed for this purpose, say a function created by nature?

In a more abstract setting, the collection of possible concepts or ?objects? is represented by a function class {f_c}, where c is a description of the (unknown) object. We do not have control over the function class {f_c}, but rather it is given by ?nature?. We assume that an object c is chosen (by ?nature?) from a distribution which has a sufficiently large entropy. Each input x represents a different measurement, or ?query?, that can be made to the object. The goal is to design an algorithm for learning c (or a ?good approximation? of c) by making queries x and observing the answers y = f_c(x). The algorithm should have the following nontrivial hiding property. Any computationally bounded eavesdropper who only observes the sequence of queries and responses (x,y) cannot learn any ?useful information? about c. NeTS:Small:Collaborative Research:Holistic Transparent Performance Assurance within the Crowded Spectrum This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Cheap commercial off the shelf wireless devices are being increasingly deployed for performance sensitive applications such as patient monitoring with body sensors and home networking for multimedia and gaming. However, wireless communications may interfere with each other when they use the same or adjacent radio frequencies. This becomes a growing issue as the public 2.4GHz spectrum is being populated by a variety of devices, including 802.11b/g routers, ZigBee sensors, Bluetooth headsets, and cordless phones. Existing interference mitigation schemes are tightly tied with the physical/MAC layers of particular platforms, and hence often cannot co exist in the same network without sacrificing the system performance.

This project develops a Holistic Transparent Performance Assurance (HTPA) framework to support performance sensitive applications in the crowded spectrum. HTPA consists of 1) a spectrum profiler that models the spectrum usage and dynamic external, intra and inter platform interferences in heterogeneous wireless environments; 2) a virtualized medium access control layer that provides unified interfaces for transparently representing, monitoring, and scheduling the underlying radio resources; and 3) a stream and system performance assurance component that schedules radio resources for each individual data stream, and harnesses network interference based on the system knowledge, MAC abstractions, and dynamic spectrum models.

HTPA enables network designers to systematically understand and mitigate the impact of complex interference that exist between heterogeneous wireless devices, and provides reference models for the standardization of future commercial wireless platforms operating in unlicensed spectrums. CSR: Small: System and Network Support for Harvesting aware Solar Sensor Networks This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Many sensor network deployments are in harsh terrain that lack infrastructure but also need to use solar energy harvesting for self sustained operation. The design of harvesting aware solar sensor networks raises numerous systems and networking challenges. From a systems perspective, a design should be able to operate under a wide dynamic range of energy harvesting scenarios, from plentiful sunlight to very limited light (e.g., thickly forested areas). From a network perspective, a protocol stack should be able to adapt to spatio temporal variability in harvesting rates due to foliage and day to day vagaries.

In this project, we will design systems and network support for harvesting aware solar sensor networks. Our systems contributions include novel sensor platforms that use thin film batteries for harvesting under low light conditions, predictive duty cycling under highly variable harvesting conditions, and lightweight checkpointing and restart to handle intermittent loss of power. Our networking contributions include a link layer that is optimized for duty cycle variability, a network layer that uses harvesting aware path metrics, and a transport layer can offer high throughput under diverse harvesting conditions.

Our educational plans consist of K 12 teaching through a summer high school outreach program, and inter disciplinary REU programs in collaboration with Harvard Forest for deploying solar powered sensor networks for ecological monitoring. SHF: Small: In Vivo Software Monitoring: Architectural and Compiler Support Monitoring of software s execution is crucial in numerous software development tasks. Current monitoring efforts generally require extensive instrumentation of the software or dedicated hardware it resembles studying the software specimens in vitro. To fully understand software s behaviors, the production software must be studied in vivo in its operational environment. To address these fundamental software engineering challenges, this research addresses a framework for in vivo monitoring and observation of software intensive systems.

Three fundamental requirements are placed on in vivo monitoring
frameworks: non intrusive, low overhead, and predictable. These frameworks must also allow low level monitoring and be highly flexible to enable a broad range of monitoring activities. To satisfy these requirements, this research changes how software is compiled and how hardware is designed by pursuing the following specific aims: (i) provide flexible architectural support shared by a variety of monitoring activities; (ii) develop a monitor aware compiler that generates the monitor together with the software to be monitored; and (iii) develop state extraction optimizations to efficiently extract program states from an executing application and forward the states to the monitor. The resulting framework is empirically evaluated to assess its performance as compared to related solutions and assess its flexibility for a variety of software engineering monitoring tasks. NeTS: Small: Connected Coverage of Wireless Sensor Networks in Theoretical and Practical Settings This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Deployment is a fundamental issue in Wireless Sensor Networks (WSNs). In many missions today, sensors are deployed deterministically in a planned manner. Instances include airport/harbor monitoring, intruder tracking on government property, etc. This project studies optimal deployment patterns in WSNs, which are those patterns that achieve desired coverage and connectivity requirements with the fewest sensor nodes. Knowledge of these patterns can help avoid ad hoc deployment to save costs, minimize message collisions, improve network management, etc. However, exploration of these patterns is difficult and far from mature in WSNs. This project comprehensively studies them in both theoretical and practical settings for WSNs in 2 and 3 dimensional spaces. Three major research tasks are carried out: (1) exploring optimal deployment patterns for 1 coverage and k connectivity in 2 dimensional space; (2) exploring such patterns for connected m coverage (m > 1) in 2 dimensional space; (3) exploring such patterns for connected coverage in 3 dimensional space. If successful, this project will result in a set of optimal deployment patterns in theoretical and practical settings to achieve multi coverage and multi connectivity in 2 dimensional and 3 dimensional spaces with different ratios of sensor communication range to sensing range. The research can help establish theoretical foundations and practical guidelines for planned deployment not just for WSNs, but also for other wireless networks, such as mesh and cellular networks. In addition, the research results will broaden understanding of applications of computational geometry and topology in computer networks. SHF: Small: Automated Debugging Techniques for Modern Software Systems Proposal Number: CCF 0916605
TITLE: SHF: Small: Automated Debugging Techniques for Modern Software Systems
PI: Alessandro Orso

Debugging, which consists of identifying and removing software faults, is a human intensive activity responsible for much of the cost of software maintenance. Existing approaches for automated debugging can help lower this cost, but have limitations that hinder their effectiveness and applicability. This project aims to overcome these limitations by developing a family of debugging techniques that (1) target realistic debugging scenarios, in which faults can involve multiple statements and manifest themselves only in specific contexts,
(2) apply advanced static and dynamic analysis techniques to automatically reduce the amount of both statements and inputs that developers must examine when investigating a failure, (3) leverage information collected from the field to increase the relevance and effectiveness of the debugging process. These newly defined techniques will be evaluated through rigorous experimentation performed on real software, in real settings, and under realistic assumptions. Debugging tools, infrastructure, and experimental subjects developed within the project will be made freely available to researchers and practitioners to help dissemination and enable further research. By advancing the state of the art of debugging, this research will help developers build more reliable software systems and, ultimately, increase the overall quality of our software infrastructure. AF: Small: Algorithmic Problems in Applied Computational Geometry Computer technologies play an important role in modern medicine and life
sciences, especially in diagnostic imaging, human genome study, treatment
optimization, and medical data management. Many computational problems
arising in the field of biomedicine call for efficient and good quality
algorithmic solutions. This project aims to develop new geometric
computing and algorithmic techniques for solving computational problems
in biomedical and engineering applications.

This research investigates a number of geometric optimization problems
that are theoretically challenging and practically relevant. The target
problems belong to fundamental topics of computational geometry, such as
geometric partition, covering, shaping, approximation, motion planning,
and clustering; they also arise in important applied areas such as
radiation cancer treatment, medical imaging, biology, computer aided
manufacturing, and data mining. Some of the algorithms and software
developed during the preliminary studies of this research have produced
significantly better solutions for real application problems (for example,
much improved radiation cancer therapy plans over those computed by the
current commercial radiation treatment planning systems). The research
will draw diverse techniques from other theoretical areas such as graph
algorithms, combinatorial optimization, discrete mathematics, and
operations research. It will also provide a rich source of interesting
new problems/questions and new ideas to prod further development of
algorithmic techniques in computational geometry and other theoretical
areas. This project is expected to generate broader impacts beyond
computational geometry and even computer science. It will produce
efficient and effective algorithms and software for solving key problems
in radiation cancer therapy and surgery, medical imaging, biology, and
other applied areas. Furthermore, the newly developed algorithms and
software will be incorporated into practical applications such as
clinical radiation cancer treatment systems. Hence, this research will
help unite and integrate the power of computational geometry, computer
algorithms, and modern biomedicine for diagnostic imaging, radiation
cancer treatment, and other applications, and improve the quality of
life for the patients. RI: Small: Computational Models of Context awareness and Selective Attention for Persistent Visual Target Tracking Although persistent and long duration tracking of general targets is a basic function in the human vision system, this task is quite challenging for computer vision algorithms, because the visual appearances of real world targets vary greatly and the environments are heavily cluttered and distractive. This large gap has been a bottleneck in many video analysis applications. This project aims to bridge this gap and to overcome the challenges that confront the design of long duration tracking systems, by developing new computational models to integrate and represent some important aspects in the human visual perception of dynamics, including selective attention and context awareness that have been largely ignored in existing computer vision algorithms.

This project performs in depth investigations of a new computational paradigm, called the synergetic selective attention model that integrates four processes: the early selection process that extracts informative attentional regions (ARs), the synergetic tracking process that estimates the target motion based on these ARs, the robust integration process that resolves the inconsistency among the motion estimates of these ARs for robust information fusion, and the context aware learning process that performs late selection and learning on the fly to discover contextual associations and to learn discriminative ARs for adaptation.

This research enriches the study of visual motion analysis by accommodating aspects from the human visual perception and leads to significant improvements for video analysis. It benefits many important areas including intelligent video surveillance, human computer interaction and video information management. The project is linked to educational activities to promote learning and innovation through curriculum development, research opportunities, knowledge dissemination through conferences and the internet as well as other outreach activities, and the involvements of underrepresented groups. RI: Small: Organizing recognition: the uses of perceptual organization Recognizing objects in images is the central problem of computer vision. One approach (``bag of words ) compiles simple statistics on the image brightness patterns and recognizes by correlating these statistics, via learning, with the imaged objects. It cannot exploit important information on the image s spatial layout. Another approach, perceptual organization (PO), computes a distinctive, structural description of the image contents and recognizes based on this. Researchers agree that PO is a crucial early stage of recognition and that its results are unreliable. (Images compress the 3D world and are ambiguous; PO cannot eliminate the ambiguity since it has no high level knowledge of what the image is ``about. ) This leads to a fundamental dilemma: How can a recognition system use the result of PO if it cannot be trusted?

To exploit perceptual organizations without succumbing to their unreliability, this project uses a strategy that averages over all possible organizations weighted by their probability, instead of computing a single, ``most likely image description. This strategy is applied to diverse tasks such as matching images by the shapes of the objects within; recognizing articulated objects such as people and animals; tracking objects through video; and computing stable perceptual organizations. The project also studies the integration of this approach with older ones into a flexible and capable recognition system. The result will be new techniques for the analysis, manipulation, and search of images. The methods developed will be integrated in the curriculum and disseminated to researchers, and the software will be made publicly available. III:Small: Commugrate A Community based Data Integration System The goal of this project is to build Commugrate, a community driven
data integration system that capitalizes on the information gained
from the interactions of communities of humans with data sources.

Commugrate tackles key challenges raised by the increase in the
number of information sources used in science, engineering, and
industry, as well as the need for large scale data integration
solutions to enable effective access to these sources. These challenges
include schema matching and mapping, record linkage, and data repair.

More specifically, Commugrate (i) utilizes both direct and indirect
contributions from different types of human communities with a focus on
the latter contributions, (ii) solves key data integration issues using
new evidences like usage and behavior data which have not been previously
used, (iii) adopts a new technique for schema matching, which defines
a new class of its own, namely usage based schema matching,
(iv) introduces the first of its genre technique for record linkage based
on entities behavior, and (v) provides an adaptive feedback system to
improve the quality of the data by making the best use of users feedback.

Commugrate has a broad impact across multiple segments of society as data
integration is by far the most important and in the same time vexing issue
in many areas in sciences, engineering, and industry. Furthermore,
leveraging users interactions with data sources, especially indirect
interaction, may provide several benefits and help solve many intractable
data integration tasks which cannot be done without human intervention.

PhD students will pursue research in this project. Publications, technical
reports, software and experimental data from this research will be
disseminated via the project web site at
http://www.purdue.edu/cybercenter/commugrate. RI: Small: Ensemble Modeling of Speech Signals for Automatic Speech Recognition This project is aimed at developing a new framework of ensemble modeling of speech signals to address the long standing challenge of robust and accurate recognition of spontaneous speech. Toward this goal, random forests based allophonic clustering is used to construct ensemble models of allophones by random sampling the variables underpinning the allophonic variations; data sampling is used to enrich the diversity of the ensemble models by balancing within set data sufficiency and between sets data diversity; functional discriminative training is used to further optimize the efficiency and accuracy of the ensemble models. Experimental evaluations of these methods are performed on a standard speech recognition task to facilitate direct assessments of their efficacy by the speech research community. The ensemble modeling approach promises higher accuracy performance and lower computation costs than the current multiple system integration approach, owing to the improved likelihood scores contributed by the ensemble models in local steps of decoding search. The approach as advocated in this project opens up a new paradigm for investigating the many issues in speech acoustic modeling, it offers a new way for ensemble modeling of structured data generally, and therefore it has the potential of significantly impacting the fields of speech recognition and other machine learning applications. The research findings are disseminated via journal publication, conference presentation, and a website. The methods of this project have broad applications in speech recognition and structured data classification, and particularly they are employed to improve the accuracy performance of a telemedicine automatic captioning system. CIF: Small: Cost and Value of Information for Resource Allocation in Wireless Networks Resource allocation in wireless networks concerns the decisions of what rate each node injects data into the network, where and when it sends out data packets with potentially different destinations, and what level of transmit power it uses at every point in time, based on the information available at the node on the system state. It is well known that feedback of relevant ``network state information is of paramount importance to the efficient utilization of the scarce network resources. Therefore, in a realistic system, the cost and value of obtaining the necessary network state information is of crucial importance and demands a comprehensive study, which constitutes the broad objective of this research. The inclusion of these factors causes drastic changes to the widely used design paradigms.
To that end, the main objectives of this project are: to develop a well founded theoretical framework for (1) establishing the value and cost of information obtained by distributed network state estimators; (2) building a joint estimation and resource allocation component for new generation wireless networks; and (3) designing distributed network controllers with asymmetric partial information. This study requires fundamental developments in wireless communications, estimation theory, information theory and optimization theory. NeTS: Small: Multicommodity Flows in Multihop Wireless Networks Multicommodity flows, or multiflows in short, are central topics in all types of communication networks, whether wired or wireless. Computing a maximum (concurrent) multiflow in multihop wireless networks is particularly challenging because the capacities of the communication links, rather than being fixed, vary with the underlying link schedule. A unique challenge is thus how to compute a link schedule subject to the wireless interference constraint which induces a link capacity function supporting a maximum (concurrent) multiflow. This project establishes both the computational hardness and approximation hardness of computing maximum (concurrent) multiflows in multihop wireless networks, and develops practical approximation algorithms with provably good performance. A polyhedral approach is taken by the project to construct various polynomial approximate capacity subregions of multihop wireless networks. These approximate capacity subregions not only are the algorithmic foundation of the computing maximum (concurrent) multiflows, but also serve as a basis for interesting future projects on network capacity and cross layer design and optimizations in multihop wireless networks. They are also of independent interest to the theoretical computer science community at large. This project provides scholarships to graduate students and offers research topics for strong dissertation works on multihop wireless networks. The outcome of this project will not only be disseminated to the professional researchers through journals and conference proceedings, but also be integrated into the lecture notes targeted for senior undergraduate students and graduate students. RI: Small: Algorithms for Sampling Similar Graphs Using Subgraph Signatures Abstract below:

Graphs and networks are a natural representation across a wide range
of disciplines and domains. Statistical tools have recently been
brought to bear on the analysis of graphs, yielding rich dividends in
various application areas. The aim of this project is to use tools
from statistics and graph theory to develop algorithms that generate
similar graphs efficiently. Since graph data is often expensive to
collect, it is desirable to synthetically generate graphs. To be
widely applicable however, the generated graphs need to both preserve
the semantics of the original data (i.e., be drawn from the same
distribution) and be efficient to compute.

Two key questions form the core emphasis of the current project.
First, how does one measure similarity between two graphs? Second, how
can this notion of similarity be used to generate new graphs? On the
topic of similarity, the project will investigate representations to
preserve global properties, propose new, efficient, representations
for signatures, and explore sampling techniques and their convergence
behavior. On the topic of generation of new graphs, the project will
develop an exponential random graph model using signatures,
investigate feature selection via regularization, propose novel
methods to sample from the exponential random graph model and novel
techniques to produce proposal graphs, and provide rigorous empirical
validation across a range of application areas.

The project will facilitate the study of large complex structures of
the kind frequently encountered in domains like theoretical ecology,
social networks, and chemo informatics, allowing researchers in these
domains to leverage statistical network analysis tools to identify
significant patterns and understand algorithm performance.

For further information see the project web page:
URL: http://www.stat.purdue.edu/~vishy/graphs.html SHF: Small: An Adaptive Architecture Fabric for Constructing Resilient Multicore Systems As semiconductor technologies scale continues to shrink, building fault tolerant, defect free microprocessors becomes increasingly difficult. Tighter design constraints, lower operating voltages, and increasing power densities have lead to circuits that are more susceptible to manufacturing defects, transient faults, and wearout related failures. It is anticipated that future designs will consist of 100 billion transistors, many of which will be unusable due to manufacturing defects and many will fail over time due to wearout and other errors. Traditional mainframes and mission critical systems rely on redundancy to overcome such failures. Reliability is essentially viewed as a tax that is levied in the form of additional silicon area devoted to constant double and triple checking of results that end users must pay to ensure correct operation. As reliability concerns invade the desktop and cellphone environments, large scale redundancy is impractical due to the high cost and energy overheads.

This research proposes StageNet, a new style of microprocessor architecture in which reliability is not a tax, but rather built in to the natural operation of the system. StageNet is both introspective to enable continuous monitoring and adaptation to reliability hazards and reconfigurable at a fine grain level to minimize the lifetime performance impact that individual failures have on the system. StageNet consists of three major components: an adaptive computing substrate that enables dynamic reorganization of individual microprocessors, armored cache designs to provide high defect tolerance with low area overhead for on chip caches, and a dynamic adaptation system to manage the execution of applications and organization of the hardware over its lifetime. The broader impact of this research is that it creates cost effective ways of dealing with faulty transistors that will enable the proliferation of embedded computers into more aspects of life, where robustness and reliability are current barriers. III: Small: Automatic Incremental Design for Next Generation Database Systems Data management systems are undergoing a sea change. Specialized
engines with specialized physical storage structures are emerging to
address the extreme data volume and performance requirements of modern
applications. At the same time, the complexity of these systems, the
applications in which they are used and the platforms on which they
are built is increasing. Thus, there is a growing need for automatic
database design for these emerging database systems. Unfortunately,
previous work in automatic design cannot be used directly. While the
conceptual framework of previous research is useful, the specifics
must be reworked in order to adequately take advantage of the
opportunities and address the challenges that these new structures
present. Furthermore, most design tools only create complete designs.
There is also a need for automatic designers that can produce a new
design that is sensitive to the cost of migrating an old design to a
new one. Such an incremental designer would often generate a
sub optimal design if that design will produce 80% of the benefit with
20% of the work.

The PIs propose to investigate this incremental automatic design
paradigm for newly emerging database systems. Specifically, the
project addresses three data management platforms: column stores for
OLAP, main memory, cluster based systems for OLTP, and an extension to
row stores for exploiting correlations in data attributes.

The PIs expect to extend the work on current design tools by
demonstrating workable incremental designers as described above. There
is a strong need for such tools, and if this research is successful,
it should enable successful deployment of many new style data
managers. Moreover, incremental design ideas based on sound
cost benefit analysis are applicable to other data intensive computing
environments and constitute an important direction towards truly
autonomic computing. Further information on the project can be found
on the project web page: http://database.cs.brown.edu/projects/auto/ III:HCC:Small: Measuring and Monitoring Technical Debt A major obstacle to delivering the increasingly complex software systems that society demands is the resource drain from maintaining existing systems. The high expense of maintenance is related to the tendency of software quality to decline over time. Maintenance is often performed under tight resource constraints, with the minimal amount of effort required. Typically, there is a gap between this minimal amount of work and the amount required to maintain the software s quality. This gap can be viewed as a type of debt, which brings a short term benefit (usually shorter release time) but which might have to be paid back, with ?interest? (decreased productivity), later. Many practitioners find this metaphor intuitively appealing and it is already transforming the way that long term software maintenance is viewed. But its lack of a sound theoretical basis, empirically based models, and practical implementation hinder its ability to transform how maintenance is done. Thus the contribution of this work is to provide empirically based models describing, and validated mechanisms for managing, technical debt. This project also supports the PI s activities in mentoring a diverse population of students, as well as UMBC s nation wide prominence in the advancement of women and minorities in science and technology. CIF: Small: Structured Transmission Strategies for Wireless Networks Abstract
0916713 CIF: Small: Structured Transmission Strategies for Wireless Networks

In this project, we use lattice and other structured coding techniques to induce alignment in wireless networks. This effort uses these codes on three different fronts: a.) Interference networks, where we use structured codes to align the interference seen at each receiver. The objective is to determine the capacity of this channel to within a constant gap using these codes. b.) Cognitive networks where we use lattice codes to mitigate the interference seen by both the licensed and the cognitive radios. We exploit code structure to both (partially) learn the interfering signal at the cognitive radio and then use this knowledge to precode/align our interference signal. c.) Secure wireless networks: again, we utilize the structure of the codebook to determine simple transformations at the source in order to keep eavesdroppers in the network at bay. We employ these codes to detect, and depending on code structure, correct for modification attacks on the codebook. SHF: Small: Energy Recycling VLSI Systems This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).

ID: 0916714
Papaefthymiou Marios
University of Michigan Ann Arbor
SHF:Small: Energy Recycling VLSI Systems

This research project will investigate novel technologies for the design of very large scale integrated (VLSI) computer systems that achieve unprecedented levels of energy efficient operation through energy recycling. In contrast to conventional computer systems that consume all the energy supplied to them while computing, energy recycling computers reclaim and reuse any energy that remains undissipated during their operation. Therefore, they have the potential to operate with substantially lower energy consumption than conventional computers. This project will encompass a broad spectrum of design technologies for energy recycling computers, including circuitry, computing architectures, and design methodologies. The effectiveness of these technologies will be assessed through the design, fabrication, and experimental evaluation of proof of concept hardware prototypes.

With power consumption in high performance microprocessors exceeding 100Watts, the design of energy efficient computers has become a top priority in electronic design due to reliability concerns caused by excessive heat generation. Furthermore, energy efficient computers play a key role in the development of new mobile applications due to battery life considerations. And last, but not least, the power requirements of computing devices, including high performance servers, desktops, and laptops, is placing an increasing burden on the power grid, with emissions from all these sources growing at a reported annual compound rate of 6% and thus posing a serious environmental concern. The outcomes of this research project can therefore be transformative, resulting in innovative design technologies for realizing next generation computer systems that achieve unprecedented levels of reliable and energy efficient operation, enable new mobile applications, and promote sustainability. CIF: Small: A Stochastic Approximation Approach to Network Communications with Feedback Communication across networks with feedback and relays is an important, challenging, and largely open problem in information and communication theory. Systematic and scalable communication algorithms will enhance reliability, decrease coding complexity or delay, impart robustness to communication schemes by exploiting diversity, and in some cases increase the achievable communication rates. Since all modern communication systems employ multiple feedback mechanisms, a better understanding of communication feedback is imperative.

This research focuses on a deeper study of this topic by viewing it through the lenses of consensus algorithms and more general stochastic approximation algorithms. This approach yields scalable and robust algorithms for cases of extreme relevance to modern communication systems, e.g., network scenarios with noisy feedback. A confluence of problems, techniques and tools from information theory and distributed dynamic systems are utilized, with a potential transformative impact on both fields. In addition to analytical techniques and simulations, the team also utilizes facilities and experience realizing novel communication algorithms in a wireless network testbed based upon software defined radios.

Any impact on the problem of communication across networks with noisy feedback will have immediate applications to most modern communication systems. Moreover, this research furthers the unification of the two aspects and research communities relevant to a more general information theory one that considers both the transmission of data (classical information theory), as well as its utilization (classical dynamical systems). An emphasis on organizing special sessions in conferences, offering new graduate courses, including undergraduates in the experimental research, mentoring minority students, and furthering outreach to high school students interested in engineering is an integral part of the project. RI: Small: AquaSWARM: Small Wireless Autonomous Robots for Monitoring of Aquatic Environments The goal of the AquaSWARM project is to design and develop small, energy efficient, autonomous underwater robots as sensor rich platforms for dynamic, long duration monitoring of aquatic environments. A novel concept of gliding robotic fish is investigated, which merges the energy efficient design of underwater glider with the high maneuverability of robotic fish. Gliding motion, enabled by pitch and buoyancy control, is exploited to realize dive/ascent and large distance horizontal travel. Soft actuation materials based flexible tail fins are used to achieve maneuvers with high hydrodynamic efficiency. The research is focused on understanding gliding design for small robotic fish, and addressing the energy efficiency issue from a systems perspective. Schools of such autonomous robots are deployed in lakes at the Michigan State University Kellogg Biological Station to detect harmful algal blooms (HABs) and validate models for HAB dynamics.

The project is expected to result in cost effective, underwater robots that can perform uninterrupted, long duration (several months), long travel (hundreds of miles) operation in aquatic environments. This will provide a novel, viable, versatile, cyber physical infrastructure for aquatic environmental monitoring, with applications ranging from understanding the impact of global warming, to environmental protection, drinking water reservoir safety, and seaport security. The project also offers an interdisciplinary training environment for graduate and undergraduate students, and provides outreach opportunities to inspire pre college students and train highly qualified teachers. Robotic fish based HAB detection will also be used as a tool to engage communities at local lakes and stimulate their interest in novel technology and environmental issues. SHF: Small: Managing Non Determinism in Multithreaded Software and Hardware This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

In the 21st century, the dominant computing platform has shifted to multicore chips that implement cache coherent shared memory and run multi threaded applications. Unfortunately, these chips do not provide a deterministic model to either software or hardware developers. Reasoning about and testing for multiple possible executions is much harder than reasoning about and testing for a single correct sequential execution, as was possible under the von Neumann model that dominated in the 20th century. Easing the burden of programming multicore chips is critical to provide society with the rapid, cost effective performance gains that we have all come to expect. Moreover, broad impact requires practical solutions that do not ask industry to discard or rewrite billions of lines of existing general purpose thread based software.

To this end, research under this proposal will develop solutions for managing non determinism with alterative implementation approaches that provide complementary benefits and opportunities. (1) Work will expand techniques of recording executions for deterministic replay to improve replay parallelism and extend the scope of record/replay to hardware debugging and fault tolerance. (2) Work will develop and advance a deterministic coherence model that eliminates a major source of non determinism in shared memory multiprocessor systems: memory races. (3) Work will develop both all software and hardware accelerated implementations of deterministic coherence, in part, through extensions to the Wisconsin GEMS simulation infrastructure. (4) Finally, work will explore rebuilding coherence upon a formal deterministic foundation. Broader impacts will include embodying the proposed work in public software releases (e.g., GEMS) as well as dissemination to students and through courses, talks, industrial affiliates, and commercial influence. TC: Small: Taint Based Information Tracking in Networked Systems TC: Small: Taint Based Information Tracking in Networked Systems
PI: Nick Feamster (Georgia Tech)

Network operators must control and monitor the flow of information within and across their networks. Existing mechanisms for controlling information flow are primarily host based: operating systems can taint portions of memory or applications based on the inputs to a particular process or resource. Unfortunately, if a host is compromised or otherwise breached, that information may propagate in unintended ways. Once information has leaked, tracking the provenance of the leaked data is challenging. This project is developing a mechanism for tracking and controlling information flow across the network to cope with these problems. This mechanism would allow operators to control how information propagates within and between networks and to devise more complex policies; for example, it might be used to control which application traffic was allowed on which part of the network. The information carried in the network traffic might ultimately be attributed to a specific user or process, thus allowing operators to express policies according to the process and user that generated the traffic.

We are addressing several research challenges. First, we are exploring the appropriate granularity for tainting that preserves semantics without imposing unacceptable memory and performance overhead. Second, we are designing the system to minimize performance overhead on applications. Third, we are exploring translation mechanisms between host based taints and network based taints, so that taints carried in network traffic convey meaningful semantics without imposing prohibitive network overhead. The research will result in an information tracking and control system that is deployed
in experimental settings (e.g., the Georgia Tech campus network) using the existing and forthcoming programmable switch implementations, and integrated into undergraduate and graduate networking and security courses. III:Small:Integrated Digital Library Support for Crisis, Tragedy, and Recovery Today people make novel uses of social networking and other internet software to respond to tragic events in creative and dynamic ways. Since shortly after the April 16, 2007 shootings at Virginia Tech, this research group has integrated digital library, data and text mining, information visualization, and social network analysis techniques to help with understanding and recovery from this tragic school crisis. This proposal is meant to research and develop a next generation domain specific digital library software suite, the Crisis, Tragedy and Recovery (CTR) toolkit, building upon 17 years of work on digital libraries, as well as expertise in information retrieval, data and text mining, database management, human computer interaction, and sociology. Advanced intelligent information integration methods have not been sufficiently applied to this domain. The impact of events is felt over extended periods, requiring longitudinal perspectives to understand their complexity and inter dependencies. Consequently, with the Internet Archive and other partners, the group will begin to create CTRnet, an integrated distributed digital library network for providing a rich suite of CTR related services. Such work will help ensure that further tragic events might be better understood and prevented. NeTS:Small:View Upload Decoupling: A Redesign of Multi Channel P2P Video Systems This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Although there are several large scale industrial deployments of peer to peer (P2P) live video systems, these existing systems have several fundamental performance problems, including huge channel switching delays, large playback lags, poor performance for less popular channels, ISP unfriendliness. In these traditional systems, a peer only redistributes the video it is currently watching. In this research, the PIs are exploring a radically different approach to P2P live video streaming, View Upload Decoupling (VUD). The main idea of VUD is to have each peer distribute one or more channels, with the assignments being made independently of what the peer is viewing. This novel approach has three major advantages over the traditional isolated channel designs: channel churn immunity; cross channel multiplexing; and the enabling of structured streaming. The PIs are developing tractable analytical performance models for multi channel P2P video streaming systems, for both VUD and traditional design approaches. The analytical results not only highlight the advantages of the VUD approach, but also provide important ``rules of thumb for the design of VUD systems. The PIs are developing dynamic VUD provisioning algorithms that are both robust with respect to channel churn and also adapt to dynamic channel popularity and flash crowds. The PIs are developing VUD provisioning, management and streaming schemes that take into account ISP locality and largely reduce the video streaming traffic imposed on ISP networks. The PIs and their PhD students are also developing an open source VUD prototype. SHF: SMALL: Stateful Interfaces Interfaces are used to specify and verify the interaction between components of a system in a wide variety of programming languages and distributed systems. Stateful interfaces add expressive power by allowing the possible interactions to change over time. The goals of this project are to develop a foundational framework for stateful interfaces, and to apply the framework in two domains: (a) typesafe, generic components for efficient XML stream processing, with application to web services and related distributed system components; (b) memory consistency specifications, with application to reliable shared memory programs that take advantage of increasingly common multi processor systems. This project integrates session types for communication centered programming and typestates for object protocols by providing a common foundational framework that includes the following key features: (a) polymorphism, allowing reuse of typed code; (b) copyable, non linear use of objects and channels, allowing several clients to share a single reference to a server; (c) expressive quantificiation, allowing the specification of memory consistency guarantees that are ubiquitous in shared memory programming. The research will advance foundations that will help improve software development and debugging of shared memory multi core programming. NeTS: Small: Secure and Efficient Multipath Based Network Utilization: A Game Theoretic and Optimization Approach This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

This project addresses the design and utilization of high capacity data networks, with the goal of increasing throughput, network security, and reliability. These are issue of paramount importance in the current development of high speed networks as a computation and communication infrastructure.

The research is using routing methods that employ multi paths to enhance throughput and security. Techniques from optimization and approximation algorithms are being applied for designing efficient algorithms that optimize throughput with latency, jitter, reliability, and costs constraints. Multi paths also allow protection against attackers attempting to either eavesdrop or disrupt communications. Efficient algorithms for the computation of the Nash and Stackelberg equilibria in the resulting two player designer attacker games are being designed to provide strategies (selection of routes) for the network administrator and insight for the design of a network of low set up cost and high throughput.

This project contributes to more secure and better utilized high speed networks, which find applications in collaborative computations for astronomy, bioinformatics, and high energy physics. The results of the project will lead to the design of provably efficient algorithms for routings which achieve desired throughput and security, and for computing exact or approximate equilibria in designer attacker games. Simulations will illustrate the benefits on both synthetic and real life backbone networks. This will open the way for implementation of multi path routing. The results will be published in technical conferences and journals. A significant part involves training of undergraduate and graduate students in computer networks. SHF:Small: Locality Driven Architectures for Scalable Multicore Systems The successive innovations in semiconductor manufacturing over the last 35 years of Moore s law have turned what used to be a room sized computer system into a single chip composed of billions of transistors. These levels of integration have forced a change toward parallelism in computer system design, including both single chip multiprocessors and systems on a chip. Today, these chips have a few tens of individual processors but future scaling will make possible hundreds or thousands of them on a chip. Two important challenges have emerged which threaten to hinder performance scaling in multicore systems. First, while technology scaling will continue to enable increased transistor counts for the foreseeable future, power and thermal limitations will prevent all but a small fraction of them to be operating simultaneously at full speed. Second, the speed of the communication paths from the multicore chip to its external memory and to other processors is increasing at a slow rate. Because these communication paths must be shared by more and more on chip processing cores, the paths must be used as efficiently as possible to prevent them from becoming a bottleneck in the system.

This project seeks to develop new computer hardware and software mechanisms that exploit data locality in high performance systems, including repeated use of a data item as well as use of multiple data items that lie near one another in memory. In particular, the project will develop hardware mechanisms for bulk data transfers that support renaming, packing, and integration into the virtual memory system. The PI will also develop hardware mechanisms in the on chip memory system that will allow it to adapt to different programming primitives as well as to different coherence needs among the processing cores. The mechanisms will be evaluated in terms of effectiveness and programmability using a range of applications.

This research aims to develop technologies critical to emerging parallel multicore chips, without which such chips will not be able to meet performance and power goals. Enabling enhanced performance in a power efficient manner is critical to all deployments of future computing platforms, including those for science, commerce, and national security. The broader impact of this research will include training graduate and undergraduate students as researchers, while also working to increase participation of underrepresented groups in computing. The primary outreach activity will include participation in a summer camp to attract high school girls to computer science. SHF: Small: Thermal Aware High Performance DRAM Architectures in Multicore Technologies Environment Protection Agency estimates that by 2011 data centers nationwide would consume electricity amounting to the equivalent output of about 30 power plants. Clearly, improving the power and thermal behavior of these systems has a direct impact on their energy efficiency and reliable operation. DRAM memories constitute a significant fraction of the power consumed in computers. In addition, in the dawning era of multi /many core processors, the performance of a system is largely dependant on its main memory efficiency. This project investigates ways to operate DRAMs at full bandwidth utilization while spending the minimum power per unit of data communicated and maintaining lower operating temperatures. Specifically, this work aims at answering three fundamental questions: a) how can we enhance processor architectures to balance the activity on different DRAM chips to protect chips under thermal stress, b) how can we enhance the DRAM systems to reduce their power consumption and peak operating temperatures, and c) how can we enhance the operating systems to improve the thermal behavior of DRAM systems?


Techniques developed in this project will improve the thermal behavior of DRAM systems and hence will reduce the cost of thermal management and decrease the system energy consumption. Furthermore, these improvements will enable new generations of high performance processors. This, in turn, will enhance the computational capabilities of future computing systems and enable progress in various fields. Finally, projects derived from this work will be integrated into courses contributing to the training of an engineering workforce for an energy efficient and sustainable society. RI: Small: Statistical Modeling of Dynamic Covariance Matrices Suitable models for dynamic covariance matrices can be extremely useful in several application domains, such as in text mining and topic modeling, where one can study the evolving correlation between topics; in financial data ranging from stock/bond returns to interest rates and currencies, where the paramount importance of tracking evolving covariances has been widely acknowledged; in environmental informatics to study trends in dynamic covariance among disparate variables from the atmosphere as well as the biosphere. In such domains, it is not sufficient to simply compute the sample covariance at each time step; the goal is to discover any trends there may be in the evolution of the covariance structure.

This project introduces and investigates a novel family of Dynamic Wishart Models (DWMs), which has the same graphical model structure as the Kalman filter, but tracks evolution of covariance matrices rather than state vectors. Similar to the use of multivariate Gaussians in Kalman filters, the models use the Wishart and inverse Wishart family of distributions on covariance matrices. Unlike Kalman filters, an analytic closed form filtering may not be possible in DWMs, but the models still have enough structure to allow efficient approximate inference algorithms. The project focuses on approximate inference for filtering, smoothing, and related problems in the context of DWMs; develop suitable numerically stable recursive updates in order to prevent numerical loss in positive definiteness; and investigate generalizations of DWMs including mixture models for tracking complex covariance dynamics.

The development of effective tracking algorithms for covariances will permit the modeling of dynamic systems where the states really represent the varying relationships between multiple entities. The key contribution of the research is in leveraging the existing literature of dynamic latent state vectors to create equally powerful methods for dynamic latent covariance matrices. Such a transformation will have direct impact on applications in text analysis and topic modeling, financial data analysis, social network analysis, environmental informatics, and several other domains, and will spawn new opportunities for bringing together researchers and students across these disciplines, thereby broadening participation in computer sciences. SHF: Small: The Chip Is the Network: Rethinking the Theoretical Foundations of Multicore Architecture Design Project ID: 0916752
Title: The Chip Is the Network: Rethinking the Theoretical Foundations of Multicore Architecture Design
PI Name: Radu Marculescu
Institution: Carnegie Mellon University

ABSTRACT
Recent advances in CMOS technology allow the integration of tens or hundreds of individually programmable processing elements, together with large amounts of dedicated memory, on the same system on chip (SoC). In such multiprocessor systems, individual processing nodes can communicate and coordinate via networks on chip (NoCs). Therefore, a major challenge is to determine the mathematical techniques for designing and optimizing such on chip networks in a rigorous manner. Traditional queuing and Markov chain approaches to buffer allocation are helpful to a certain extent, but capturing the traffic variability represents a major problem. Starting from these overarching ideas, this project introduces a new statistical physics approach for performance analysis in multiprocessor SoCs. More precisely, we develop a completely new mathematical description of network traffic using an analogy between a Bose gas and the information flow in the communication network. This new modeling paradigm where networks are seen as gases can be further used to develop efficient on chip buffer assignment algorithms.

The new design methodology enables the development of more efficient multiprocessor SoCs which have a dramatic impact on society via applications ranging from entertainment to gaming to security and to bio and gene engineering. More broadly, the results of this project impact significantly other research communities by improving the level of understanding of networking concepts needed to design and control complex systems. SHF: Small: Interleaving Constrained Parallel Runtime System for Tolerating Concurrency Bugs Future processor chips are expected to have hundreds or even thousands of processor cores. To take advantage of this massive computing power, programmers need to parallelize their applications. Parallel programming, however, is notoriously difficult. Almost all the production concurrent software systems used today contain bugs costing billions of dollars. To address this challenge, this research is developing parallel runtime mechanisms that could make it possible for even buggy software to run correctly in a production system.

The fundamental problem with the current parallel programming models is that they expose an unbounded number of thread interleavings to the parallel runtime system, and a majority of the interleavings in a production system remain untested. This research is exploring two directions to avoid incorrect interleavings from manifesting in a production run. The first approach uses a sampling based low overhead data race detector for detecting incorrect interleavings, which are then avoided. The second approach constrains production run thread interleavings to a set of tested interleavings, which could provide comprehensive immunity against most types of concurrency bugs. Software tools developed as part of this research will help software developers and researchers. Students will also use these productivity tools in their course projects. AF:Small:Collaborative Research:Studies in nonuniformity, completeness, and reachability Computational complexity theory classifies computational problems into various complexity classes based on the amount of resources needed to solve them. This classification is done by measuring various resources such as time, space, nonuniformity, nondeterminism, and randomness. A better understanding of the relationships among these various resources shed light on the computational difficulty of the problems that are encountered in practice.

This project explores several central questions regarding nonuniformity, complete problems, and space bounded computations. This project attempts to discover improved upper bounds for problems with high circuit complexity. Regarding complete sets, non relativizing properties of complete sets will be explored. Space bounded computations will be investigated in the context of planar graph reachability problems.

This project addresses several basic questions in computational complexity theory. The results from this project will further our understanding of computational resources such as nonuniformity, nondeterminism, and space. Research results will be published in peer reviewed journals and will be presented at national and international conferences, thus enabling broad dissemination of the the results to enhance scientific understanding. New courses will be created and taught along the themes of this project, thus integrating teaching and research. The project supports various human resource development activities such as supporting and mentoring graduate students and inviting visitors. NeTS: Small: Toward High Performance WLANs: Bridging the Physical Layer Divide This award is funded under the American Recovery and Reinvestment Act
of 2009 (Public Law 111 5).

This project architects high performance wireless local area networks (WLANs) by understanding the impact of physical layer attributes on performance and incorporating them in the design and control of next generation WLANs. The research is comprised of three parts. The first part develops models of spatial diversity, the dominant physical layer feature, that help understand and predict the performance of infrastructure mode WLANs. The second part integrates spatial diversity with cross layer protocol analysis that allows evaluation of the influence of physical layer attributes on both lower and higher layer protocols. The third part investigates new network controls that harness opportunities provided by spatial diversity that help mitigate, and in some cases, transcend their detrimental performance effect including unfairness and throughput degradation. The control dimension extends to large scale WLANs covering city blocks and campuses that inject complex spatial coupling. The project employs a combination of simulation, experimentation, and analysis to achieve its goals. The broader impact of this project lies in narrowing the performance gap between wireless and wired networks, which facilitates ubiquitous high speed Internet access. The project also helps educate students in the fundamentals and intricacies of wireless communication. The results from the project will be disseminated at conferences, seminars, and through the project web site where data and tools are made publicly available. SHF: Small: Specification, Verification, and Semantics of Higher Order and Concurrent Software This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

To extend the scale and generality of separation logic and grainless semantics
to become applicable to the specification and verification of more complex and
varied software systems, the following research initiatives are being pursued:

Higher Order Design Patterns Using higher order separation logic to
specify design patterns in object oriented programing, and to verify their
implementation.

Extension to a Higher Order Programming Language extending separation
logic to an Algol like language with a heap, including procedures that
permit global information hiding.

Grainless Semantics devising a new form of grainless semantics that gives
a more abstract and concise description of shared variable concurrency by
avoiding any commitment to a default level of atomicity.

Implementation of Logics and Semantics Using the Isabelle/HOL system to
implement the machine aided construction of proofs in separation logic in a
way that ensures both the validity of the proofs and the soundness of the
logic and its extensions.

The overall goal is to substantially increase the variety of programs that can
be verified by separation logic, and to facilitate soundness arguments for this
and other logics for shared variable concurrency. RI: Small: Novel structured regression approaches to high dimensional motion analysis The ability to estimate motion of objects from video is a fundamental scientific problem that arises in many tasks: finding out how the human body moves, tracking vehicles movements on a highway or the motility of schools of fish. Despite many advancements the problem remains hard because of sudden, often highly nonlinear changes and the high dimensionality of the object s configuration spaces. Much prior work has focused on building complex physics based models, in an analysis by synthesis paradigm dominated by expert s domain knowledge. When such knowledge is lacking, the resulting models may produce inaccurate predictions.

To address these issues, this project investigates a new paradigm of using limited amounts of carefully collected data to learn direct predictive models of high dimensional motion. We approach the problem as that of the structured regression, a novel generalization of traditional statistical methods that specifically exploits the spatio temporal structure of the data to avoid the need for analysis by synthesis . This research will result in a set of robust techniques and computational algorithms that support this new modeling framework.

The tools and techniques developed here will have wide applicability in many areas of technology and industry that rely on design of accurate prediction models in complex space time domains, leading to more general and sustainable forecasting solutions. Through engagement of graduate and undergraduate students in key research activities, the project also provides advanced technical training vital for success of a new generation of computer scientists. SHF: Small: High level Programming Models and Frameworks for GPGPU based Computing This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Graphics Processing Units (GPUs) have emerged as a promising alternative in
the transition of the computing industry to mainstream parallel computing.
Enabling applications to benefit from their potential requires that GPU programming be made accessible to the average programmer. This research focuses on the challenges making GPU programming easier through new high level programming models, and enabling efficient GPU execution through compilation frameworks for these models.
Two complementary GPU programming models are proposed OpenMP, which is widely used for shared memory parallel programming, and Parallel Operator Data Flow Graphs (PO DFGs), which naturally represent algorithms in a wide range of current and emerging application domains. Various optimization techniques are developed for programs written to these models, including partitioning the program between the host CPU and GPUs, stream optimizations that render the program s memory access characteristics to be more amenable to the GPU s memory system, minimizing data transfer between the host and GPU memory, and GPU architecture specific optimizations. The research contributes to the evolution of GPGPU programming from manual ports of applications using low level APIs, to the use of high level parallel programming models. SHF: Small: Dynamic Power Redistribution in Failure Prone CMPs Future multi core microprocessors will be capable of deconfiguring faulty units in order to permit continued operation in the presence of wear out failures. However, the unforeseen downside is pipeline imbalance in which other portions of the pipeline are now overprovisioned with respect to the deconfigured functionality. Such an imbalance leads to sub optimal chip wide power provisioning, since power is now allocated to pipeline functions that no longer provide the benefit they did with a fully functioning chip.

This research proposes to dynamically redistribute the chip power under pipeline imbalances that arise from deconfiguring faulty units. Through rebalancing achieved by temporary, symbiotic deconfiguration of additional functionality within the degraded core power is harnessed for use elsewhere on the chip. This additional power is dynamically transferred to portions of the multi core chip that can realize a performance boost from turning on previously dormant microarchitectural features. The technical deliverables of this project will be: (1) a novel resilient multi core system architecture including dynamic power redistribution management algorithms that achieves much higher performance than one that is oblivious to pipeline imbalances; and (2) detailed simulations that quantify this performance advantage for various multi core workloads.

The broader impacts of this project relate to integrated research and education, enhanced infrastructure for research, broad dissemination of results, and potential societal impact. Furthermore, the PI will recruit women and underrepresented minority students to work on the project. SHF: Small: Achieving IC Quality through On Going Diagnosis and Customized Test Proposal ID: 0916828
PI name: Blanton Shawn
Institution: Carnegie Mellon University
Title: Achieving IC Quality through On Going Diagnosis and Customized Test

ABSTRACT
The main objective of computer chip testing has and continues to be the separation of good chips from bad ones (i.e., ones that do not meet the desired operational characteristics). Test is now however being expanded as a value added endeavor. In this project, we are data mining test data in order to continuously monitor chip quality. We propose to use diagnosis extracted models of chip failures along with a new technique for estimating chip quality. Both are incorporated in an on line, quality monitoring methodology that ensures a desired level of quality by changing the actual tests applied to a computer chip to better match the characteristics of currently failing chips.

This approach to quality is dynamic in nature and is a radical change from the typical approach. Without exception, each chip manufacturer (Intel, IBM, etc.) assumes that any type of defect can occur anywhere within their chip which means that each manufactured instance has to be thoroughly tested at considerable expense. This is akin to prescribing drugs for all possible diseases/ailments for every patient without performing one diagnostic examination. Opposed to the traditional approach, this proposed work instead diagnoses chips that have failed in the past to determine what ?diseases? (i.e., defects) actually are occurring within the fabricated ICs. The ?prescriptions? (i.e., the tests applied to the chips) can therefore be changed and/or minimized to match the diseases found instead of over testing as is done now, leading to improved chip quality at minimal cost. RI: Small: Modeling and Recognition of Landmarks and Urban Environments The goal of this project is to design a scalable and robust system for modeling and representing the spatiotemporal and semantic structure of large collections of partially geo referenced imagery. Specifically, the project is aimed at Internet photo collections of images of famous landmarks and cities. The functionalities of the system include 3D reconstruction, browsing, summarization, location recognition, and scene segmentation. In addition, the system incorporates human created annotations such as text and geo tags, models scene illumination conditions, and supports incremental model updating using an incoming stream of images. This system is designed to take advantage of the redundancy inherent in community photo collections to achieve levels of robustness and scalability not attainable by existing geometric modeling approaches. The key technical innovation of the project is a novel data structure, the iconic scene graph that efficiently and compactly captures the perceptual, geometric, and semantic relationships between images in the collection.

The key methodological insight of this project is that successful representation and recognition of landmarks requires the integration of statistical recognition and geometric reconstruction approaches. The project incorporates statistical inference into all components of the landmark modeling system, and includes a significant layer of high level semantic functionality that is implemented using recognition techniques.

Potential applications with societal impact include virtual tourism and navigation, security and surveillance, cultural heritage preservation, immersive environments and computer games, and movie special effects. Datasets and code produced in the course of the project will be made publicly available. The project includes a significant education component through undergraduate and graduate course development. SHF:Small: Next Generation Test Compression Technology PROPOSAL NO: 0916837
INSTITUTION: University of Texas at Austin
PRINCIPAL INVESTIGATOR: Touba, Nur
TITLE: Next Generation Test Compression Technology


Abstract

Testing integrated circuits (ICs) requires storing large amounts of test data on a tester and transferring it to/from the chip under test. The bandwidth between the tester and chip is very limited due to limited pins and tester channels. Test data volume continues to grow dramatically with increasingly dense system on chips (SOCs) and three dimensional ICs as well as the need for additional tests to target defects in nanometer designs. A major development in the field over the past decade has been the emergence of test compression technology which stores test data on the tester in compressed form and decompresses it on chip. The commercialization of this technology has helped immensely in keeping up with rising test data volume. However, going forward, there is a need for a next generation of test compression technology that can provide significantly greater compression to handle the larger designs of the future. This research will develop new theory, concepts, and architectures that are fundamentally different from existing commercial technology and have the potential for providing an order of magnitude or more improvement for test stimulus compression as well as output response compaction.

Society increasingly relies on correct and dependable operation of electronic devices. The impact of this research will be to develop new technology to keep test costs down and make it economical to fit in more tests to improve product quality. This will be critical as the manufacturing process becomes increasingly difficult to control at smaller geometries. Participation of undergraduates, women, and minorities will be actively encouraged. III: Small:Using Data Mining and Recommender Systems to Facilitate Large Scale Requirements Processes Problems related to requirements definitions account for numerous project failures and translate into significant amounts of wasted funds. In many cases, these problems originate from inadequacies in the human intensive task of eliciting stakeholders needs, and the subsequent problems of transforming them into a set of clearly articulated and prioritized requirements. These problems are particularly evident in very large projects such as the FBI Virtual Case File or NASA s Space Station, in which knowledge is dispersed across thousands of different stakeholders. On one hand, it is desirable to include as many people as possible in the elicitation and prioritization process, but on the other hand this can quickly lead to a rather chaotic overload of information and opinions. The work proposed under this grant will develop a new framework that utilizes data mining and recommender systems techniques to process and analyze high volumes of unstructured data in order to facilitate large scale and broadly inclusive requirements processes. The proposal is based on the observation that the requirements elicitation process of many large scaled industrial and governmental projects is inherently data driven, and could therefore benefit from computer supported tools based on data mining and user modeling techniques.

INTELLECTUAL MERIT
The proposed research will lead to a robust requirements elicitation framework and an associated library of tools which can be used to augment the functionality of wikis, forums, and specialized management tools used in the requirements domain. Specifically, this research will enhance requirements clustering techniques by incorporating prior knowledge and user derived constraints. A contextualized recommender system will be designed to facilitate appropriate placement of stakeholders into requirements discussion forums generated in the clustering phase.

BROADER IMPACT
The proposed work has potential for broad impact across organizations that develop stakeholder intensive systems. Technology transfer can be expected due to collaborations with organizations such as Siemens and Google planned as an integral part of this research. Educational materials will be developed specifically for requirements engineering and recommender systems courses, and will be broadly disseminated.

Key Words: Recommender systems; Data mining; Clustering; Requirements engineering; Requirements elicitation. TC: Small: Minimalist Hardware Trojans through Malicious Side Channels In order to provide system security, hardware modules which function as trust anchors are used in an ever increasing number of devices. The majority of laptops and PCs are now equipped with Trusted Platform Modules (TPMs), and a large number of pervasive computing systems such as smart cards, electronic passports or high speed routers make use of hardware for cryptographic algorithms and key storage. In almost all such applications the security of the entire system hinges on the assumption that the hardware modules are trustworthy. Recently, due to the increasing use of potentially untrusted semiconductor foundries, the threat of maliciously manipulated hardware has been raised, Since hardware manipulations, including hardware Trojans, are difficult to detect and, perhaps more importantly, even harder to repair, they form a very serious threat to system security for today s and future applications.

The standard approach to Trojan hardware consists in adding extra logic to a given IC design which weakens the system. The main drawback of this approach, from an attacker s perspective, is that extra function blocks can potentially be detected through a host of techniques, including, e.g., optical inspections at different layers of the design, or power and EM fingerprinting. Our malicious circuit manipulations are orders of magnitude more subtle than previously known Trojans, but can nevertheless totally compromise secure hardware blocks by leaking cryptographic keys. The core idea is to create malicious side channels, in particular power supply channels, through small modifications of circuit elements, e.g., at the transistor level. We will refer to these covert channels as Trojan side channels (TSC). The core parts of the research are modeling of the assumptions, development of channels and modulations schemes, their realization on the circuit level, and proof of concept implementations.

In addition to posing a threat to system security, Trojan side channels can also be used constructively. For instance, they have applications in anti counterfeiting: illegal copies of ICs with the same functional behavior will not leak the same side channel ID and can thus easily be detected. Also, TSC could be used for conveying internal status information about a circuit, increasing the testability of a circuits. Moreover, because TSC can be viewed as a form of physically encryption one can imagine other cryptographic protocols and applications using TSC as primitives. TC: Small: Dynamic Early Filtering of Botnet Garbage Traffic Currently in the Internet there is an increasing number of unwanted, unsolicited garbage packets mainly generated by botnets, which can launch Distributed Denial of Service attacks, worm attacks, and spam. These garbage packets are allowed to traverse the Internet to cause severe traffic burdens, waste communication resources, and disrupt the Internet?s normal functions. Such packets need to be discarded as close to their sources as possible to increase the availability and reliability of the Internet.

This project aims to establish a comprehensive and sustainable architecture that coordinates the routers in the Internet to achieve early filtering of botnet garbage packets in Internet traffic. The architecture comprises four major components: rule generation component, rule dissemination component, rule management component, and rule security component. The objective is to investigate and quantify the tradeoff between the saved bandwidth originally consumed by the garbage traffic and the throughput slowdown introduced by the routers extra filtering overhead, and find optimal solutions under the tradeoff function. The evaluation plan will use benchmarks developed under various traffic traces and network topologies to evaluate the performance of the developed algorithms and technologies, and derive insights on how far and wide the filtering rules should be disseminated and installed under different attack scenarios in order to optimize the performance.

Completion of the project will create techniques and software that improve the security and capacity of the Internet overall. Additionally, early filtering of garbage traffic will limit the damages caused by large scale botnet attacks, reduce operational costs for ISPs, enhance the performance of many online services and applications, and increase the reliability of critical national infrastructures. CSR: Small: Infrastructure free Human Context Awareness with a Wearable Sensing and Computing System This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The major objective of this project is to develop the fundamental theoretical framework and algorithms that realize human context awareness in an infrastructure free fashion and validate them through physical experiments using a body sensor network. This new human context awareness approach will open up many new research opportunities in pervasive computing, as well as human computer interaction and human robot interaction. This project will have many societal impacts. First, the research results can be applied to track and monitor emergency responders, therefore significantly improving the operation efficiency and personnel safety. Second, this project will benefit health research and health care practices. Being able to monitor human s activity and location in a daily setting will enable researchers to study many health problems related to human behaviors. When used by physicians to monitor their patients or elderly people, it can greatly improve the accuracy of disease diagnosis and human health assessment. The proposed research and education integration activities will help boost the enrollment and retention rate in the ECEN program at Oklahoma State University. The proposed outreach activities will stimulate prospective and current college students, especially Native American, female students, to pursue degrees or careers in science and engineering. RI Small: Exploiting Comparable Corpora for Machine Translation (CC4MT) Parallel corpora, i.e. texts that are translations of each other, are an important resource for many natural language processing tasks, and especially for building data driven machine translation systems. Unfortunately, for the majority of languages, parallel corpora are virtually non existent. To be able to develop machine translation systems for those languages, we need to be able to learn from non parallel corpora. Comparable corpora ? i.e. documents covering at least partially the same content ? are available in far larger quantities and can be easily collected on the Web. Examples include news published in many languages by Voice of America or BBC, and the multi lingual Wikipedia.
To make best use of comparable corpora it is not sufficient to extract sentence pairs, which are sufficiently parallel, thereby building a parallel corpus and then using proven training procedures. Rather, new techniques are required to find sub sentential translation equivalences in non parallel sentences. To extract phrase pairs from comparable corpora requires a cascaded approach:
find comparable documents using, for example, cross lingual information retrieval techniques;
detect promising sentence pairs, i.e. those, which may contain translational equivalences;
apply robust phrase alignment techniques to detect phrase translation pairs within non parallel sentence pairs;
The main focus of the project lies on this third step: developing novel alignment algorithms, which do not rely on aligning all words within the sentences, as traditional word alignment algorithms do, but can separate parallel from non parallel regions.
The long term benefit of this work will be that machine translation technology can be applied to those languages, for which so far no translation systems are available, due to the lack of the language resources required by current technology. This will enable communication across language barriers, esp. in critical situations like medical assistance or disaster relieve. RI: Small: Learning Based Systems for Single Image Photometric Reconstruction This project focuses on developing algorithms and datasets that can transform photometric reconstruction systems from hand designed systems into learning based systems that are optimized on real world data.
Photometric reconstruction systems derive cues from the perceived intensity of different locations on a surface. Shape from shading, where the surface is assumed to have a diffuse reflectance, is a well known example of photometric reconstruction. This project produces the datasets and methods necessary to use machine learning techniques to build models for photometric reconstruction.

This learning based approach enables systems to be optimized on real world data so that they produce the most accurate results possible. In addition, this learning based approach enables the development of more sophisticated methods with more parameters than typically used in hand designed systems. The ability to find optimal parameters in an automated fashion can not only improve existing approaches, such as by incorporating image data more effectively, but can also enable the development of algorithms that push the boundaries of current systems. In particular, algorithms are developed for estimating the shape of objects without knowing the illumination or even trying to explicitly model it.

The power of the learning approach cannot be realized without data for training and testing. A major task in this work is the construction of a database of images and ground truth 3D reconstructions of the objects in the images. The 3D models can be found using an example based photometric stereo technique. RI: Small: Cooperative Coevolutionary Design and Multiagent Systems Cooperative coevolution is a potent approach to doing large scale stochastic optimization. The unsolved game theoretic challenges inherent in this computational method are complex and of significant interest to the evolutionary computation community. This project is advancing the state of the art in coevolution and is applying it to significantly larger problems than commonly found in the literature. These challenges, and their solution, have potentially transformative impact on other co adaptive environments such as multiagent reinforcement learning, estimation of distribution algorithms, agent modeling, and swarm robotics. Coevolution has strong applicability to fields that use multiagent system models, including multirobotics, biology, economics, land use, and political science. Better models in these fields can positively affect society, policy, homeland security, and the environment. NeTs: An open architecture for the evolutionary design of routing protocols Traditional routing systems have been designed by network engineers based on complex collections of objectives, policies, principles and past experiences. Due to limited human experience and capability, this manual design process severely impedes the design process of routing. The objective of this project is to bring a revolutionary change to this design process by building a highly flexible architecture, called Orchestra, for the automatic assembling and testing of a great variety of routing designs. Orchestra stores a large set of reusable genes . Each gene is a small piece of computer code that implements a particular design for a small component of a routing system. The correctness of the genes and their mutual compatibility are automatically verified. Orchestra assembles various routing systems from verified genes and then tests them in both simulation and real environment. Based on the performance of the assembled protocols, Orchestra uses evolutionary algorithms to switch and tune designs of routing components to eventually identify the best design for a network setting.

Orchestra will greatly ease a network engineer s burden of implementing and evaluating an entire routing system. It can efficiently explore a much larger design space for routing systems than any single network engineer can. New areas for routing designs that are not explored by humans can be automatically discovered by Orchestra. The large collection of component designs in Orchestra will also provide a common platform for comparing and evaluating different design choices as well as serving education purposes. CSR: Small: Operator Proof Systems Management This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
Many studies have shown that human mistakes are an important source of system failures. Further, repairing mistakes is often time consuming, leading to high unavailability. In this project, we will explore a novel approach to dealing with human mistakes called operator proof systems management. In an operator proof system, an omnipresent management infrastructure will enable the system to defend itself against operator mistakes. The infrastructure will constantly monitor operator actions and the system state to decide when and how the system should defend itself. Possible defensive measures include blocking operator actions that could lead to a mistake and/or limiting operator access to prevent mistakes from spreading throughout the system. Blocks are later lifted if the system can test the correctness of the operator actions.
To explore our ideas, we will design and implement two very different prototype operator proof systems: an Internet service and an enterprise system. We will explore the design space and evaluate the overall approach by running a large set of experiments, where volunteer operators of different levels of experience are asked to perform a variety of tasks on the prototype systems.
Broader impacts. Our research will provide a concrete step toward the realization of a model where large computer systems can be operated at lower cost by less skilled individuals. Our investigation will also expose a large number of students (acting as volunteer operators) to system management issues and our proposed solutions. CIF: Small: Multichannel Signal Processing for Dense Optical Communication Networks Abstract:

The dramatic increase in throughput demands from backbone transport data networks has propelled the development of all optical wavelength division multiplexed (WDM) networks. As the capacity demands on these systems increase, the physical layer degradation becomes so severe that sophisticated signal processing techniques are necessary to maintain the quality of service. Channel nonlinearity differentiates multichannel fiber optic systems from conventional wireline and wireless systems. Consequently, novel signal processing and communication theoretic approaches are required to design and analyze the channel; this research project addresses this need. As a consequence, WDM system designs are improved and the data throughput available to society though these networks is substantially increased.

This research develops a discrete time wavelength nonlinear model for the WDM system and uses this model to design powerful signal processing techniques. Up to now, the WDM channel has been considered as a set of parallel channels, ignoring the cross channel effects or modeling them simply as noise. Motivated by techniques that have been so successful in wireless communications, such as multiuser detection, MIMO processing, multichannel precoding, etc., this research develops algorithms to apply to the WDM fiber channel that can produce substantial capacity gains. A two dimensional discrete time polynomial model for a WDM system is formulated to account for intra channel and inter channel linear and nonlinear effects. Multichannel processing algorithms across time and wavelength for interference mitigation are designed and evaluated. Since nonlinear interference limits the performance of networks, constrained coding to diminish this interference is used to trade capacity for performance. In all optical networks, crosstalk emanating from other lightpaths can limit performance. Employing idle lightpaths judiciously provides multiuser coding and path diversity (redundancy and memory) to the entire network.

Level of effort statement:

At the recommended level of support, the PI will make every attempt to meet the original scope and level of effort of the project. NeTS: Small: Collaborative Research: Holistic and Integrated Design of Layered Optical Networks The dramatic increase in throughput demands on transport systems has propelled the development of all optical networks. These networks can provide tremendous capacity when they are designed with their own limitations in mind, such as coarse wavelength granularity and physical impairments. In this research we consider the holistic design of optical networks that include the interdependence of three network layers: the traffic grooming layer, the lightpath management layer, and the physical fiber layer.

The network is first viewed from the top down, where sub wavelength circuit requests arrive with specific quality of service requirements. Current traffic grooming approaches are altered to incorporate their dependence on the lightpath management and physical layer constraints.
The system is then examined from the bottom up, so that the quality of transmission and efficiency of resource utilization can be optimized as the higher layer protocols evolve. Total capacity is measured from an information theoretic view point and system optimization uses ideas from game theory.

The results of the research will be practical algorithms for improved capacity and survivability of future optical networks as well as providing a quantitative proof of their superiority. The enhancement of network capability will help satisfy our society?s ever increasing need for information. It encourages the development of applications that require significant bandwidth. It also stimulates cross fertilization of ideas from the two fields of networking and communications. The algorithms and software will be made publicly available via a web site. The research enhances the education of the diverse group of graduate and undergraduate students participating in it. SHF:Small:Exploring the Synergy between Software Design and Organizational Structure Many successful large scale software systems share a fundamental characteristic: their modular structures enable system wide advances though distributed and parallelized improvement of modules. However, merely breaking software into modules, without assessing the interplay between a design and the organization that must instantiate it, does not always ensure that parallelized, module wise evolution is effective. In particular, mismatches between design and organizational structures can result in expensive inter team communication costs, exacerbated by barriers such as differing time zones, languages and cultures. This research aims to formally express and quantitatively assess the key characteristics of software structures that allow for system wide evolution through distributed module wise contributions, and to account for the relationship between design structure and organizational structure, as it impacts software quality, productivity, and survival. The work will explore a computable socio technical model, associated metrics and automated analysis techniques to improve the conduct of software development. The approach will allow designers to assess and manipulate software designs at early development stages so that modules can be defined and implemented by independent teams, shortening development time, facilitating changes, and minimizing coordination costs. The results will be demonstrated on large software systems, working with industrial partners who wish to understand the impact of these techniques. CIF: Small: Large Scale Software Dissemination in Stochastic Wireless Networks Commercial providers are increasingly permitting third party developers to write and implement their own software applications on wireless devices, ranging from sensors to 3G cellular phones. As the number of applications and users grow, reliable software dissemination is quickly emerging as a key enabling technology, providing fundamental reprogramming services, such as software download, updates and security patching.


Intellectual Merit
This project aims to develop analytical foundations for efficient software dissemination in loss prone wireless networks, measured in terms of delay and communication/computational speed. Planned research will proceed along four thrusts: 1) PERFORMANCE LIMITS: mathematically formalizing the problem of software dissemination in multi hop wireless networks using stochastic shortest path optimization based on the theory of Markov Decision Processes; 2) LARGE SCALE ASYMPTOTICS: analyzing performance at high node densities using extreme value theory and comparing state of the art technologies, including rateless coding, packet level channel hopping, and physical layer cooperation; 3) ACK LESS PROTOCOLS: eliminating control traffic (e.g., ACKnowledgments), with high probability, using a combination of extreme value theory and shifted rateless codes; 4) IMPLEMENTATION: implementing theoretically based software dissemination protocols on sensor motes and, subsequently, on Android capable smartphones.



Broader Impact

This work promises a broad impact to various societal needs. On an education level, open cell phone programming expertise will be incorporated into innovative class assignments and labs taught by the PIs, including Software Engineering, Algorithms, and Networking. On a commercial side, the research identifies and provides directions for fundamental issues that will face private enterprises attempting to capitalize on emerging smartphone capabilities. The PIs will also expedite research transfer through liaisons with local and international industrial partners. Finally, the project will establish theoretical connections between disconnected fields, most notably bringing tools primarily used in civil and financial engineering into computing and communication. SHF: Small: IMUnit: Improved Multithreaded Unit Testing CCF 0916893
SHF:Small: IMUnit: Improved Multithreaded Unit Testing
PIs: Darko Marinov and Grigore Rosu

Computing is going through a paradigm shift from a mostly sequential worldview to a mostly parallel worldview. The currently dominant programming model for parallel code is that of shared data, where multiple threads of computation communicate by accessing shared data objects. Unfortunately, developing and testing multithreaded code is very hard. To significantly improve testing of multithreaded code, this project develops a set of new techniques and tools for multithreaded tests. A multithreaded test is a piece of code that creates and executes two or more threads. Executing a test follows some schedule/interleaving of the multiple threads. The key of the new approach is to allow the explicit specification of a set of relevant schedules for each test, while traditional tests implicitly specify all possible schedules. This project addresses three important challenges for multithreaded tests: (1) How to describe a set of schedules and which schedules from a given set to explore? (2) How to automatically generate multithreaded tests, especially schedules, for given code? (3) How to select and prioritize rerunning of the multithreaded tests when the code changes? Improved unit testing of multithreaded code has the potential to substantially increase the quality of developed code. SHF: SMALL: Ant: Automatic and Manual Debugging Support for Massively Parallel Programs This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).
The research funded by this award targets the difficult problem of how to debug programs running on large parallel systems. The state
of hundreds to hundreds of thousands of parallel tasks that form a single
computation are too complicated for a programmer to usefully analyze.
This project will develop tools to find similarities between the state of
different processes, simplifying the task of the programmer. The challenge
in finding these similar tasks is to do it efficiently, without imposing an
overhead so hight that the tool is useless. A naive implementation would
compare the state of all processors against one another, and would introduce
overheads increasing as the square of the number of processors. Our approach
will successively refine sets of similar processes, will use key attributes
of program behavior (e.g. communication patterns) to perform this grouping.
We will also investigate the use these groups of similar processes to allow
invariance and statistically based techniques developed for sequential
programs (such as value and PC invariance) to be effectively adapted to
parallel programs. Because these techniques look for rarely occurring program
activities, applying them to disparate processes together will introduce
noise into the analysis, severely diminishing their accuracy. The use of
our grouping strategy will allow effective parallelization of the
techniques, allowing them to be applied with significantly less overhead
than when used with sequential applications. CSR: Small: Binary rewriting without relocation information This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

This project focuses on the development of a new binary rewriter that can be use to statically transform binary code that does not have relocation information and to do so without the overhead of dynamic binary rewriting.

Binary rewriters are pieces of software that accept a binary executable program as input, and produce an improved executable as output. The output executable usually has the same functionality as the input, but is improved in one or more metrics, such as run time, energy use, memory use, security, or reliability. Many optimizations for binary rewriting have been proposed to improve all these metrics. However, existing binary rewriters have a severe limitation ? they are unable to rewrite binaries that have no relocation information. Unfortunately, linkers typically discard relocation information, and hence virtually all commercial binaries lack relocation information. Consequently, they cannot be rewritten by existing rewriters.

In this project, the PI is building a binary rewriting infrastructure that can rewrite binaries that do not contain relocation information. This will have the following broader impacts: (i) the ability for anyone to rewrite any binary to improve its performance, security, or memory consumption, or to monitor its resource consumption, is a powerful new ability that could unleash innovation and engender a new class of commercial applications; (ii) the resulting new commercial applications will be a net gain for the nation?s economy; (iii) the improvements in those applications will boost the productivity and security of their users; and (iv) a strong educational program with instructional and outreach components. CSR: Small: Feedback Controlled Management of Virtualized Resources for Predictable Escience The use of virtual machines for Escience has been advocated both within the enterprise to replace aging machines and as the underlying technology of cloud computing whereby scientific researchers can ?rent? servers on demand. However, both scenarios can lead to inadequate performance. Within the enterprise, with incorrect planning or under unexpected heavy or even moderate load, there might not be enough physical capacity for every virtual machine to achieve reasonable performance. In cloud computing based scenarios, the ?renters? are largely subject to the informal service promises of the cloud provider based on a granularity that can be too coarse or at the wrong level of abstraction. This project pursues a novel unified framework to ensure predictable Escience based on these two dominant emerging uses of virtualized resources. The foundation of the approach is to wrap an Escience application in a performance container framework and dynamically regulate the application?s performance through the application of formal feedback control theory. The application?s progress is monitored and ensured such that the job meets its performance goals (e.g., deadline) without requiring exclusive access to physical resources even in the presence of a wide class of unexpected disturbances. This project extends this foundation and early results in three important dimensions: creating support for non specialists to use the framework; implementing these techniques in Eucalyptus, one of the major open source cloud computing frameworks; and applying the techniques to ?Software as a Service? (SaaS), in which applications in the cloud are regulated to provide predictable performance. CSR:Small:Multi Layer Support for Virtualizing Heterogeneous Architectures This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Heterogeneous architectures with specialized coprocessors are gaining momentum due to compelling performance and energy efficiency benefits. Heterogeneity introduces two main challenges: how to select the best coprocessor for a particular task, and how to map a task onto a particular accelerator. Programmers should be able to focus on algorithm and software design without worrying about hardware details or sacrificing portability, so rewriting code for every new coprocessor is untenable. While it is possible to schedule and map tasks at compile time, competing tasks and varying workload characteristics mean that effective scheduling must happen at runtime.

The solution proposed with this project is to virtualize the underlying system heterogeneity, which frees programmers from the burden of considering this heterogeneity in their implementation. At runtime, the system software layers schedule tasks to the most effective targets and map the programmer s virtualized task description to that target s architecture. The overarching goal of this project is to design an end to end solution to the challenge of heterogeneous code generation and optimization. This project develops the required capabilities. The intellectual merit of this research lies in the novel advances required in the compiler, runtime, and operating system. This work in turn provides broad impact by increasing programmer productivity and providing software tools that enable future research and software development. This work will also train graduate and undergraduate students in cutting edge computer systems concepts and design skills, and develop new educational and outreach materials. This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5). RI: Small: Robust Automatic Speech Recognition in Highly Reverberant Environments Speech processing systems, including automatic speech recognition and speaker identification, are the key enabling technologies that permit natural interaction between humans and intelligent machines such as humanoid robots, automated information providers, and similar devices. For example, it is now commonplace to encounter speech based intelligent agents handle at least the initial part of a query in many types of call center applications. While we have made great progress over the past two decades in overcoming the effects of additive noise in many practical environments, the failure to develop techniques that overcome the effects of reverberation in homes, classrooms, and public spaces is the major reason why automated speech technology remains unavailable for general use in these venues. Reverberation remains one the most difficult unsolved problems in speech recognition in open acoustical environments.

This project develops novel approaches that combat the effects of reverberation through two complementary perspectives: contemporary knowledge of human auditory processing and state of the art application of statistical source separation techniques that build on techniques in image and music processing. The synergistic development of these approaches is expected to provide substantially improved speech recognition and speaker identification accuracy in reverberant acoustical environments, along with a principled structure that enables us to understand on a much deeper level why the solutions to these problems are effective. This work is expected to have an enormous impact in extending the applicability of automatic recognition of natural and casual speech to highly reverberant environments. CIF: Small: New Approaches to the Design and Analysis of Graphical Models for Linear Codes and Secret Sharing Schemes NSF Proposal 0916919

New approaches to the design and analysis of graphical models for linear codes and secret sharing schemes

Abstract

Error correcting coding enables one to design reliable systems of transmission and storage of information and is used universally for sending packets over the web, in writing data on CD s and flash memory devices, and other similar means of modern communication. A very efficient method of encoding information for error protection is the so called iterative decoding, which assumes that every binary digit of transmission is recovered based on its realiblity and the realibility of a few other, carefully selected bits of the encoded message. This method of error correction is analyzed based on the representation of the encoding as a graph in the plane in which recovery from errors proceeds by successive exchange of information between the nodes of the graph in an iterative procedure performed in a number of rounds. One of the main goals of this research is to reduce complexity (the number of rounds) needed for reliable recovery of the transmission from errors in the communication medium.

Graphical models of linear codes have so far been restricted to trellises, i.e., cycle free graphs, and graphs with exactly one cycle (tail biting trellises). This research studies complexity of realization of codes and iterative decoding algorithms on connected graphs with cycles, deriving complexity estimates from the tree decomposition of graphs. One of the goals of this research is to find methods of constructing low complexity realizations of codes for such well known code families as Reed Muller and Reed Solomon codes, and explore the optimality gap of these representations. Methods of matroid theory used in the study of graphical models will also be explored in the analysis of access structures of secret sharing schemes and secure multi party computation protocols. RI: Small: Simplifying Text for Individual Reading Needs A surprisingly large number of Americans read below their grade level, either because of limited education or because their native language is not English. Low reading levels impact a child?s progress in school and an adult?s job opportunities as well as limiting information access.
This project aims to improve access by developing new language processing technology for selecting and transforming text to obtain material at lower reading levels, extending current paraphrasing work that focuses on summarization as compression to include explanatory expansions. In addition, the goal is to develop adaptive models that can be tuned to a specific domain and an individual s needs. The approach involves analyzing corpora of comparable text collected from the web, developing models of paraphrasing aimed at generating simplified English, developing a discourse sensitive clause selection method for expanding or omitting details, and exploring representations of language that facilitate domain and user adaptation. The language processing contributions of this work include development of text resources to support language technology in education applications, new representations of reading difficulty, and advances in automatic methods of paraphrasing. The broader impact of this project includes making information more accessible to people with limited English reading proficiency. In addition, students working on the project will have the opportunity to interact with teachers from a local school so as to better understand the impact of their work and guide their approach, and their work will be showcased in University of Washington diversity oriented outreach programs. CIF: Small:Sparse and Geometric Representations of Images and Multidimensional Signals Project:
Sparse and Geometric Representations of Images and Multidimensional Signals


Abstract:

One of the most fundamental characteristics of natural images and multidimensional signals is the presence of geometric regularities due to smooth boundaries between smooth regions. Current representations for this class of signals such as curvelets, contourlets, shearlets, and surfacelets, are constructed in the frequency domain, which lead to basis functions with large spatial support and Gibbs oscillations.


This research develops new sparse representations for multidimensional signals with geometric regularities. These representations allow successive approximation from coarse to fine and will be digital friendly.
The investigators focus on spatial domain constructions based on true multidimensional and geometric lifting schemes that would lead to a new generation of geometric wavelets. In addition, geometric tiling dictionaries with low coherence are explored. The research aims for a precise connection between the continuous domain, where geometric regularity is characterized, and the discrete domain, where input signals and transforms are defined on sampling grids. Finally, the research develops new processing algorithms that exploit the gained knowledge of geometrically regular signals and developed sparse representations. NeTS: Small: Collaborative Research: Logical Localization for Mobile Devices through Ambience Sensing The notion of location is broad, ranging from physical coordinates (latitude/longitude) to logical locations (like Starbucks, WalMart, museums). Logical locations are gaining importance because they express the context of the user, allowing applications to carry out context specific interaction. For instance, a person entering Starbucks may receive an advertisement for purchasing coffee. Evidently, such applications expect to learn the existence of the user within the confines of a logical place (like Starbucks). The rich body of localization literature based on GPS/WiFi/GSM is dominantly physical in nature, and translating them to logical locations is error prone. This is because a dividing wall may separate two logical locations and an error margin of few meters may place the mobile device on the incorrect side of this wall.

This proposal proposes a unique system called SurroundSense that breaks away from physical localization and investigates direct methods of recognizing logical locations. The main idea in SurroundSense exploits the observation that a logical place has a distinct ambience in terms of its background sound, light, color, RF signals, and layout. The attributes of the ambience can be sensed through mobile phone sensors and suitably combined to form a fingerprint of that place. This fingerprint can then be used to identify the logical location of the phone by comparing it against an existing database of location fingerprints. As an example, a Starbucks may include sound signatures from coffee machines and microwaves that are different from the clinking of forks and spoons in restaurants. Similarly, signature colors in the decor may vary from store to store, lighting styles and store layouts may be different, and so will the WiFi SSIDs audible at that location. The combination of all these attributes, a fingerprint, is likely to exhibit reasonable diversity for high quality, logical localization. SHF: Software Tools and Techniques for Maximizing Realism on Multi core Processors High performance computing is used to run models of the real world. This is true in both ultra high performance simulations used in scientific computing to study Physics, Biology etc. as well in designing approximations of real world involving applications such as gaming, social networks etc. Maximizing realism of the world being modeled and mimicked is the key to get this right. This work attacks the problem of realism on two fronts : first a framework is developed to speed up sequential parts of the computation using a large number of available cores based on a new concept of probabilistic speed up. Secondly a runtime solution is devised to maximize realism in immersive applications such as gaming under the constraint of responsiveness. The frameworks involve development of programming models, interfaces, APIs and run time system to solve the above problems by managing the underlying parallelism and computation. Apart from research this effort involves developing a cross cutting graduate level course that spans between simulations, algorithms and programming languages. CSR: SHF: Small: Automata Theorectic Approach to Hardware/Software Co Verification This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Computer systems such as personal computers and embedded systems are increasingly pervasive. People s everyday lives depend on these systems. Therefore, they must be high confidence. To boost system performance, hardware and software are often closely coupled for optimizations. As a result, separate verification of software and hardware cannot guarantee the correctness of the entire system; thus hardware/software (HW/SW) co verification, verifying the hardware and software together, is highly desired. There are four major benefits from co verification: (1) advocating design level HW/SW interface specifications, (2) facilitating early hardware and software verification in the system context, (3) reducing verification complexity by properly leveraging HW/SW couplings, and (4) broadening property coverage to the entire system.

This project develops an automata theoretic approach to HW/SW co verification. Major research tasks include: (1) developing a co specification scheme for hardware and software, (2) developing an automata theoretic model for co verification and its model checking algorithms, (3) developing abstraction/refinement algorithms for co verification, (4) developing a co verification toolkit supporting this approach, and (5) evaluating this approach and the toolkit on device driver co verification.

Education and outreach efforts include (1) collaborating with industrial partners on device driver co verification, (2) creating video demonstration of this research and make it available online for education and research purposes, (3) integrating research activities and results of this project into a three course software engineering sequence, (4) advising and mentor undergraduate and graduate students to conduct research in this project, and (5) broadening participation through outreach efforts. SHF: Small: SISA: A System Level ISA for Power Performance Management in CMPs Over recent decades, power and complexity challenges have limited the ability to derive continued growth in computing performance through clock frequency scaling. Future growth will come through implementing systems with many independent processors on the same chip. Challenges lie, however, in writing software for these systems, and in creating software that is portable across several hardware generations. For parallel software to smoothly exploit a chip s computation and communication capabilities, hardware needs better information regarding software s structure and resource requirements. Analogous to the traditional, fine grained instruction set architecture (ISA), this research proposes a higher level, coarse grained System level ISA or SISA. SISA provides information on computational chunks and the data or synchronization dependencies between them. Expressing software as a coarse grained directed graph, SISA enables efficient, adaptive scheduling of parallel computation and communication. It also offers other benefits for reliability, energy efficiency and portability.

The proposed research program will have broad impact in several ways. First, the PI has a solid track record of knowledge dissemination and technology transfer. This includes extensive collaborative relationships with industry, and several patents. In addition, the PI has a track record of releasing high impact software tools for external use. The Wattch power modeling tool is one of five major tools distributions from her group, in use by thousands of computing researchers worldwide. The PI also has a strong track record of support for undergraduate research and underrepresented groups, and has also advised summers of undergraduate research with women and under represented minorities through CRA W and Princeton programs. She has been involved in teaching non STEM students and multidisciplinary efforts, and will continue and broaden such activities through this research. CSR: Small: High Assurance at Low Cost in Data Centers Using Virtualization This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
As our reliance on online services continues to grow, the need for maintaining high availability of these services has become a pressing need as well as a challenge. The challenges include hardware failures, software bugs, operator error, malicious break ins, and wide area disasters such as power blackouts and natural catastrophes. This project is developing techniques to build data center services that remain highly available under such challenging conditions using virtualization.
The project makes two contributions. First, is a novel approach for Byzantine fault tolerance (BFT) with a replication cost factor close to F+1 in order to tolerate up to F failures, thereby halving the cost of state of the art BFT approaches. The savings in replication cost as well as commensurate gains in service throughput are realized using virtualization and are applicable to data centers hosting multiplexed services. Second, is a novel approach to seamlessly recover from wide area disasters with negligible service replication cost during typical conditions. These gains are realized using virtual machine checkpointing, asynchronous replication, and semantics aware storage techniques. The project demonstrates the feasibility of these approaches using theoretical analysis as well as prototype driven experiments. The prototype implementations will be made available to other researchers and practitioners. NeTS: Small: Collaborative Research: Transmission Re Ordering in Wireless Networks: Protocols and Practice This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The edge of the Internet continues to spread rapidly over the wireless medium. To cope with the escalating demand, every available opportunity in wireless networking will has to be maximally exploited, particularly at the border of physical and higher layers. One such opportunity is the capability of successfully capturing a frame even in the presence of an interfering frame, i.e., message in message (MIM). With MIM, two concurrent transmissions may be successful if they are activated in a specific order, but not the reverse. Towards harnessing MIM, this project aims to: i) Understand the intrinsic potential of MIM through theoretical analysis; ii) Draw on this understanding to design efficient higher layer protocols; iii) Develop a prototype testbed to validate/evaluate the MIM aware protocols.

This project offers obvious benefits to the broader community as it attempts to satiate the growing demand for bandwidth hungry applications over wireless networks by extracting the most from the scant spectral resources. Additionally, this project facilitates: i) Development and strengthening of the laboratory and curriculum for wireless networking at University of South Carolina and Duke University; ii) Involvement and mentoring of undergraduate students in wireless networking research; iii) Deployment of experimental wireless technologies within The Duke SmartHome. SHF: Small: 3D Integration of Sub Threshold Multi core Co processor for Ultra Lower Power Computing This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

A fundamentally different approach for circuit design that has recently gaining popularity is sub threshold circuits, where the power supply is set below the transistor threshold voltage to obtain energy savings when speed is not the primary constraint. Individual circuits or processing element designed at sub threshold region operates slowly, but throughput of the circuit can be improved by operating many of them in parallel. Hence, a processing fabric with many sub threshold cores can be a suitable platform for throughput oriented parallel applications. This proposal explores the viability of 3D die stacking of sub threshold and super threshold circuits for low power computing. More specifically, our sub threshold die contains many small sub threshold cores, whereas our super threshold die consists of several large super threshold cores. The sub threshold cores are used as co processors to execute massively parallel, high throughput, low/mid performance tasks at ultra low power, while the main super threshold processors handle high performance sequential tasks.

This new cooperative computing hardware is expected to provide a viable platform for many useful embedded software applications that require both high performance/power tasks as well as low performance/power tasks. 3D integration will allow each die to utilize the process technology optimized for sub and super threshold operation and be seamlessly integrated into a single system. TSV (through silicon via) based 3D communication requires no off chip access, thereby reducing power consumption further. Sub threshold circuits are currently used for some low power applications such as watches, hearing aids, distributed sensor networks, filters, and even pipelined micro processors. 3D integration is already available in the embedded domain, and the high performance processor industry is actively evaluating this technology for general purpose computing. This research will fill a critical gap that is needed to make 3D integrated sub threshold multi core co processors a reality. CSR: Small: Automated Software Failure Causal Path Computation Automating debugging has been a long standing grand challenge.
Central to automated debugging is the capability of identifying failure causal paths, which are paths leading from the root cause to the failure with each step causally connected. It is key to understanding and fixing a software fault. The project develops a novel scalable debugging technique. Given a failure and the desired output, the technique produces the failure causal path.

The following enabling techniques are devised. Given a failure and the desired output, the first technique is to search for a dynamic patch to the failure such that the patched execution generates the desired output. Sample patches include negating the branch outcome of a predicate execution. The second technique is to align the failing and the patched executions to facilitate later comparison. It consists of control flow alignment and memory alignment. Two runs may differ in control flow so that correspondence between execution points need to be established. A data structure may be allocated to different memory locations so that memories also need to be aligned. The third technique is to efficiently compare the program states of the two runs at the aligned places to generate the causal path. The ramifications include reducing resource consumption of debugging and improving software productivity and dependability. TC: Small: Denial of Service Attacks and Counter Measures in Dynamic Sepctrum Access Networks Abstract: 0917008: TC: Small: Denial of Service Attacks and Counter Measures in Dynamic Spectrum Access Networks

This project studies denial of service (DoS) attacks that are unique to dynamic spectrum access (DSA) networks: (a) DoS attacks by incumbent user emulation; (b) DoS attacks by protocol manipulation. In the first case, one or more malicious nodes pretend to be the primary by mimicking the power and/or signal characteristics to deceive legitimate secondary nodes into vacating the white space unnecessarily. In the second case, the malicious users either modify spectrum sensing related information or falsify their own sensing data thereby affecting the final decision. A number of mathematical models for the DoS attacks and several counter measures based on game theory, decision theory, stochastic learning, cryptography and Byzantine fault tolerance are developed in this project. Some defense mechanisms and protocols developed through this project will be tested on SpiderRadio (a cognitive radio test bed being developed in the PIs? laboratory).

Broader Impact: Since DSA networks are expected to play an important role in first responder networks, the solutions proposed here are expected to impact design of such networks. Since research in DSA network protocols and architectures are still in the formative stages, the proposed security solutions can be incorporated into the design phase of the system rather than being added on as an after thought.

Graduate and undergraduate level courses and Technogenesis scholarships will be used to disseminate the finding of this project. Special efforts will be taken to mentor women and minority students via established connections. CSR: An initial study of the potential of a new approach to storage QoS for dynamic, multi service infrastrctures CMU is working toward a new approach to providing storage quality of service (QoS) in distributed environments supporting multiple services with time varying workloads. Today s shared storage, whether implementing QoS or not, allows different services workloads to mix without consideration of the large efficiency swings caused by unpredictable interference. Ideally, however, each service would see full efficiency within the fraction of the I/O system s time allocated to it, which would increase effective utilization and enable practical storage QoS control. We are formulating and beginning exploration of a resource management architecture in which QoS control is layered atop robust performance insulation and equipped with algorithms for dataset assignment and slack exploitation. The eventual outcomes seeded by this project will include enabling the storage QoS critically needed for emerging virtualized and cloud computing environments, enhancing education at CMU and elsewhere by providing insights taught in our storage systems and distributed systems classes, and being integrated into a deployed instance used by scientists sharing a cluster for their work. CIF:RI:Small:Content Based Strategies of Image and Video Quality Assessment Abstract
Emerging demands on ubiquitous multimedia access continue to push coding algorithms to ca pitalize on content based properties of images and video. For example, by identifying and pre serving regions of interest, or by synthesizing textures at the decoder, it is possible to dramati cally reduce bandwidth requirements while preserving visual quality. These next generation coding strategies must be accompanied by next generation quality assessment algorithms that can handle the unique coding artifacts. Yet, determining quality in a manner that agrees with human perception remains a grand research challenge. Current quality assessment methods use a fixed analysis, whereas human perception adapts to the image?s content. In order to meet increasing demands on bandwidth, mobility, and IP streaming, there is a critical need to push the state of the art in quality assessment toward such a content adaptive approach.

In this research, the investigator conducts a series of studies designed to examine the utility of content adaptive models of human vision for quality assessment of images/video containing degradation and enhancement. The first study will collect a large set of subjective ratings for enhanced and degraded images and video. This effort will provide ground truth data for training and validation. Using these data, the investigator will: (1) Research new methods of quality as sessment that can deal with images containing enhancement. (2) Research and model the mul tiple strategies employed by the human visual system during quality assessment, including de veloping content based neural models and image adaptive techniques of strategy selection. (3) Research the relationship between quality and regions of interest. This research will lead to more accurate and robust methods of quality assessment, and it will lay the groundwork for next generation perceptual models that take into account the adaptive nature of human vision. NeTS: Small: Collaborative Research: Transmission Reordering in Wireless Networks: Protocols and Practice This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The edge of the Internet continues to spread rapidly over the wireless medium. To cope with the escalating demand, every available opportunity in wireless networking will has to be maximally exploited, particularly at the border of physical and higher layers. One such opportunity is the capability of successfully capturing a frame even in the presence of an interfering frame, i.e., message in message (MIM). With MIM, two concurrent transmissions may be successful if they are activated in a specific order, but not the reverse. Towards harnessing MIM, this project aims to: i) Understand the intrinsic potential of MIM through theoretical analysis; ii) Draw on this understanding to design efficient higher layer protocols; iii) Develop a prototype testbed to validate/evaluate the MIM aware protocols.

This project offers obvious benefits to the broader community as it attempts to satiate the growing demand for bandwidth hungry applications over wireless networks by extracting the most from the scant spectral resources. Additionally, this project facilitates: i) Development and strengthening of the laboratory and curriculum for wireless networking at University of South Carolina and Duke University; ii) Involvement and mentoring of undergraduate students in wireless networking research; iii) Deployment of experimental wireless technologies within The Duke SmartHome. AF:Small:RUI:Concrete Computational Complexity The goal of computational complexity theory is to determine the computational resources needed to carry out various computational tasks. The resources measured may involve hardware (such as gates used to construct a circuit, or area on a chip) or software (such as the time or space used in the execution of a program on a machine), and the tasks considered may range from simple addition of two integers to a large algebraic or geometric computation.

This project will deal primarily with low level complexity theory, in which the resources required grow modestly (at most quadratically) with the size of the task. Examples of such tasks are furnished by the arithmetic operations (addition, subtraction, multiplication, division and square root extraction) performed by the executions of single instructions in a computer. For these tasks, hardware oriented resource measures are most appropriate in most cases.

The broader impacts of the project lie in the opportunity it will provide to explore a new model for undergraduate research. The most common model for undergraduate research is to give students problems that they may reasonably be expected to solve within the time allowed (typically an academic year for a senior thesis, or ten weeks for a summer assignment). This project will explore a new model, wherein students are assigned the task of working on an authentic research problem (one that has resisted solution for many years, and is unlikely to be resolved with a single new stroke), with the goal of making a contribution (even a small one) that might play a role in an eventual resolution. If explored with imagination, this new model should provide a valuable complement to the established practices for undergraduate research. NeTS: Small: Automated Design of Medium Access Control Protocols This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The performance of wireless networks critically depends upon the careful design of networking protocols. Networking protocols are traditionally designed based only on intuition, and through a complex design process that is particular to the network at hand. The disadvantage of the traditional approach is that protocol design cycles are long, costly, and prone to human error. The lack of a general methodology also means that either the protocols are general purpose and hence low performance, or are specialized and hence inflexible and require complete re design when the application specifications change. The major research contribution of this project is the establishment of a general methodology for the automation of networking protocols, taking the Medium Access Control (MAC) protocols as a first prototype. The MAC protocol performance is critical in high performance wireless networks. The following design process chain is created to generate a MAC protocol based on resource constraints and application specific goals: (1) Optimization Program: Formulation of the network protocol problem that models the effects of control information exchanges, (2) Solvers: Optimal waveform generation, which specifies the optimal exchange of both control information and data, (3) Protocol Extraction: The extraction of the optimal protocol as a minimal description of the optimal waveforms. This technology opens the way to the design of networking protocols that will be based on automation design tools in the future. This project also integrates the development of prototype tools into graduate level coursework. CSR: Small: Efficient and Reciprocal File Replication and Consistency Maintenance in Pervasive Distributed Computing Advancements in technology over the past decade are leading to a promising future for pervasive distributed computing (PDC), where access to information and services is provided anytime and anywhere around the world. The proliferation of Internet scale applications poses new challenges for how future file systems are designed and managed to enable this future. The difficulty of these challenges grows with the number of users and the intensity of the data, and is further compounded by the need to support replication and consistency. However, previous file replication and consistency maintenance approaches are not efficient enough to be scalable for PDC. These two issues are typically addressed separately, despite the significant interdependencies between them. This research addresses this issue through the development of a file management system with Efficient and Reciprocal file Replication and Consistency maintenance (ERRC). The ERRC system will incorporate swarm intelligence based file replication and coordinated multi factor oriented file replication algorithms to achieve high efficiency in both file replication and consistency maintenance. The system also contains self adaptive file replication and consistency maintenance mechanism that allows each node to determine the need of file replication and consistency maintenance in a decentralized manner. Understanding and insight gained as a result of this project will be disseminated through technology transfer to industry partners, in addition to publication and software release channels. The multi disciplinary nature of this research also lends itself to cross disciplinary education and well rounded training of students. CIF: SMALL: Information Theoretic Multi Core Processor Thermal Profile Estimation Dynamic thermal management is the process of controlling surges in the operating temperature especially of a multi core processor during runtime, based on limited measurements by on chip thermal sensors. Managing thermal sensors and processing their measurements presents a rich vein of theoretical and practical challenges including: deciding on the number, location and type of thermal sensors; estimating the thermal profile; and characterizing the fundamental performance tradeoffs between sensor quantity and complexity in guaranteeing estimate accuracy. In this interdisciplinary project, we propose a new approach to the problem of thermal profile estimation in multi core processors which relies on fundamental information theoretic principles. Our approach rests on new problems in information theory that capture the salient features of on chip thermal profile estimation. Our new formulations are inspired notably by rate distortion theory and also bear a similarity to compressed sensing. Furthermore, our approach has wider applicability to general problems of parameter estimation based on limited sampled and quantized measurements.

The project has a home in computer architecture, VLSI design as well as one in information theory and compressed sensing. Its potential impact is twofold: (a) improvement in the performance and reliability of multi core processors; and (b) new models and problem formulations in the fields of information theory and compressed sensing. The broad reach of this project will provide a valuable learning environment for the investigators and their graduate and undergraduate students. The basic elements of the technical approach will be discussed in Special Topics courses and seminars with the participation of graduate students. NeTS:Small:Dynamic Coupling and Flow Level Performance in Data Networks: From Theory to Practice This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).

The performance of network users best effort flows, e.g., file transfer delays and web browsing responsiveness, depends on the resources they are allocated over time. When varying traffic loads share these resources and/or wireless nodes transmission capacities depend on each other through interference, the allocated resources and thus flows performance are coupled through the traffic and interference dynamics. This project investigates such performance coupling in data networks and, in turn, how it affects protocol and network design. This is a significant problem, as most data networks share these characteristics, and we currently have no robust tools to effectively predict and thus optimize performance. Expected results include new theory, approximations and performance bounds enabling the analysis of such systems which are also applicable to other domains. Expected results also include using these tools to investigate: (1) the benefits of, and capacity allocation, in networks supporting multipath transport and/or shared wireless access networks; and (2), the development of improved protocols and algorithms that are sensitive to such performance coupling, e.g., preliminary work suggests the state of the art base station association policies can be substantially improved by factoring the performance coupling resulting form interference. The work s impact will be assured through broad dissemination in the research community, and by leveraging industry partners in evaluating practical applications and technology transfer opportunities. III:Small:RUI:Integrating Image and Non Image Geospatial Data This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

This project develops novel methods for integrating image and non image geospatial data to (1) advance the state of the art of automated remote sensed image analysis and, in turn, (2) improve the coverage and fidelity of the non image repositories.

A characteristic of remote sensed imagery which has not been sufficiently exploited by the analysis community is that the images can be georeferenced to extensive repositories of non image geospatial data such as maps and geographic dictionaries termed gazetteers. In particular, these associations represent a rich source of labeled data needed to train the analysis algorithms.

The first part of this project develops a framework for learning appearance models for a large set of geospatial objects indexed by an extensive gazetteer in an unsupervised fashion. Besides the standard challenges such as choice of features and form of the model, this problem is made interesting by the fact that current gazetteers only specify the spatial footprint of indexed objects using a single point location. Methods are explored for simultaneously estimating the model parameters and the spatial extents of the known objects.

The second part investigates using the learned models to update the gazetteer from imagery. This includes estimating the spatial extents of known object instances as well as detecting unknown or novel instances.

The project has research and educational synergies with a Spatial Analysis Research Center that the PI is establishing at the new University of California at Merced. The PI plans to open imagery so that evaluation datasets can be made publicly available through the project website (http://vision.ucmerced.edu/projects/integrating/). NeTs: Small: Interdomain X ities: Toward Five Nines Availability in Internet Routing This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Various studies over the past decade have shown that network availability on the Internet is about 99%, which pales in comparison to other utility services such as power grids and telephone networks. The primary cause of network unavailability today is due to problems related to interdomain routing that are unlikely to go away with technology trends or further growth as they are due to systemic limitations of the protocol architecture. This project is developing techniques towards the design of an interdomain routing architecture that provides high availability under flexible routing policies, link and node failures, and router misconfiguration.

The project has the following thrusts. First, it develops a quantitative foundation for interdomain X ities , a term used to describe metrics desired in an interdomain routing protocol such as availability, stability, policy flexibility, accountability, predictability, deployability etc. Second, it develops routing protocols based on insights from the theory of distributed systems, namely, using redundancy to mask failures, and treating consistency as a safety property. Specifically, the project builds upon multiprocess routing , an approach that runs multiple parallel routing processes that select primary or backup routes to deliver packets with high probability under multiple link and node failures; and consensus routing , a consistency first approach to ensure high availability under flexible policies. The project adapts these approaches to tolerate failures as well as to limit the impact of misconfiguration. These new proposals will be compared with existing research proposals for interdomain routing based on the X ities axes. The protocol designs will be made available to researchers and practitioners through open source implementations. NeTS: Small: Scalable Wireless Mesh Network Technology at the Edge for Efficient Mobile Video Data Access on Demand This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The growing popularity of wireless networks at the periphery of the Internet, combined with the need for ubiquitous access to video on demand (VoD) applications, calls for a new generation of video streaming technology. This research is investigating techniques to enable efficient VoD service over wireless mesh networks (WMN). Four methods are studied: wPatching is a multicast technique designed to leverage the broadcast nature of WMN to support VoD applications; Dynamic Stream Merging utilizes smart mesh nodes to merge video streams in order to conserve wireless resources; a hybrid environment uses mobile nodes as relay nodes to extend the service range of the WMN; and Constrained Multicast provides efficient and seamless handoffs for mobile users in the WMN by eliminating Internet router (layer 3) handoff delay. Along with extensive simulation studies, a prototype is developed to experiment with the Dynamic Stream Merging and Constrained Multicast techniques. These new streaming methods provide an enabling technology to extend VoD applications beyond the traditional wired environment to reach mobile users on wireless networks at the edges. The result is a new paradigm for efficient management of VoD traffic over a WMN. Applications such as digital libraries, distance learning, public information systems, electronic commerce, and video entertainment will benefit from this innovative streaming technology. The software and protocols developed for this project will be made available to other researchers to facilitate and encourage future research and application experimentation. NeTS: Small: The Impact of Message Passing Complexity on QoS Provisioning in Stochastic Wireless Networks This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Abstract (NSF 0917087):

Wireless networks operate under hostile conditions and often exhibit multi scale stochastic dynamics. In such dynamic environments, network algorithms for quality of services (QoS) provisioning hinge heavily on state information exchange, and network functions are intimately tied with the complexity of message passing. This project aims to pursue a systematic characterization of the impact of message passing complexity, a fundamental yet under explored area. Under such a common theme, the proposed research is organized into two coordinated thrusts.

1) Thrust I focuses on the impact of message passing complexity on effective throughput and delay performance of wireless scheduling. Novel vacation models are developed to account for signaling complexity, and effective throughput is studied using the fluid approach and delay analysis is carried out by diffusion approximation.

2) In Thrust II, noisy feedback models are devised to account for message passing complexity in distributed rate control algorithms using various optimization methods, and stochastic stability is characterized accordingly. The framework here provides a platform to compare different rate control algorithms in terms of complexity and robustness.

This project will significantly advance the understanding of the impact of message passing complexity on QoS provisioning in stochastic wireless networks. The study on open problems, such as delay performance of wireless scheduling, will open up new research directions in this area. Undergraduate students will get involved to carry out network performance measurements in this project. SHF: Small: Change Theory for Variation Aware Programming The management of changes is a difficult and error prone process in software maintenance and many other fields. Due to a lack of theory, there are few tools and methods available to systematically change objects while preserving important properties along with such changes. This problem is addressed by the theory of structured change. The overall research objective is to find principles for sound change management and to establish a theoretical foundation for the development of supporting tools. One particular goal is the development of domain specific languages, which users can employ to effectively manage changes in all kinds of software applications.

The following technical approach is pursued. First, a flexible change representation is developed. Based on this representation, laws of a change algebra will be established to identify property preserving transformations that can support sound change transitions. This work is accompanied by the development of algorithms for the incremental checking of property preservation. Ultimately, the theory will enable the development of tools that help users to effectively manage the evolution of objects. This goal is supported by the investigation of interaction principles that underly the editing and exploration of structured objects and their changes. Building on top of the theoretical foundation, the development of domain specific languages will provide concrete help for users to express more sophisticated transformations and combinations than the simple one step operations that are offered by the underlying formal model. Specifically, in the area of change representations for programs such a DSL will yield new, theoretically founded support for feature oriented programming and software product lines. In the context of spreadsheets, web sites, etc. the development of such DSLs will empower millions of users to deal in a more systematic way with changes. AF: Small: Analysis Algorithms: Continuous and Algebraic Amortization Adaptive numerical algorithms are widely used to solve continuous problems in Computational Science and Engineering (CS & E). Unlike discrete combinatorial algorithms which predominate in Theoretical Computer Science, such algorithms for the continua are typically numerical in nature, iterative in form, and have adaptive complexity. The complexity analysis of such algorithms is a major challenge for theoretical computer science. In particular, it is necessary to properly account for the adaptivity that are inherent in such algorithms. Until now, all complexity analysis that accounts for adaptivity (for example, in linear programming) must invoke some probabilistic assumptions. The broader impact of this project lies in the push to extend the scope of theoretical algorithms into the realm of continuous computation. The project is seen as part of a research program to develop a computational model and complexity theory for real computation, one that can account for the vast majority of algorithms in CS & E.

This project develops a new non probabilistic analysis technique called continuous amortization. It is able to quantify the complexity of an input instance as an integral, and reduce the problem to providing explicit bounds on the integral. The success in producing the first example of such adaptive bounds for the 1 dimensional case is now extended to higher dimensions. In order to bound these integrals, one needs another form of amortization called algebraic amortization. This generalizes the usual zero bounds by simultaneously bounding a product of individual bounds. These advances build upon the principal investigator s work in previous NSF projects on Exact Geometric Computation. The project also validates its algorithms by implementing them using the open source Core Library software. AF: Small: Toward mechanical derivation of Krylov space algorithms The objective of this research is to systematize the derivation of algorithms in the field of iterative linear system solving: iterative methods, preconditioners, multigrid. The approach is by extending the Formal Linear Algebra Methods Environment (FLAME), a system originally developed for deriving dense matrix algorithms.

The merit of this research lies firstly in the fact that it facilitates experimentation, since it makes derivation of new algorithms essentially simpler than the lengthy induction arguments that are traditionally necessary. Secondly, it will lead to algorithms being proved correct by the very mechanism of derivation. Finally, a complete systematization of FLAME may take the form of a symbolic system, where algorithm implementations are derived mechanically, steered by the user but otherwise autonomously, from a specification of their properties rather than from an algorithmic description.

The impact of this research will be on the computational community, since it lowers the threshold to exploring new algorithmic strategies, and on software developers, since it makes it easier to derive correct implementations of algorithms. Additionally, it will impact the way the subject of iterative linear system solving is taught, since FLAME worksheets offer a simpler and more insightful description of algorithms than is used traditionally. NeTS: Small: Addressing Research Challenges in Low Duty Cycle Wireless Sensor Networks Energy efficient wireless communication is critical for long term sensor network applications, such as military surveillance, habitat monitoring and infrastructure protection. To reduce the energy costs of RF listening, a node has to reduce its duty cycle by sampling wireless channels very briefly and shutting down for long periods. Consequently, the connectivity of low duty cycle networks becomes time dependent. Previous research in this type of networks predominately focused on physical and link layer designs. This project is positioned to provide significantly added value to these earlier successful research by conducting the first systematic research at the network layer for low duty cycle communication under a wide spectrum of network configurations covering a large design space. The key research challenge addressed by this project is to optimize networking performance (e.g., delay, reliability, and cost) in the presence of sleep latency and other practical considerations including (i) unreliable links, (ii) dynamic energy availability, and (iii) mobility. With a successful outcome from this project, long term sensor applications can be supported by low duty cycle networking technologies, leading to significantly reduced costs in development, deployment, and maintenance. These applications, in turn, can improve the safety of transportation, the quality of education and learning, and the development of innovations in many scientific frontiers.

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5). NeTS: Small: Greed Resistant Protocols This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Communication protocols on the Internet are typically designed first to work, then to interoperate and evolve, and eventually to be resilient to misbehavior and selfishness. Unfortunately, these extensions to protocols often sacrifice backward compatibility, because otherwise a misbehaving user can claim to be legacy. At the other extreme, the ``incentive compatible protocols are often designed assuming selfish users are also myopic, focusing narrowly on the short term and unwilling to donate resources for no immediate gain. This project investigates the iOwe, a new primitive for exchanging value in competitive protocols, and applies it to two new overlay network protocols, PeerWise and HoodNets. The iOwe is a verifiable promise of future service that can be exchanged by nodes that trust the issuer. PeerWise and HoodNets are protocols by which users can achieve better network performance, either in terms of latency (PeerWise) or wireless bandwidth (HoodNets). Both rely on a non simultaneous exchange of service: PeerWise because simultaneous use is unlikely, HoodNets because channel bonding achieves better apparent bandwidth only when both links aren t already saturated. The project results are expected to provide new facilities to help build protocols in which repeated interactions permit users to operate at a short term loss for long term gain. Similarly, the protocols being prototyped can aid in the robustness and performance of network service. CIF: Small: Nonintrusive Digital Speech Forensics: Source Identification and Content authentication Current digital media editing software allows malicious amateur users to perform imperceptible alterations to digital speech communications This creates a serious threat to the knowledge life cycle . The proposal seeks to develop theories, methods and tools for extracting and visualizing evidence from digital speech content for the purpose of media source identification and content authentication. The strategy will be based on the important paradigm of nonintrusive media forensics. The hypothesis is that the physical devices and associated signal processing chains may leave behind intrinsic fingerprint that are detectable by statistical methods. The hypothesis clearly indicates the purpose of the project. RI:Small:RUI: Towards the Next Generation of Stereo Algorithms This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

This project provides challenging test data and benchmarks designed to advance stereo vision methods to a level of practical relevance. It aims to bridge the gap between the sophisticated but brittle methods that perform best on current benchmarks and the robust but simple methods employed in real world applications.

The project provides new high resolution datasets with accurate ground truth, taken with different cameras under different lighting conditions, and depicting complex indoor and outdoor scenes with non Lambertian surfaces and outliers such as moving people, reflections, and shadows. The project explores novel algorithmic approaches for dealing with such challenges, including ways to leverage resolution, deriving color and noise models on the fly, and designing local region growing techniques that allow deferring global optimization from the pixel level to the region level. Undergraduate students are actively involved in all components of this research.

The project has strong potential impact along several fronts. The datasets and benchmarks resulting from this work serve as catalyst for new research and enable machine learning approaches. The algorithmic contributions allow harnessing the explosion of images available online. Robust matching techniques that can handle the variety of images available on the Internet enable a host of new applications with broad impacts on the population at large, including visual localization and navigation, as well as automatic 3D reconstruction and visualization of whole cities. Finally, the project exposes undergraduates at a liberal arts college in rural Vermont to the world of research, experimentation, and discovery. RI: Small: Efficient Reinforcement Learning for Generic Large Scale Tasks Recent advances in autonomous agents research are pushing our society closer to the brink of the widespread adoption of autonomous agents in everyday life. Applications that incorporate agents already exist or are quickly emerging, such as domestic robots, autonomous vehicles, and financial management agents. Reinforcement learning (RL) of sequential decision making is an important paradigm for enabling the widespread deployment of autonomous agents. However, a few notable successes notwithstanding, state of the art reinforcement learning algorithms are not yet fully capable of addressing generic large scale applications.

This project is advancing in four directions to scale up application of RL systems. Specifically, the project is (1) developing algorithms to automatically structure the input, output, and policy representations for learning; (2) introducing parallelizable reinforcement learning algorithms so as to exploit modern parallel architectures; (3) unifying abstraction and hierarchical reasoning with model based learning for the purpose of enabling intelligent exploration of large scale environments; and (4) enabling reinforcement learning algorithms to benefit from low bandwidth interactions with human users. Finally, we intend to unify the four research thrusts above into a single algorithm and conduct empirical evaluation on real world/large scale applications, to include biped robot balancing and walking, robot soccer in simulation and with real robots, and a full size autonomous vehicle capable of planning paths in an urban environment.

In addition to research advances and implications for improving national infrastructure, the project will contribute to undergraduate and graduate curriculum development. RI: Small: Scaling Up Inference in Dynamic Systems with Logical Structure Stochastic inference problems arise naturally in many applications that interact with the world, such as natural language processing (NLP), robot control, analysis of social networks, and environmental engineering. Current real world applications require inference mechanisms that can scale to thousands and more interactions. This project is scaling up and improving precision of automated inference and learning in dynamic partially observable domains, with later application to decision making, by combining the complementary computational strengths of logical and probabilistic methods. The inference and learning algorithms being developed are being assessed by theoretical and experimental means, with aims of improving question answering from text narratives and environment state estimation for mobile robots. Examples of societal benefits are enabling people to access and examine details hidden in large amounts of textual information as well as enabling more helpful and autonomous mobile robots. III: Small: Making and Tracing: Architecture centric Information Integration In many scientific domains information is scattered across numerous interrelated and heterogeneous artifacts. In software engineering in particular, artifacts such as requirements, design documents, and code are often isolated by tools, development groups, and geographic locations. Current traceability approaches ? attempts to link related information ? fall short in tracing across heterogeneous artifacts and in supporting user customized links. This project is aimed at crossing these information barriers, enabling the creation of traceability links between related artifacts, to support tasks such as impact analysis and software maintenance.

This project targets automated architecture centric traceability which centers links on the architecture, enabling scalable and flexible link capture. Furthermore, stakeholders control link capture, enabling them to directly benefit from the links. The prevalent approach to automatic traceability is to recover links from existing artifacts. In contrast, this project pursues prospective generation of trace links which captures links in situ, while artifacts are generated or modified, enabling the capture of contextual relationships. Open hypermedia adapters enable the capture of links across heterogeneous artifacts and the rendering of resources at different levels of granularity. Users can determine the artifacts to trace and the link semantics to assign via externally customizable rules.

The approach is applicable to data provenance capture in e Science. Prospective capture can aid in inferring experiment design and capturing links across heterogeneous artifacts like publications, data files, and plots. The results will also be valuable to the development of safety critical systems where satisfaction of all requirements is part of safety assurance. SHF: Small: Rethinking Computer Architecture for Secure and Resilient Systems Our society, economy and national security are critically dependent on computers and computing devices. With a few hops, commodity computers and mobile phones can be connected to secret or sensitive information or to critical infrastructures. However, mainstream commodity computers have not been designed with security in mind, for the last three or more decades. Rather, they have been designed to improve performance, energy efficiency, cost or size, with security added on as an after thought. While some specialized secure computers have been built, up to now, one had to sacrifice performance (or cost and convenience) for security. In this research, the PI plans to explore what is feasible if we allow ourselves a clean slate design, where security is a first class goal, on par with performance and other goals. The investigation will rethink computer architecture from first principles to significantly improve both the security and the performance of future computers.

The research has two thrusts: how to design computer architecture to enable more secure software and systems, and how to design computer hardware components that are themselves more trustworthy. The PI will develop a new threat based design methodology for computer architecture, examining how to build security awareness into the design of each basic aspect of computing. New architectural foundations for secure processing, secure memories, secure caches, secure virtual memory translation, secure storage, and secure control flow will be developed. The research will focus on providing the cornerstones of security: Confidential and Integrity of critical information, and Availability, in the sense of resilient attestation and execution of security critical tasks even when parts of the system may have been corrupted. The solutions will also consider hardware attacks, in addition to the software and network attacks considered by software security solutions and the current state of the art hardware TPM (Trusted Platform Module) solution.

The intellectual contributions of this research will be new architectural foundations, and a new dimension of threat based design in the research and development of all future computers. The broader impact of this research is to provide core security technology that can be built into commodity computing devices and their servers. These can be used in computer, communications, control, entertainment and embedded systems to build significantly more secure systems that will provide a leap forward in information and cyber security, benefiting our society. SHF:Small:Reasoning about Specifications of Computations The research will develop the foundations for a framework for reasoning about the formal properties of programming languages, compilers, software specifications, concurrent systems and other related computational systems. The framework will be based on two separate but interacting logics. One logic will be geared towards specifying and prototyping varied software systems. The second logic, referred to as the meta logic, will provide flexible and powerful mechanisms for reasoning about specifications written in the first logic. The objects to be specified and reasoned about typically have complex syntactic structures, often involving some form of binding. Use will be made of a higher order approach to representing syntactic structure in both logics to facilitate a natural treatment of such objects. Useful new logical capabilities will be exposed and embedded in actual computer systems that can be used in prototyping and reasoning tasks in the intended domains. The insights and the tools produced will be used pedagogically to expose high school students and beginning undergraduates to important ideas in logic and computation. A close collaboration with a group of French researchers will provide an international dimension to the research, co funded in part by the NSF Office of International Science and Engineering. In the long run, mechanized formal specification of (and reasoning about) programming languages has clear application to the improvement of software infrastructure in the real world: its correctness, reliability, maintainability, and security. RI: Small: Recursive Compositional Models for Vision Detecting and recognizing objects from real world images is a very challenging problem with many practical applications. The past few years have shown growing success for tasks such as detecting faces, text, and for recognizing objects which have limited spatial variability.

Broadly speaking, the difficulty of detection and recognition increases with the variability of the objects ? rigid objects being the easiest and deformable articulated objects being the hardest.
There is, for example, no computer vision system which can detect a highly deformable and articulated object such as a cat in realistic conditions or read text in natural images. This project develops and evaluates computer vision technology for detecting and recognizing deformable articulated objects.

The strategy is to represent objects by recursive compositional models (RCMs) which describe objects into compositions of subparts. Preliminary work has shown that these RCMs can be learnt with only limited supervision from natural images. In addition, inference algorithms have been developed which can rapidly detect and describe a limited class of objects. This project starts with
single objects with fixed pose and viewpoint and proceeds to multiple objects, poses, and viewpoints. Theoretical analysis of these models gives insight and understanding of the performance and computational complexity of RCMs.

The expected results are a new technology for detecting and recognizing objects for the applications mentioned above. The results are disseminated by peer reviewed publications, webpage downloads, and by university courses. III: Small: Techniques for Integrated Analysis of Graphs with Applications to Cheminformatics and Bioinformatics A number of scientific endeavors generate data that can be modeled as graphs: high throughput biological experiments on protein interactions, high throughput screening of chemical compounds, social networks, ecological networks and food webs, database schemas and ontologies. Access and analysis of the resulting annotated and probabilistic graphs are crucial for advancing the state of scientific research, accurate modeling and analysis of existing systems, and engineering of new systems. This project aims to develop a set of scalable querying and mining tools for graph databases by integrating techniques from databases and data mining. The proposed research work is theoretical as well as empirical. New theoretical ideas and algorithms are being developed and these are being applied to the domains of Cheminformatics and Bioinformatics.

The first research thrust examines primitives for graph data management and graph mining. A declarative query language for graphs is being investigated. This language is based on a formal language for graphs and a graph algebra, and separates the concerns of specification and implementation. Scalability of techniques for similarity search on graphs and mining for significant patterns is being investigated as a part of this thrust.

The second research thrust applies the developed techniques to the domain of Cheminformatics. Specific tasks that are being examined are search for similar compounds, mining for significant motifs, diversity analysis, and analysis of macromolecular complexes.

The final research thrust applies the developed methods to the domain of Bioinformatics. There has been an explosion of data of widely diverse biological data types, arising from genome wide characterization of transcriptional profiles, protein protein interactions, genomic structure, genetic phenotype, gene interactions, gene expression, proteomics, and other techniques. Techniques being developed can integrate and analyze data from multiple sources and models efficiently, while accelerating (interaction and function) prediction, and pathway discovery.

Further information about the project can be found at the project web page http://www.cs.ucsb.edu/~dbl/0917149.php. RI: Small: Region based Probabilistic Models for Descriptive Scene Interpretation This project focuses on the task of providing a consistent, semantic interpretation of all components of an image of an outdoor scene. The image is automatically segmented into large regions, each of which is a coherent scene component that is labeled with a rough geometric configuration (distance from the camera and surface normal), and with one of a subset of semantic classes, which include both background classes (such as water, grass, or road) and specific object classes (such as person, car, cow, or boat). The approach is based on the development of a holistic probabilistic model (a Markov random field) whose parameters are automatically learned from data. The model exploits both scene features and contextual relationships between scene components (e.g., cows are typically found on grass and boats on water). It also utilizes object shape and appearance models to identify specific object instances in the image. To address the complexities of reasoning using these richly structured models, new probabilistic inference algorithms are developed.

This project helps train graduate and undergraduate students within the PI s group, as well as students in an annual project class in this area. The project also develops significant infrastructure, including an extensive data set of labeled images and efficient inference algorithms, which are freely distributed to the research community.

The ability to provide a coherent interpretation of a scene composition is an important step towards automatic image annotation, with benefits both for image retrieval and for providing image summaries to visually impaired users. AF: Small:Explorations in Computational Learning Theory A primary goal of machine learning is to have computers learn from data, and to make predictions based on what they have learned. Machine learning has been used in many applications, such as identification of spam emails, detection of suspicious computer network traffic, and detection of malignant tumors. The use of machine learning is based on the implicit assumption that there is a mathematical function that describes, with some accuracy, the relation between the inputs to the prediction problem and the correct prediction. The function is not arbitrary; instead, it is of a certain restricted type. However, not all types of functions are efficiently learnable. Also, learnability depends crucially on the type of data that is available.

This project focuses on the learnability of Boolean functions, a central topic in computational learning theory. The research in this project falls into three main categories: learning from random examples, learning with costs, and DNF learning and minimization. Problems in the first category address core open questions in the standard PAC learning model and explore the extent to which access to data from different probability distributions can aid in learning. Problems in the second category are motivated by concrete problems in protein engineering, databases, and cyber security, where there are costs associated with determining the value of inputs, or in obtaining data. The third category concerns problems of properly learning DNF formulas using DNF hypotheses, related complexity theoretic problems concerning DNF minimization, and problems concerning the complexity of certificates of DNF size.

Broadly, this project seeks to expand our understanding of which types of functions are efficiently learnable by computers, and under what conditions. The research on learning with costs can yield advances in the application areas that motivate it. DNF minimization is a central problem in both complexity theory and in the design of logic circuits; the research on DNF has the potential for impact in both these areas. CSR: Small: Workload Shaping for Capacity and Power Provisioning in Storage Data Centers Workload Shaping for Capacity and Power Provisioning in Storage Data Centers

Abstract

This award is funded under the American Recovery and Reinvestment
Act of 2009 (Public Law 111 5).

Tremendous growth in hosted services and data sharing are motivating a new generation of storage virtualization technologies that encompass the entire data center or ensembles of data centers. Accurate dynamic resource provisioning is critical to achieving cost effective and environmentally sensitive operation of these service centers in order to (i) reduce fixed costs for floor space, server hardware, and infrastructure for power distribution and cooling, (ii) lower the operational costs for management and electricity, and (iii) mitigate the hidden environmental and social costs of excessive energy use.

The goals of this project are to develop effective dynamic resource provisioning and scheduling schemes for storage servers using novel workload shaping techniques to improve cost and energy consumption. The research involves the design, analysis, implementation, and performance evaluation of new algorithms for workload shaping with superior performance characteristics. This will result in significant improvements in resource provisioning, capacity prediction, and admission control, as well as energy and power consumption. Unlike traditional server scheduling techniques in which the negative effects of workload bursts affect even well behaved portions of the workload, the proposed workload shaping approach isolates the bursts and confines their effects to localized regions using online workload decomposition and recombination.

The project will support the education of graduate students in an emerging sector of Information Technology, and will also involve undergraduate engineering students in research in a topical technical area. SHF:Small:Multicore Hardware/Software CoDesign using Fast, Cycle Accurate Simulators Tuning software for improved performance/power is often extremely difficult because one cannot easily observe the entire inner workings and power consumption of running hardware at a fine time resolution. Thus, determining the root cause of performance and power issues can often only be done by careful crafting of experimental code, running that code, and measuring the limited information that can be observed.

This proposed research is aimed at designing a new class of very fast and accurate simulators that, with proper instrumentation, can provide significantly more visibility than the modeled hardware. However, such simulators can result in information overload, making finding the issues difficult. Thus, the PI proposes to investigate methods to effectively and automatically find such issues so that they can more quickly and easily be addressed. Such methods could result in more efficient computer systems that can solve problems faster and/or use less energy, potentially impacting hardware and software developers and users alike. TC: Small: V2M2: Towards a Verified Virtual Machine Monitor Virtualization is rapidly becoming a key technology for computing
systems, promising significant benefits in security, efficiency, and
dependability. Fully realizing these benefits depends upon the
reliability of virtual machine monitors (hypervisors). This research
aims toward developing a very high assurance, commercial quality
hypervisor by:

(1) designing practical tools and techniques to make reasoning about
hypervisors feasible;
(2) honing them through the formal verification of an existing simple
research hypervisor; and (3) applying them to develop and prove a graduated series of more and
more realistic hypervisor models, aiming toward a fully
functional, formally verified hypervisor.

The project proceeds along two tracks. In the first, an existing
research hypervisor is modeled and verified using the ACL2 formal
analysis tools. In the second track, tools and methodologies are
developed applicable to a wide variety of hypervisor implementations.
This work aims to advance the formal analysis of an important class of
software applications and to advance the state of the art in formal methods,
particularly management of large and complex formal proofs. This
research is a collaboration between the formal methods and systems
research groups at the University of Texas at Austin. SHF: Small:Transforming Linear Algebra Libraries Ever since the first FORmula TRANslator, compilers have purported to
take an algorithm, formulated in human terms, and compile it to an
efficient executable. Practice clearly does not live up to this
ideal, as programming languages force execution oriented design
decisions on the programmer, and compiler imperfections necessitate
further manual optimizations. While the ordinary programmer can
ignore this problem by using libraries, the library developer cannot.
The problem appears more desperate now than ever, since it is not
known what form future architectures will take, combined with a trend
towards pushing complexity away from the architecture (for example to
lower power consumption) and onto the program and compiler.

The FLAME project already allows linear algebra libraries to be
developed and coded at a high level of abstraction that better
captures the underlying algorithms. It has already been shown that
these technique greatly simplify the porting of libraries to different
platforms ranging from conventional sequential computers to exotic
multiGPU accelerators. However, the APIs for coding at a high level
of abstraction incur a considerable performance penalty for operations
that in each step perform relatively little computation (e.g., level 2
BLAS operations and ``unblocked algorithms). As a result, a
nontrivial part of libflame, the library that has resulted from the
project, still requires coding at a low level.

The project will develop a source to source translator will take
algorithms expressed at a high level of abstraction and will transform
these into a range of representations, including high performance
low level code. This will overcome the final legitimate objection to
the FLAME methodology since coding at a high level of abstraction will
no longer carry a performance penalty. The approach will be
generalized so that code transformations that an expert applies by
hand in order to target different platforms will be made mechanical.
Together, these represent a major departure from traditional
approaches to library development. CSR: Small: Scalable Applications Start with Scalable Virtual Machine Services This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The goal of this proposal is to devise and implement a scalable virtual machine that includes scalable garbage collectors, profilers and just in time (JIT) compilers.

As hardware vendors deliver chip multiprocessors (CMPs) as the next generation general purpose computing substrate, programmers are increasingly choosing managed languages for their next generation applications. The virtual machine (VM) stands between the two, integrating them together. However, VMs are not yet capable of providing scalable performance to applications on parallel systems composed of CMPs. VMs are limited in part because key features, such as dynamic profiling, compilation, and garbage collection, are often not themselves scalable. This VM scalability bottleneck is an impediment to application scaling. For example, because VMs are oblivious to application data partitioning, they can introduce unnecessary communication such as false sharing.

This project seeks to solve this problem by designing and implementing scalable virtual machine (SVM) services. SVM includes novel dynamic profilers, just in time (JIT) compilers, and garbage collectors that are themselves scalable and support parallel applications. The project explores how to design and build a framework for parallel dynamic analysis, JIT compilation, and garbage collection that uses novel cost models of the application, analysis, JIT, and collector work to optimize for scalable high performance. In particular, analysis, JIT, and garbage collector work is divided and scheduled in novel ways to mirror application partitions and threads.

SVM will be developed within a Java Virtual Machine, but the algorithms will apply to other managed languages such as C #, JavaScript, Python, and Ruby. The investigators will make both SVM and parallel applications publicly available, adding to the national research infrastructure. CSR: Small: System Support for Planetary Scale Services This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
Increasingly, computation and storage are moving into a planetary cloud accessible across ever widening Internet pipes. Unfortunately, asynchrony, failures, and heterogeneity make it difficult to harness available computing power and storage, with inherent tradeoffs in performance, data consistency, and availability. The goal of this research is to raise the level of abstraction for building planetary scale services, in particular to: i) make fault tolerant and high performance storage a baseline abstraction for a variety of services, and ii) simplify the process of building extensible and distributed data structures that match the requirements of a range of services.
Taken together, this project has the potential to effect a qualitative shift in our ability to deploy next generation high performance and highly available network services. First, application developers will be able to leverage fault tolerant, locality aware data storage as a given for their distributed applications. Second, the research will deliver primitives to ease the problem of deploying new distributed data structures. We will provide the abstractions to manage distribution, replications, and faults in these environments.
The broader impacts of this project will include leveraging our infrastructure to conduct studies of large scale service architectures, first in advanced graduate courses and later in undergraduate courses, and a public release of the replication and extensible data structure software underlying our work for research and educational purposes. AF: Small: Parallel Methods for Large, Atomic scale Quantitative Analysis of Materials Parallel Methods for Large, Atomic scale Quantitative Analysis of Materials
Project Summary

Experimental advances in materials science have enabled near atomic scale imaging of materials, making it possible to relate atomic arrangements to macroscopic material properties and enabling smart materials design. Driven by the needs of the materials science community facing a rapid adoption of this new technology, the goal of the proposed research is to develop comprehensive algorithmic foundations for quantitative analysis of atomic scale data from interpreting atom probe tomography images to their analysis and feature extraction. Parallel algorithms and high performance implementations will be emphasized due to the large data set sizes which can reach several hundred million to a billion atoms and beyond. The investigators will develop parallel algorithms to 1) deal with inherent limitations and errors in atomic sampling, 2) determine crystallographic orientation, 3) perform autocorrelation based clustering analysis, and 4) extract coherent structures and intrinsic geometric features through linear and non linear manifold learning. The research will be carried out by an interdisciplinary team of PIs with expertise in parallel algorithms, high performance computing, computational and differential geometry, and materials science. The research is intended to advance materials science from visualization and simulation of atomistic scale solid state phenomena to its direct observation, quantitative analysis and interpretation. SHF: Small: System Theoretic Analysis and Design for Dynamic Stability of Memory Devices in Nanoscale CMOS and Beyond Proposal ID: 0917204
Pi name: Li Peng
Inst: Texas Engineering Experiment Station
Title: System Theoretic Analysis and Design for Dynamic Stability of Memory Devices in Nanoscale CMOS and Beyond

Abstract


Data storage is essential to a broad range of electronic and biological systems. Static random access memories (SRAMs) provide essential on chip data storage for many electronic applications including microprocessors, ASICs, DSPs and SoCs. There also exists a growing effort in developing engineered genetic memory circuits to facilitate new understanding of cellular phenomena in natural organisms, cellular control and biocomputing. This work intends to facilitate the understanding, analysis and enhancement of dynamic stability, a key system property, for semiconductor SRAMs, emerging memristor and biological memories. Conventional static SRAM stability metrics are unable to capture intrinsic dynamic circuit operations and hence inherently limited in their applications. This work addresses the need for a rigorous understanding of dynamic stability in semiconductor and genetic memories via a system theoretic approach. Nonlinear system theory will be exploited to construct new dynamic noise margin concepts and design metrics with theoretic rigor and design insights. Novel system theoretically motivated numerical algorithms will be developed to facilitate analysis and optimization with significantly improved efficiency.

This work will facilitate the design of nanoscale computing systems as well as the development of synthetic gene regulatory networks. Interdisciplinary explorations will provide new opportunities for solving research problems of practical significance and offer educational opportunities to students. The PIs will promote the research participation from undergraduate students and students from underrepresented groups and engage in high school teacher enrichment programs. The research outcomes will be integrated into undergraduate and graduate curriculum and disseminated in the research community and major semiconductor companies. CIF: Small: Power Consumption in Communication Power consumption is an increasingly important issue across society.
For communication, as the ranges of links in wireless networks
continue to shrink, the power consumed in the encoding and decoding
becomes a decidedly nontrivial factor in the choice of system
architecture. This is particularly important in settings such as
wireless patient monitoring, personal area networks, sensor networks,
etc. Shannon s classical information theory only established the
tradeoff between rate and transmit power as the probability of error
goes to zero and the block length goes to infinity. This research is
about giving new conceptual tools for reasoning about the power
consumption in encoding and decoding as well. The core idea is that in
the age of billion transistor chips, the proper metric for complexity
is the power consumed by the implementation.

Just as simplified channel models have enabled sophisticated analysis
that has revealed deep insights into error correction and transmit
power, this research develops simplified implementation models that
are amenable to analysis. This reveals the fundamental tradeoffs
underlying the interplay between transmission and processing
powers. Crucially, the models developed are compatible with modern
approaches to iterative and turbo decoding by massively parallel
ASICs, while also not being limited to just the currently known
families of sparse graph codes. By developing a unified mathematical
framework, this research allows us to understand the total power cost
of meeting performance objectives like high rate, low distortion, low
delay and low probability of error. This in turn leads to an
understanding of how to better engineer wireless systems as a whole:
opening up avenues for collaboration between circuit designers,
communication theorists, and networking researchers working at higher
layers. NeTS:Small:A Unified Lookup Framework to Enable the Rapid Deployment of New Protocols in High Speed Routers High end routers and switches perform performance critical lookups
into large data based on network packet content. Current designs rely
on specialized hardware modules for high performance. While mostly
sufficient, emerging new protocols create a problem; for these to be
deployed, expensive equipment upgrades are required. To support the rapid deployment of new data plane protocols, we need a flexible hardware/software framework.

This project is developing such a unified framework for future
network devices with four components: a) an abstract execution model
to represent hardware, b) a toolchain to ease implementation, d)
quantitative performance evaluation of protocols, and d) programmable
hardware architectures specialized for the core operations in network
protocols. Leveraging trends in tiled architectures, the project is
developing a specialized tiled architecture that moves computation close
to storage and thus provide efficient lookups. The project is also
investigating how these design methodologies extend beyond protocol
processing to payload inspection as required by intrusion prevention,
application identification and other functionality.

Deploying and implementing new protocols is challenging. This project
will result in specification of flexible hardware for high end
routers. This flexibility to support multiple protocols using a single
toolchain and architecture framework will reduce development costs of
future high end routers and speedup the deployment of new
protocols. Another outcome will be the open source release to
the academic community of the toolchain for developing high speed
implementation of protocols, reference implementations of various
protocols, standard data sets, and a NetFPGA implementation of the
proposed hardware architecture. III: Small: Collaborative Research: Mining and Optimizing Ad Hoc Workflows Ad hoc workflows are everywhere in service industry, scientific
research, as well as daily life, such as workflows of customer
service, trouble shooting, information search, etc. Optimizing ad
hoc workflows thus has significant benefits to the society.
Currently the execution of ad hoc workflows is based on human
decisions, where misinterpretation, inexperience, and ineffective
processing are not uncommon, leading to operation inefficiency.

The goal of this research project is to design and develop
fundamental models, concepts, and algorithms to mine and optimize ad
hoc workflows. The project includes novel research on the following
key areas: (1) Network Modeling and Structure Mining. A network model
is built that statistically captures the execution characteristics
of ad hoc workflows, and is optimized to improve the execution of
new workflows with respect to different optimization objectives.
(2) Workflow Artifact Mining. The network model built on workflow
executions is then extended with workflow artifact mining to realize
an optimization system that is able to take advantage of both
executions and text contents. (3) Role Discovery and Relation
Assessment. A computational framework is built to analyze the roles
and relationships of agents involved in ad hoc workflow executions
in order to further optimize workflows.

Advances from this project include models to represent ad hoc
workflows, algorithms for mining hidden collaborative models, and
techniques that optimize ad hoc workflow processing. The project
bridges two emerging research areas: service science and network
science, and enriches the principles and technologies of data mining.
It also enhances research infrastructure through the collaboration of
team members from different areas (data mining, database, and
network). This research is tightly integrated with education through
student mentoring and curriculum development.

Publications, software and course materials that arise
from this project will be disseminated on the project website:
URL: http://www.cs.ucsb.edu/~xyan/smartflow.htm TC: Small: Keeping Jack in the Box: Confining the Role of Untrusted Inputs in Web Scenarios A significant number of attacks on Web browsers and Web applications
are successful through the use of malicious inputs.
For instance, attacks on web browser extensions target browsers by
exploiting vulnerable extensions (add ons) by supplying malicious input.
Malicious inputs exercise unintended behaviors leading to attacks that
compromise confidentiality, integrity and
availability of web based systems. This project systematically
examines the role that user inputs play in web browsers and
applications, and develops techniques to prevent these attacks by
confining their influence. The challenge is to develop sound and
precise automated analysis mechanisms for web based platforms
such as JavaScript. Research from the areas of static and dynamic analysis,
information flow tracking and learning program behaviors
will be used to develop robust, efficient and highly precise techniques.

Papers from the project will be distributed in popular online
resources for Web security for the widest possible dissemination and
further enhancement. Furthermore, the PIs will transition results from
this research to Web development and standards communities.
The results of this project will be integrated into the new and
existing courses in the undergraduate and graduate curricula at the
University of Illinois campuses at Chicago and Urbana Champaign.
These courses will train a growing workforce of software engineers
who will be more security aware and apply these principles for
laying a platform for a more secure Web. This project will also directly
contribute to the research training of three Ph.D. students
supervised by the PIs. This project will also support the involvement
of the PIs in outreach activities aimed at K 12 teacher training and
minority groups, and contributions by designing programs that create
awareness and interest in computer science. CIF: Small: A Resource Scalable Unifying Framework for Aural Signal Coding This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111 5).


A Resource Scalable Unifying Framework for Aural Signal Coding


The objective of this project is to achieve a comprehensive, unifying framework and corresponding universal methodologies for the coding of aural signals. The challenges are due to the combination of elusive perceptual criteria, complex signal structures, and great diversity in signal type, operational setting, complexity and delay requirements. Historical efforts were tailored to narrowly defined signal types and scenarios, such as linear prediction for low rate speech communication versus transform based music coding for storage and streaming. However, there is a growing realization of their insufficiency to handle the heterogeneous aural signals and network settings encountered in many real world applications, as evidenced in particular by recent major initiatives of the multimedia and networking industries demanding joint speech audio coding standardization.


The research formalizes the tradeoffs that underly universal aural signal coding, and develops a unifying framework and methodologies to enable efficient optimization of resource scalable coding under heterogeneous signal and network setting scenarios. The main thrusts of the project are: i) Development of a unifying resource scalable framework coupled with effective perceptual distortion criteria, which covers the continuous gamut of aural signal types and networking scenarios, and is scalable in bit rate, encoding/decoding complexity and delay, etc.; ii) Theoretical analysis of rate (perceptual) distortion performance limits within such unified compression paradigms; iii) A new class of universal methodologies and effective optimization algorithms for efficient coder design and various resource allocation within this unifying framework. NeTS: Small: Parallax Leveraging the Perspective of Ten Million Peers This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Much of the Internet s growth occurs in domains beyond the reach of current measurement techniques and platforms, such as behind NAT boxes and firewalls or in regions of the Internet not exposed by public BGP feeds. This work is motivated by the observation that, collectively, peers in large scale P2P systems have a unique and valuable perspective on network conditions, one to which today s researchers, operators and users have limited or no access. P2P systems are an ideal vehicle for accessing these views, being among the largest services and covering most of the Internet. As passive monitors of the Internet, different P2P systems can provide different, complementary views of the network, over partially overlapping space and time domains. The goal of this effort is to explore the potential for reusing such valuable view, investigating techniques for gathering, sharing, and exploiting it. The work focuses on: identifying useful metrics regarding network conditions collected by these peers and evaluating their potential for reuse (over time and across multiple systems perspectives), designing approaches for maintaining this information and making it accessible to other large scale distributed systems in a decentralized manner while preserving the privacy of the participating peers, and exploring potential applications that could benefit from this information.

Access to end host views of the network will help in the understanding and characterization of the underlying network and address the needs of new emergent Internet services and applications. SHF: Small: Multi Core Architecture, Applications, and Tools Co Design Technology limitations, emerging applications, and changing usability trends are ushering in a new era in multi core computer systems. A key problem for both application and microprocessor design is that applications are largely evolving independently from the architectural development of microprocessors. This is a problem for architectural research because developing efficient architectural solutions requires realistic characterization of the next generation applications. From a system design perspective, understanding application behavior is crucial for building an efficient system, since they must be optimized to exploit mechanisms provided in the architecture. This proposal seeks to re think next generation multi core systems both software
and hardware architectures using state of the art quantitative design
tools.

The two key ideas explored in this research are the following. First, is a hybrid task level/data level parallelism execution model for emerging applications that have abundant but not synchronization free parallelism. Second, is the development of new highly accurate and efficient quantitative models to evaluate architecture and application design alternatives, at scale and over a wide range of application workloads. The project seeks to provide a suite of quantitative tools to close the development loop of design, evaluation and analysis of software s behavior on hardware, allowing the tuning of both software and hardware. This project takes real time graphics as a challenge application and derives a full system, called Copernicus, for future real time graphics that can provide significantly higher image
qualities. The project will also integrate these quantitative models in the curriculum and disseminate to the research community.

The innovations proposed in this research have the potential for significantly aiding microprocessor and application development for future systems. The development of a full system specification for ray tracing can trigger an inflection in the evolution of both programmable processors and
specialized graphics processors. NeTS:Small: Wifu: A Software Toolkit for Wireless Transport Protocols The use of wireless networks is expanding, including mesh networks that increase the geographic coverage of an access point, and ad hoc networks that provide instantaneous and self configuring communications when wired infrastructure is not available. The viability of these networks is hampered, however, by the poor performance of TCP when it operates over multiple wireless hops. Many alternative transport protocols are being designed, but the lack of
experiments on deployed networks severely limits the impact and relevance of current research in the area.

This project designs a software toolkit that simplifies the process of developing wireless transport protocols. Protocols can be written in user space, enabling researchers to avoid the complexities of kernel development and focus their efforts on experimental evaluation. The project uses the toolkit to implement a number of leading wireless transport protocols, comparing their performance, examining the decomposition of functionality between end hosts and routers, and evaluating TCP compatibility. The end result will be a comprehensive study of wireless transport protocols in an experimental setting,
demonstrating how to obtain high throughput while working seamlessly with existing TCP stacks.

The toolkit will be developed as an open source project, so that it can be used by the research community to develop new transport protocols or to examine the interaction of wireless transport protocols with new MAC designs, network coding, or routing protocols. The project will also develop curriculum for incorporating a mesh testbed into a senior level course in networking, with accompanying labs and instructional materials. NeTS: Small: Beyond Listen Before Talk: Advanced Cognitive Radio Access Control in Distributed Multiuser Networks The award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Cognitive radio (CR) has the potential to improve spectrum utilization and expand wireless communication services by opportunistically utilizing underutilized spectrum bands. This project designs advanced cognitive radio access and power control algorithms that can achieve better spectrum efficiency while limiting interference to primary communications. Moving beyond the more traditional access strategies that rely only on secondary user (SU) spectral sensing to avoid collision with primary users (PUs), this research exploits various levels of primary network?s data link control (DLC) signaling and feedback information. Such DLC information is available in many practical wireless systems, such as transmission profile, receiver ACK/NACK, channel quality indicator, and power control information. Utilizing such information elevates the level of SU cognition. It provides more efficient spectrum sharing, better PU protection (especially in the presence of multiple distributed SUs), and multiple levels of SU and PU interaction. The major outcomes include: 1) Distributed multi SU cognitive access and power control based on PU receiver feedback information; 2) Optimal algorithms for distributed multi SU access control in multi channel cognitive environments; 3) Cognitive radio access robust to PU behavioral changes and incomplete PU feedback information; and 4) Hierarchical cognitive radio networks of users with varying degrees of cognition. This significantly broadens the future applications of wireless services in areas with limited open spectrum. The plan recruits students, especially from under represented groups, and integrates the results into the classes for computer science and electrical engineering majors. SHF:Small: Portable High Level Programming Model for Heterogeneous Computing Based on OpenMP High end systems and general purpose computers alike are increasingly being configured with hardware accelerators. A typical system, or node of a large scale platform, contains a few multicore processors that share memory, together with one or more coprocessors, such as a high end accelerator board, an inexpensive programmable graphics card (GPGPU), or an FPGA (Field Programmable Gate Array). The coprocessors will typically have a different instruction set, distinct memory, a different operating environment and markedly different execution characteristics than the multicore component of the system. As a result, these heterogeneous platforms pose tough challenges with regard to their programmability.


The goal of this research is to significantly simplify the process of developing code for heterogeneous platforms by providing a single, high level programming interface that may be used across, and within, multicore processors and a broad variety of accelerators. Language features will be designed, in the form of an extension to the industry standard OpenMP API, that will enable the application developer to specify code regions for acceleration, along with the necessary synchronization and data motion. The implementation will translate the resulting enhanced OpenMP for a variety of heterogeneous platforms. AF: Small: Topological Graph Theory Revisited: With Applications in Computer Graphics Recent research in computer graphics has shown that the classical topological graph theory provides a solid mathematical foundation and powerful tool for the development of 3D modeling systems. Despite its initial success, there is still a significant gap between the theoretical research in topological graph theory and its direct applications in computer graphics. In particular, the research in classical topological graph theory has largely neglected geometric issues. Moreover, very recent research has shown exciting connections between graph embeddings on non orientable surfaces and surface weaving, and demonstrated a need for a refined and extended study in this direction.
This proposed research, guided by its applications in computer graphics, will refine and extend the classical topological graph theory in the following two directions:
1. Graph embeddings on orientable surfaces with geometric constraints: This project will re examine the fundamental issues studied in graph embeddings on orientable surfaces that are related to topologically robust 3D modeling, by considering geometric constraints such as symmetry, planarity and conical properties. Applications of this research include development of topologically robust and highly interactive graphics modeling systems.
2. Graph embeddings on non orientable surfaces and their applications in modeling surface weaving: This project will refine and extend the study on graph embeddings on non orientable surfaces and the corresponding graph surgery operations, study their relations to 3D modeling, and build a new paradigm of modeling surface weaving. Applications of this research include creating beautiful shapes such as woven basket and topological sculptures. CSR:Small:Estimating the End system Network I/O Bottleneck Rate to Optimize Transport Layer Performance This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5). The bottleneck for the transferring data at very high speeds often turns out to be the end system performance. In the absence of definitive knowledge about the workload at the receiving end system, the sender s transmission rate often overshoots the critical bottleneck rate of the receiver. This typically results in oscillations between the extremes and poor performance. To optimize the performance of the transport protocols and achieve the important flow control functionality, it is important to estimate the receiving end system effective bottleneck rate. In this project we will use modeling and active analysis of the end system to estimate this rate. We will develop queueing network models for representing the different multicore and multiprocessor end systems running different types of workloads. We will develop a software tool to be integrated with existing transport protocols. We will carry out experimental analysis for different types of end systems with different configurations and workloads. We will apply and extend methods that have been proposed to address the limitations of queueing network models for performance analysis of computer systems with bursty workloads and correlated service times. The software tool will be made available to the research community to analyze and optimize distributed applications and systems. The research project will provide a framework to train graduate and undergraduate students in both analytical and experimental methods, and develop knowledge and intuition about next generation computer systems and distributed applications. RI: Small: Learning Biped Locomotion In a not too distant future, assistive robots will become a natural part of the human society, in hospitals, schools, elder care facilities, inner city urban areas, and eventually homes. While wheeled robots, e.g., a humanoid torso on a mobile platform, can cover a range of tasks that assistive robots will be needed for, eventually, legged robots will be the most suitable, as legs increase the effective workspace of a robot and allow maneuvering more complex terrains like steps, curbs, and cluttered and rough terrains in general.
This project investigates biped locomotion with a Sarcos humanoid robot. In contrast to most other projects in biped locomotion, it emphasizes walking over uneven and rough terrain, obstacle avoidance, recovery from unexpected perturbation, and learning methods for motor control, as these issues seem to be the most important for working in dynamic and partially unpredictable human environments. Another focus is on dexterous movement control, i.e., control with a maximal amount of compliance and minimal negative feedback gains, using advanced operational space controllers with internal model control. Dexterous, compliant control will increase safety of the robot when accidentally impacting with humans or obstacles, and it will also allow the robot to recover more easily from external perturbation simply by ?giving in?. Such a control approach requires departing from the traditional high gain position controlled humanoid systems, and focuses on torque control, reactive instantaneous control instead of finite horizon optimization, as well as efficient motion planning and learning methods. SHF: Small: Simplifying Reductions This project develops compile time techniques and a software tool (called the Reduction Simplification Engine, RSE) for optimization of equational programs with reductions (associative, and usually commutative, operations applied to collections of data). By using these techniques and tools it is possible to generate programs with lower asymptotic complexity than the original specification. Such complexity reduction is an ambitious, almost unheard of, goal in compilation: most compilers seek constant factor gains, usually a few percentage points. The techniques developed in this project build on more than twenty years of research by the PI on a formalism called the polyhedral model. The main novelty of the current effort is that in addition to polyhedral techniques, algebraic properties such as idempotency and distributivity are also used to augment the analyses performed. SHF: Small: A Foundation for Effects Modern programming languages provide sophisticated control mechanisms, making them suitable to program a wide variety of applications. It becomes a challenge for a language designer to bring together these features without creating inconsistencies or for a programmer to understand how to use them. The proposed work aims to provide a sound and robust framework for reasoning about a variety of control mechanisms in isolation and more importantly about the complex interactions among them.

The Curry Howard isomorphism will guide the development of such a framework, which entails connecting programming languages to logic. The connection to logic allows one to formalize questions regarding expressive power and to exchange and borrow results with the field of proof theory. New logics expressing dynamic properties directly, as opposed to through program transformations, will be investigated. An advantage of this approach is that compilation, execution, optimization and code safety are expressed within the same foundation level.

The study will have a direct impact on how to reason about security properties, since it provides methods to reason about dynamic properties in a continuously changing context. It will also have an impact on program verification, since it provides a better understanding of the invariants that programs should satisfy. NeTS: Small: Topology Switching for Data Centers and the Clouds Above This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

This project addresses the complex networking challenge presented by the emerging cloud computing model. Cloud providers must run a diverse set of client applications, each with potentially different networking demands, on shared data center facilities. Traditionally, a datacenter network is configured to use the same routing process to choose the best route for each flow in a datacenter, regardless of the application. For example, Ethernet frequently performs shortest path routing along a single spanning tree. Yet data center networks typically exhibit significant redundancy; routing along a single tree leaves many paths unused, sacrificing potential gains in reliability, isolation and performance.


Topology switching moves beyond this one size fits all approach providing an architecture for fine grained multi topology networking. It allows applications to create custom routing systems within a data center; they can configure multiple logical topologies that, together, are tailored to their reliability and performance requirements. From a cloud provider s perspective, a topology switched network increases efficiency by multiplexing potentially hundreds of topologies across the same shared physical network. The PIs are designing a scalable topology switched routing platform that facilitates the exploration of application interfaces, management challenges, novel routing strategies, and performance benefits of this approach.


Ultimately, the project aims to develop a flexible topology management primitive that improves administrators ability to effectively manage extremely large datacenter deployments. The research is also analyzing the benefits and costs of multi topology networking. Additional outcomes of this proposal include a public release of the topology switching platform, enabling academic and industrial feedback and adoption. CIF: Small: Cognitive Femtocells: Breaking the Spatial Reuse Limits of Cellular Systems Cognitive femtocells: Breaking the spatial reuse limits of cellular systems

Abstract:

The next generation of wireless cellular systems will be data traffic driven, providing seamless connectivity to the Internet and its services. In the areas of information and communication technologies, cellular systems and the Internet have proven to be the most transformative technologies for society. This research endeavors to optimally marry these two technologies with the goal of dramatically increasing achievable data rates and coverage. The proposed solution is to deploy very small cellular access points in residential homes and offices. These ``femtocells are connected to the network via existing DSL/Cable and do not require additional deployment of costly wired infrastructure. Several theoretical challenges are posed by the femtocell concept. In particular, due to lack of coordination with the rest of the network, femtocells interfere with the network itself. This research develops novel solutions for femtocell technology by exploiting ideas from cognitive radio, based on intelligent opportunistic usage of the shared radio resource. Femtocell base stations will be deployed without careful frequency planning and will react to the interference environment by adapting their signaling strategy, thus, in the vernacular of cognitive radio, the cellular system plays the role of the ?primary? user. Unlike the classical cognitive radio scenario, the cellular signaling protocols and coding schemes are known. In this research we build on recent results on the Gaussian
interference channel and exploit the signal strength imbalances due to path loss typical of cellular networks. Thus, the system is designed so that it operates predominantly in the regime of either weak or strong interference. Advanced signal processing and channel coding are exploited to utilize the fine structure of primary (cellular) signals, to explore the fundamental costs of cognitive radio and to develop coordination strategies to minimize this overhead. The goal of this novel design is to achieve a dramatic improvement in spatial reuse, allowing future cellular systems to achieve data rates comparable to wireless local area network while retaining the seamless connectivity and mobility of cellular networks. SHF: Small: RUI: Observationally Cooperative Multithreading Modern processors are designed to perform more tasks simultaneously, rather than to perform single tasks more quickly. These new multicore processors are powerful, but using that power is challenging; interesting problems often divide irregularly, requiring difficult and error prone coordination among subtasks. Consequently, parallel programming is considered hard to learn and harder to do. Observationally Cooperative Multithreading (OCM) is a new approach. In programs written for cooperative multithreading (CM), subtasks take turns and execute one at a time. The CM model is well known to rule out conflicts and to simplify programming. OCM takes these same programs but runs them on modern multicore machines, executing subtasks simultaneously when there are no conflicts. The result can be a speed and resource utilization benefit with no extra complexity for programmers. Potentially, OCM could make concurrency more accessible to a broad audience, including introductory students. The research will develop OCM implementations using techniques such as Transactional Memory and Lock Inference, with the aim of fostering adoption of OCM by a large user community. Realistic benchmarks will be constructed to analyze the speed and scalability of OCM implementations, and to verify ease of programming in the OCM model. III COR Small: Towards More Flexible, Expressive and Robust Stream Systems This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
Current stream engines are mostly stand alone systems, whereas in many applications, stream processing will be one component of a larger information system. Coupling with other components, such as user interfaces, transactional data and archives will be increasingly important. For stream queries to be robust over extended execution periods, they must have the means to adapt to both internal and external changes. Changes in input rates, time lags and data distributions can cause shifts in internal memory and processing loads. Operators must adapt to these changes both local and in concert with other operators. Variations in client demands create opportunities for improved resource use to which operators must adapt. For a stream processing system to be robust in the face of changing workloads and possible system faults, the architecture must have levels of flexibility and adaptivity not currently existing. The team proposes three approaches to developing the necessary flexibility. They will use a formal analysis that will provide precise notions of time and progress, in order to provide criteria and metrics for a variety of situations. In addition, they will elaborate operator and architecture design activities that will then be implemented and evaluated in two different data stream systems. In addition to faculty at Portland State University and graduate students there, the team has an ongoing collaboration with AT&T Research, who will provide access for testing the new system. The techniques developed on this project will broaden the number of applications that can be reasonably served with data stream systems, and by working with commercial systems, the adoption of those techniques into next generation products will be accelerated. In addition to a new course in steam systems, the team works with high school interns each summer, who are recruited through the Saturday Academy program at Portland State. TC: Small: Anchoring Trust with a Verified Reference Kernel Many complex software based systems must be certified in order
to be deployed in safety critical environments. Modern software
certification processes require trustworthy evidence supporting the
verification claims. While verification tools have made tremendous
gains in power, they lack the ability to generate concise and
independently checkable evidence. The V Kernel project develops a
practical approach to reconciling trust and automation. The claims
generated by the fast but untrusted front line verification tools are
certified offline by slower but verified back end checkers. The
front line analyzers can provide hints and certificates that assist the
back end tools. The verification of the checkers can be carried out by
untrusted tools as long as the end result can be independently
certified. Our approach does not constrain front line tools by
requiring them to produce proof objects. A wide variety of verification
tools including SAT solvers, decision procedures, model checkers, static
analyzers, and theorem provers can be validated using this approach. III: Small: Do It Yourself forms driven workflow web applications Emerging Do It Yourself database driven web application
projoect aims to (1) enable non programmers to rapidly build custom
data management and workflow applications and (2) to promote
a novel pattern of interaction between application owners and
programmers. Their beneficiaries are organizations, in need of long tail
web applications, that cannot afford the time and money needed to
engage into the conventional code development process.

Do It Yourself platforms must maximize two metrics that present
an inherent trade off: the simplicity of specification and the
application scope, which characterizes the class of applications
that can be built using the platform?s specification mechanism.
The proposal introduces two scopes with interesting trade off
features. Namely, in the All SQL scope and the (more limited)
forms driven workflows scope each application page consists
of a report (modeled by a nested query) and forms and actions in the
report s context, leading to updates. The limitations have practically
minor effects on the scope but they enable simple specification and
automatic optimizations, studied in the proposal, such as:
1. automatic creation of reports by choosing between the candidates
using information theoretic criteria relying on constraints captured in
the limited models.
2. summarization of applications as workflow specifications by
analyzing the dependencies between updates and queries. Vice versa,
the proposal shows how simple workflow primitives translate to queries
(reports) and updates (forms and actions).
The proposal also provides an unlimited model of web applications,
where programmers introduce code components and interface them
with the limited part via queries and updates.

The proposed models of database driven web applications will impact
the education of both Computer Science (CS) and non CS students
that need to comprehend web applications at a high conceptual level.
For further information see the project web page at
http://www.db.ucsd.edu/forward NeTS: Small: Collaborative Research: Multi Resolution Analysis & Measurement of Large scale, Dynamic Networked Systems with Applications to Online Social Networks Many large scale networked systems such as Online Social Networks (OSNs)are often represented as annotated graphs with various node or link attributes. Such a representation is usually derived from a snapshot that is obtained through measurements. These graph representations enable researchers to characterize the connectivity of these systems using graph analysis. However, captured snapshots of large networked systems are likely to be distorted. Furthermore, commonly used graph analysis characterizes the connectivity of a graph in an indirect fashion and generally ignores graph dynamics.

This multi disciplinary research program designs, develops and rigorously evaluates theoretically grounded techniques to accurately measure and properly characterize the connectivity structure of large scale and dynamic networked systems. More specifically, the project examines various graph sampling techniques for collecting representative samples from large and evolving graphs. It also investigates how multiscale analysis can be used as a powerful technique to characterize the key features of the connectivity structure of large dynamic networked systems at different scales in space and time. The developed techniques will be used to characterize fundamental properties of the friendship and various interaction connectivity structures in different OSNs.

This project promises to identify the underlying technical and social factors that primarily drive the structural properties and dynamic nature of OSN specific connectivity structures. It will produce new models for friendship and interactions in OSNs, a large archive of anonymized datasets and new tools for OSN measurement, simulation and analysis. The latter will be incorporated into newly developed courses in Computer Science and Sociology, and will be freely distributed. SHF:Small: A CAD Framework for Coupled Electrical Thermal Modeling of Interconnects in 3D Integrated Circuits The semiconductor industry is at an interesting crossroads, where traditional scaling of CMOS devices is beginning to confront significant challenges that are threatening to derail the more than four decades old Moore?s law. 3D integrated circuits (3D ICs) offer an exciting alternative, where in lieu of scaling, continuous increase in functionality, performance and integration density can be sustained indefinitely by stacking semiconductor layers on top of each other in a ?monolithic? manner. A 3D IC is comprised of two or more active (semiconducting) layers that have been thinned, bonded and interconnected using special vertical wires drilled through the active layers known as ?Through Silicon Vias (TSV)?. When TSVs (10 100 micrometer long) are used to replace the longest (several millimeters) on chip horizontal wires as well as some chip to chip connections (on printed circuit boards), significant reduction in wire delay and chip power dissipation can be achieved. Moreover, 3D ICs also offer the most promising platform to implement ?More than Moore? technologies, bringing heterogeneous materials (Silicon, III V semiconductors, Graphene, etc) and technologies (memory, logic, RF, mixed signal, MEMS, optoelectronics, etc) on a single chip.

However, modeling and analysis of interconnects in 3D ICs present new and significantly more complex problems. In contrast with traditional interconnects, the modeling of high aspect ratio TSVs embedded in a semiconducting material with non uniform currents in the third dimension, and electromagnetic coupling of interconnects with multiple conductive substrates at high frequencies, constitute new challenges for design and design automation methods. Furthermore, the high power density in 3D ICs due to multiple active layers and their limited heat removal options give rise to large three dimensional thermal gradients, making it important to consider the coupling between thermal and electromagnetic properties of interconnects and the surrounding media. Finally, the need for accuracy is accompanied by the computational challenge of handling a large number of coupled interconnects at the system level, as 3D integration further exacerbates the size of the interconnect problem.

This project will develop the necessary foundations for coupled electrical thermal modeling and analysis of interconnects and passives in 3D ICs, considering the electromagnetic coupling of general 3D interconnects with multiple substrates at ultra high frequencies as well as the physical attributes of high aspect ratio TSVs (including geometry, material and density), using thermally aware and inherently efficient techniques to enable full chip modeling of the large system of interconnects. The overall program also ties research to education at all levels besides focusing on recruitment and retention of underrepresented groups in nanoscience and engineering. TC: Small: Runtime and Static Analysis for Web Application Security Web applications are prevalent and enable much of today s online
business including banking, shopping, university admissions, and various
governmental activities. However, the quality of such applications is
usually low, and they are increasingly popular targets for attack. This
project aims at developing practical testing and analysis mechanisms and
tools to secure web applications. In particular, it focuses on
developing novel, principled techniques to address the following
research issues: (1) how to formalize security threats in web
applications; (2) how to provide runtime security for deployed
applications via dynamic monitoring; and (3) how to provide static
security enforcement during application development and testing.

The project is interdisciplinary, touching upon a number of requisite
areas including computer security, software engineering, and programming
languages. It has the potential to advance knowledge in each of these
disciplines with novel formulations of security requirements, systems
concepts, and advanced testing and analysis techniques. The project
also has the potential for significant industrial and societal impact.
Through the proposed research, education, and outreach activities, the
project will empower web application developers with the knowledge,
methodologies, and development tools to build secure web
applications. Testing and analysis tools developed in the project will
also be distributed to other institutions and the industry for teaching,
research, and experimental evaluation. CSR: Small: Managing and Indexing Exascale Archival Storage Systems This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
As our society stores ever increasing quantities of digital data and abandons analog storage, we must be able to manage, organize, and preserve digital data for decades or longer. Maintaining access to large scale digital archives will be critical in enabling future generations to access the medical records, personal information, photographs, and other data that we are generating and storing digitally. However, current approaches to archival storage are ill suited to long term preservation because they do not cope well with evolution of device technologies and rely upon centralized search indexes that do not scale well and are prone to failure.
This project is exploring techniques that allow the management of large scale archives containing 105?106 intelligent power managed storage devices connected by a network. These techniques allow seamless evolution of the archival storage system by integrating new devices and removing old, inefficient devices. The research is also developing approaches to index data in the archive, allowing users to quickly find the data they need by maintaining indexes on each device and routing queries to the devices that might contain relevant data. These techniques will allow archives to scale to hundreds of thousands of devices, allowing them to contain the vast digital legacy we are leaving to our descendants.
This research will guide the design of archives that are power efficient and can gracefully evolve as storage technology changes. Additionally, the project will train both undergraduates and graduate students in the problems facing digital archiving, an area of critical importance. RI: Small: Kernelization with Outer Product Instances Thus far kernel methods have been mainly applied in cases where observations or instances are vectors. We are lifting kernel methods to the matrix domain, where the instances are outer products of two vectors. Matrix parameters can model all interactions between components and therefore take second order information into account. We discovered that in the matrix setting a much larger class of algorithms based on any spectrally invariant regularization can be kernelized. Therefore we believe that the impact of the kernelization method will be even greater in the matrix setting. In particular we will show how to kernelize the matrix versions of the multiplicative updates. This family is motivated by using the quantum relative entropy as a regularization. Most importantly we will use methods from on line learning to prove generalization bounds for multiplicative updates that grow logarithmic in the feature dimension. This is important because it lets us use high dimensional feature spaces.

We will apply our methods to collaborative filtering. In this case an instance is defined by two vectors, one describing a user and another describing an object. The outer products of such pairs of vectors become the input instances to the machine learning algorithms. The multiplicative updates are ideally suited to learn well when there is a low rank matrix that can accurately explain the preference labels of the instances. The kernel method greatly enhances the applicability of the method because now we can expand the user and object vectors to high dimensional feature vectors and still obtain efficient algorithms. AF: Small: Learnability, Randomness, and Lower Bounds This project is motivated by new connections between the research fields of computational complexity theory and machine learning theory. Computational complexity theory aims to understand which problems can be solved efficiently on a computer by determining the amounts of computational resources such as CPU time, memory space, or circuit area that are required to solve problems. At the center of this field is the famous P vs. NP question which impacts virtually every scientific and engineering discipline, given the thousands of diverse NP complete problems that have been discovered. Machine learning theory studies the extent to which computers can learn from data and their ability to make future predictions and classifications based on what has been learned. Some powerful learning algorithms have been discovered, but whether computers can be programmed to accomplish many learning tasks remains an open question.

Both computational complexity and machine learning aim to understand the capabilities and limitations of computation, but the two fields study different types of problems and use different kinds of techniques. This research will employ techniques and ideas from each of these two fields to impact the other field, specifically with the goal of proving lower bound results. This research will be accomplished by making use of a new vantage point provided by algorithmic randomness to relate complexity and learning problems. Learning algorithms will be utilized to establish lower bounds on the computational resources required to solve problems in computational complexity. The converse direction will be investigated to apply techniques and ideas from computational complexity to show that attribute efficient learning algorithms do not exist for certain concept classes. Algorithmic randomness and Kolmogorov complexity will be used to improve our understanding of the capabilities and limitations of learning algorithms.

This research will improve our understanding of computational complexity, which is informative to many areas of science and engineering where computation plays a role. This project aims to better understand what learning tasks can be accomplished efficiently by computers, which has applications to the foundations of artificial intelligence. In particular, this research will identify new obstacles that must be overcome in order to design successful automatic learning systems. A greater synergy will be developed between computational complexity theory and machine learning theory, with the benefit of laying a foundation for future collaboration and interdisciplinary work across these fields. NeTS:Small: Mapping, Attacking and Defending the BitTorrent Ecosystem The BitTorrent Ecosystem includes millions of BitTorrent peers, hundreds of active trackers, and dozens of independently operated torrent discovery sites. The Ecosystem is further fueled with distributed trackers using Distributed Hash Table (DHT) and Peer Exchange (PEX) functionality. The Ecosystem also includes ?interdiction companies, which attempt to curtail the distribution of targeted content.

Despite its importance, both in terms of its footprint in the Internet and the influence it has on emerging P2P applications, the BitTorrent Ecosystem is only partially understood today. Many communities (including P2P researchers and developers, ISP researchers and engineers, copyright holders, and law enforcement agencies) would like to have a comprehensive and in depth understanding of the BitTorrent Ecosystem, as well as tools for mapping the Ecosystem in the future.

In this context, the PI and his graduate students are exploring two inter related research directions. First, they are developing tools and methodologies for comprehensive exploration and mapping of the entire BitTorrent Ecosystem. Second, they are examining of how the Ecosystem can be attacked and defended.

The expected results for the mapping research include: new public domain tools and methodologies for mapping and analyzing the Ecosystem; a comprehensive mapping data set, which will be more than an order of magnitude larger than any existing data set and will essentially cover all trackers and peers in the public Ecosystem; novel estimation methodologies based on importance sampling, incorporating measurement samples from both centralized and distributed trackers.

The expected results for the BitTorrent attack/defense research include: measurement and evaluation methodologies of ongoing attacks from ?interdiction companies ; an in depth study of the seed attack, whereby the attackers attempt to prevent the initial seed from distributing the file into the Ecosystem; machine learning algorithms for defending against the pollution attack in BitTorrent; and tractable deterministic and stochastic models for the dynamics of BitTorrent attacks, providing critical insight into BitTorrent vulnerabilities. MRI: Acquisition of a High Performance Computer Cluster Supporting Computational Science Research and Learning Proposal #: CNS 09 22644
PI(s): Robila, Stefan A.
Institution: Montclair State University
Title: MRI/Acq.: High Performance Computer Cluster Supporting Computational Science Research and Learning

Project Proposed:
This project, acquiring a high performance computer cluster, aims to expand interdisciplinary projects in three departments (CS, Math Science, and Linguistics) and enable collaboration with a research intensive university (Syracuse). These projects require significant computation power and describe problems where the size of the data sets continues to grow. The work involves novel approaches for parallel image and signal processing, and in particular spectral imaging. The approach includes design of parallel algorithms and the development of theories on how to parallelize the general data processing steps. Other projects use parallel processing for phylogenetic modeling, use dynamic systems for disease modeling, parallelization techniques for the use of scattering theory, and the cross lingual morphosyntactic tagging.

Broader Impacts: This project helps to attract more students to hands on multidisciplinary research and to establish and strengthen a cross institutional research partnership between an undergraduate serving teaching institution and a research university. The research on spectral imaging parallel applications allows for the use of spectral sensors in new applications improving the timeliness and accuracy of results. In terms of dynamic systems, computer modeling and computational analysis of disease dynamics might contribute to determine the best policy for a given epidemic situation. Applications and contributions of scattering theory to science include deep earth seismology, exploration of underground resources, engineering, mine detection, and other military applications. In phylogeny, the ability to reconstruct optimal evolutionary trees based on objective criteria impacts directly the understanding of the relationships among organisms, human evolution, and spread of infectious disease. MRI: Acquisition of High Speed Network Infrastructure for High Performance Distributed Computation in the Colleges of Science and Engineering This award will fund the purchase of a Cisco 10 Gigabit/s switch and wavelength division multiplexing equipment to construct a three way high speed data link between science and engineering research groups affiliated with the Computational Science Research Center (CSRC) at San Diego State University (SDSU), the TeraGrid hub at the San Diego Supercomputer Center (SDSC), and the California Institute for Telecommunications and Information Technology
(CalIT2) at the University of California, San Diego (UCSD). This network infrastructure will provide core data transmission resources for research groups at SDSU in physical chemistry, geophysics, astronomy, bioinformatics, thermal science, combustion engineering, fluid dynamics, nonlinear dynamics, ocean modeling, and computer science who routinely use SDSC and CalIT2 computing resources. The wavelength division multiplexing equipment will enable rapid network transactions between the two campuses, thereby allowing researchers at SDSU to rapidly transfer terabytes of data generated on TeraGrid clusters back to SDSU for local visualization and post processing.
The core 10 Gigabit/s switch will initially interconnect server room facilities at SDSU in two separate colleges: the College of Science and the College of Engineering. The switch will also provide several additional 10 Gigabit/s interfaces for individual research laboratories to connect and thereby form a high speed research backbone network on the SDSU campus. Research projects that will immediately make use of this network infrastructure include (1) the development of a service oriented architecture/cyberinfrastructure of chemical equilibrium and kinetic services used to model gas turbine NOx emissions in syngas combustion, (2) the development of a cyberinfrastructure for the General Curvilinear Ocean Model (GCOM),(3) computational quantum mechanical studies of floppy reactive intermediates, (4) the numerical investigation of high order particle source in cell (PSIC) methods for the simulation of pulse detonation engines, and (5) the development of PetaShake, a petascale implementation of the TeraShake parallel Anelastic Wave Model (AWM) code that simulates earthquake scenarios using several billion grid points. These projects currently place heavy demands on the existing SDSU campus network. SDSU researchers now require greater throughput and shorter delay when exchanging datasets and this grant will provide a much needed improvement to SDSU UCSD network connectivity by funding the necessary equipment to facilitate rapid data transfer between the two universities.
In addition to faculty and graduate student use, this network infrastructure will be used by undergraduate students working on projects that include (1) senior thesis research projects in computer science and computer engineering, (2) programming projects in courses such as computer networking, high speed network design, client server programming, and distributed computing, and (3) undergraduate research carried out under existing, externally funded fellowship programs. SDSU occupies a strong position for making these opportunities available to minorities underrepresented in science and engineering research, as one of the top ten campuses in the nation granting bachelor s degrees to ethnic minorities. MRI: Acquisition of Equipment to Develop an Energy Efficient and Reliable Wireless Sensor Network for Urban Landscape Irrigation Management System Proposal #: CNS 09 22888
PI(s): Sun, Bo
Makki, Kami; Osborne, Lawrence J.
Institution: Lamar University Beaumont
Beaumont, TX 77705 5748
Title: MRI/Acq.: Acq. of Equipment to Develop an Energy Efficient and Reliaable Wireless Sensor Network for Urban Landscape Irrigation Management System
Project Proposed:
This project, designing and implementing an energy efficient and reliable Wireless Sensor Network (WSN) for Urban Landscape Management System (ULIMS), aims to provide significant cost savings and alleviate urban water shortage problem by combining WSNs, data management, system integration, and web based delivery. The constructed WSN will consist of a long lived, real time reliable network with remote control capability. Based on engineer portable, low energy sensor nodes that can provide sensed data, the resulting system (W ULIMS) addresses fundamental constraints faced by WSNs deployed in outdoor harsh environments, including energy supply, limited memory, the need for unattended operation for a long period of time, and the lossy and transient behavior of WSN communication. The project presents the following activities:
Randomized scheduling with a connectivity guarantee,
Extremely low duty cycle data forwarding, and
Remote control (using the idea of Trickle).
The randomized scheduling with connectivity guarantee and extremely low duty cycle data forwarding protocols under study are expected to collect sensor data in an energy efficient manner. The testbed provides an infrastructure to study practical estimation, data collection and dissemination, and energy conservation faced by WSNs in harsh environments.

Broader Impacts: This project applies new technologies to traditional agricultural environments and other applications while advancing the complex behaviors of WSNs. It should contribute to alleviate the significant water shortage problem facing many cities by providing cost savings through efficient usage of agricultural labor and timely notification of situations requiring managerial decisions. Furthermore, it establishes an environment to broaden student?s knowledge and research experience. MRI: Acquisition of a Supercomputing Cluster for Data Intensive Computational Research using the Open Science Grid (OSG) Cyberinfrastructure This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Bellarmine University will acquire a 64 node Supercomputing Cluster with approximately 22TB of RAID6 disk storage space to conduct grid enabled computational research in High Energy Physics using the Open Science Grid (OSG) cyberinfrastructure. The collaborative research activities will be carried out in close partnership with University of Oklahoma, State University of New York at Albany, Purdue University at Calumet, Southwestern University, and a Historically Black Minority Serving institution, Benedict College. Using the grid middleware resources, this partnership will essentially create a virtual organization that will link all the research participants by bringing together geographically and organizationally dispersed computational resources and people from the six institutions from six different states by developing a platform for shared systematic use of the supercomputing cluster and effective utilization of computational tools. We plan to search for the Higgs Boson and SUSY (Supersymmetric) particles in specific decay channels using the LHC (Large Hadron Collider) data. The physics studies will first be carried out and tested on detailed simulations of particles and their interactions with the detector using the GEANT monte carlo package. This multi institutional research team will focus on the development of evolutionary algorithms and computational code to carry out the complex data analyses tasks by studying the various detector parameters focusing on data reduction and filtering methodologies and pattern recognition techniques for extracting ?discovery level signals? from petabyte scale datasets.

The availability of a supercomputing cluster will serve as a powerful stimulus for the recruitment of undergraduate students in the sciences, especially from the underrepresented groups. The computational research activities with OSG will not only have a significant and a broad societal impact on the educational training of undergraduate students of all races and ethnicity at Bellarmine University, but also has the potential to bring in a new era of grid enabled scientific research capability at this institution via infusion of this new knowledge. The grid enabled supercomputing cluster will create precisely the kind of exciting learning environment that will provide the next generation of students at all seven participating institutions with an accessible entry to research and education at the frontiers of computational science, information technology, and cluster/grid computing. These trained students will acquire a solid computational background with broad technical skills and expertise in high performance parallel cluster computing techniques and grid middleware technologies. Using this supercomputing facility, we plan to incorporate the grid enabled research activities into the academic curriculum by developing an innovative interdisciplinary course in ?Computational Science using Grid Technology.? Additionally, we plan to host a hands on OSG Summer Grid Workshop/School at Bellarmine University for the undergraduate students from the regional public, private, and minority serving institutions in Kentucky. MRI: Acquisition of a high performance computer cluster for the computational study of complex chemical systems: from small molecules to biological nanomachines The best work in theoretical chemistry combines innovative creation of new mathematical methods, the development of new algorithms for the computational implementation of these methods, and large scale, computer intensive simulations on systems of key importance in chemistry in its broadest sense. Excellence in all three of these area characterizes the work of Millard Alexander, John Weeks, Daniel Kosov, Christopher Jarzynski, and Dev Thirumalai of the University of Maryland. Their work covers chemistry and chemical biology in the broadest sense, ranging from including non adiabatic effects in small molecule dynamics (Alexander), thermodynamic properties of complex systems (Weeks), non equilibrium electron transport through nanostructuctures and molecules (Kosov), thermodynamic properties of complex systems (Jarzynski), and the mechanism of ribosome assembly (Thirumalai). The principal investigators have already established transformational, forefront research programs in these areas. In addition, several of the principal investigators have contributed to the development of state of the art software packages in their individual areas which are now being used in many research groups external to the University of Maryland. A major challenge in theoretical chemistry and chemical biology is the extension of accurate modeling to systems of larger complexity and dimensionality. The computational resources currently available to the principal investigators are either outdated or oversaturated, and thus have become insufficient for their research and training activities. The present MRI proposal will allow purchase of a new cluster, based on dual core, multi cpu server nodes.
This insight gained from the proposed research will have potential application in many areas of chemistry and chemical biology, ranging from combustion, to atmospheric chemistry, to complex fluids and interfaces, to current flow in nanomaterials, to the understanding of thermodynamic equilibrium in complex systems and to the functioning of biological nanomachines. All of these areas are of crucial importance to our society as a whole, in particular to the development of cleaner combustion, more efficient lubrication, higher throughput fluid transport, new nano materials and for a better understanding of fundamental biological transformations. Also, the proposed research will lead to the continued development of theoretical methods and computational codes for accurate, fast simulations in the areas described above, making use of both micro and macro parallelization. These codes will be freely distributed to the scientific community at large. In addition, future research scientists undergraduates, graduate students and postdoctoral fellows will be trained in the use of the latest tools in computational chemistry and biophysics and in the development of strategies for efficient use of high performance, massively parallel computer architectures. These skills are crucial to ensure the technological leadership of the United States. MRI: Development of A Versatile Service Oriented Wireless Mesh Network for Disaster Relief And Environmental Monitoring in Puerto Rico This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Proposal #: CNS 09 22996
PI(s): Lu, Kejie; Rodriguez, Domingo
Institution: University of Puerto Rico Mayaguez
Title: MRI/Dev.: Versatile Service Oriented Wireless Mesh Network for Disaster Relief & Environmental Monitoring

Project Proposed:
This project, developing a versatile service oriented wireless mesh network (WMN), VESO MESH, that can be quickly built to respond to natural disaster, establishes a data intensive environmental monitoring application specifically addressing hurricanes and earthquakes. The VESO MESH architecture aims to
Integrate distributed processes and storage devices into the network so as to provide effective data process and access inside the network,
Construct high throughput backbone in WMN emphasizing the transmission of a larger volume of data, and
Design and implement a set of service oriented protocols to efficiently provide services to users inside the network and effectively utilize the network resources in terms of energy, processing, storage, etc.

Broader Impacts: The work promises to enable high throughput data applications in disaster relief and environmental monitoring scenarios. It also plays a critical role in improving recruitment and retention of minorities at a large minority serving institution with a high focus in engineering and a large enrollment of women. Curriculum materials will be developed and disseminated. The project provides educational and training opportunities for students at all levels in an EPSCoR jurisdiction. MRI: Development of Software Defined Communications Testbed for Radio and Optical Wireless Networking Proposal #: CNS 09 23003
PI(s): Dandekar, Kapil R.
Fontecchio, Adam K.; Johnson, Jeremy R.; Kim, Youngmoo E.; Kurzweg, Timothy P.
Institution: Drexel University
Title: MRI/Dev.: Software Defined Communications Testbed for Radio and Optical Wireless Networking

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Project Proposed:
This project, developing a multi purpose Software Defined Communications (SDC) testbed to be used for rapid design and prototyping of the next generation wireless communication networks making use of radio, optical, or ultrasonic modalities, responds to the impending need for new high bandwidth, inexpensive, flexible, and upgradable wireless communication technologies to meet the growing demands of future applications. Among others, the SDC testbed aims to enable many projects including high speed secure data transmission, thru metal relay and control networks, localization and tracking, real time wireless video transmission, and enhanced home entertainment systems. The integrated plan challenges the existing radio frequency centric view of software defined radio by addressing the requirements of high bandwidth, robustness, and easy configurability for future telecom applications. Supporting multiple signal propagation media within a single framework, the project aims to
Develop a modular hardware and software platform which can be used to prototype technology and algorithms for macro and micro scale communications, local area networking, and localization applications.
Provide software reconfigurability at all layers of the protocol stack (in contrast to conventional SDR which only provides physical layer flexibility) so that cross layer UWB and optical communications techniques can be developed and field tested.
Disseminate transceiver hardware block diagrams and FPGA/DSP software modules to allow the SDR community to prototype high performance radio, optical, and ultrasonic communications systems.
Build propagation channel data repositories for high data rate UWB, optical, and ultrasonic wireless systems which can be used by industry and academia to design and evaluate new algorithms.
Demonstrate, experimentally for the first time, the augmentation of UltraWideBand (UWB) multicarrier, UWB impulse radio, diffuse free space optical, line of sight optical communications with spectrum efficient MIMO space time coding techniques.

Broader Impacts: This project offers educational opportunities for graduate and undergraduate students, as well as for K 12 students and teachers. It catalyzes advanced telecommunications development, but also trains and excites future engineers and general public. The project creates an Entrepreneur Development program in ECE; it will be used as a model for undergraduate senior design students in the context of starting a small business; facilitates independent study and cooperative work, and conducts outreach programs with local students and teachers. MRI: Acquisition of Two REMUS Autonomous Underwater Vehicles for Multiple Cooperative Marine Vehicle Research Proposal #: CNS 09 23031
PI(s): An, Edgar Beaujean, Pierre Philippe J.; Xiros, Nikolaos I.
Institution: Florida Atlantic University
Title: MRI/Acq.: Two REMUS Autonomous Underwater Vehicles for Multiple Cooperative Marine Vehicle Research
Project Proposed:
This project, acquiring two REMUS 100 autonomous underwater vehicles (AUVs), addresses the research challenge of multiple cooperating vehicles systems (MCVS), which involves a fleet of AUVs and autonomous safety vehicles (ASVs). Since successful operation of any MCVD often rests upon efficient communication among the vehicles, the dynamic mission planning problem (including search and classify and mapping capabilities) is studied utilizing a fleet of heterogeneous AUVs and ASVs. This challenging problem, considering underwater acoustic communication, navigation, and sensing constraints, differs from those usually studied and reported in the underwater literature. The work is contrasted to unmanned aerial or ground vehicles in which communication and navigation capabilities are generally NOT considered significant bottlenecks. The project capitalizes on two previous NSF achievements:
An underwater vehicle network simulatorsand
A location aware source routing (LASR) protocol developed for multiple communicating AUVs subject
to realistic underwater communication constraints.
Nonlinear and stochastic variations with the environmental conditions and sensor characteristics are expected. Because a closed form solution for an optimal controller design is not anticipated, given that control performance tends to be cost prohibitive, when completely via experimentation. Hence, given the wide range of mission and environmental scenarios and controller objectives, the problem will be studied combining modeling, simulation, and experimentation. Some of the simulated results will be validated by carrying out targeted at sea field experiments using the instrumentation. Local cost definitions will be explored as a means to constrain individual vehicle?s maneuvering and cooperation. Attention is paid to maximizing the overall mission efficiency while minimizing the impact of uncertainty. Expectations include development of:
An advanced event base planning and control algorithms to improve robustness of communication and
navigation uncertainty, throughput, and bandwidth efficiency,
An advanced multiple cooperating vehicle modeling software to support mission performance analysis, and
A science base for multiple AUVs and undersea acoustic networks.

Broader Impacts:
The research problems constitute ideal topics for theses and dissertations. This multidisciplinary field will be integrated into the existing course curriculum, providing valuable theoretical and simulation knowledge to the students, as well as hands on experiences. Using underwater robotics as a primary domain application, the existing effort, sustaining the teachers and students with interest in math and science, will be continued giving special consideration to female and underrepresented groups. Workshops, competitions, and summer classes are also planned to expose students the logistics of marine vehicles. The understanding gain is likely to provide insights to the U.S. Navy for missions using a fleet of autonomous underwater vehicles to better search and classify in homeland security missions. MRI: Acquisition of a Parallel Computing Cluster and Storage for the Marquette University Grid (MUGrid) This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The Marquette University computational grid (MUGrid) brings together a diverse team of investigators from mathematics, statistics, computer science, chemistry, electrical and computer engineering, mechanical engineering, and biomedical engineering to develop and utilize grid computing as an integral tool for computational research.

This project expands the developing MUGrid with the acquisition of a new parallel computing cluster, storage and network infrastructure upgrades which will also provide the cyberinfrastructure for Marquette University s new graduate programs in Computational Sciences. The Computational Sciences program, part of the Department of Mathematics, Statistics and Computer Science, will serve as the central location for formal training on cyberinfrastructure technologies for research computing. In addition to formal training, an open access seminar and regularly offered MUGrid ?boot camps? will provide training for students and faculty across campus, including individuals from traditionally under represented groups in STEM disciplines. MUGrid s success will be evaluated through usage reports, longitudinal tracking of student career progress, and user surveys, which are aimed at monitoring job preparedness and impact on training a diverse workforce.

The new parallel computing cluster and storage infrastructure will also be made available to the broader community through outreach activities. The resource will be part of Southeast Wisconsin High Performance (SeWHiP, http://www.sewhip.org) computing consortium s developing regional grid that will directly support academic/industry partnerships for healthcare, energy and water research in Southeast Wisconsin, the most diverse region of the state. Summer programs for K 12 teachers and students will provide training and access to MUGrid and SeWHiP resources. Projects developed on MUGrid will be used to promote science and engineering through partnerships with local museums such as Discovery World. Resources will be also made available to the Open Science Grid and TeraGrid. Resource usage tracking, longitudinal studies of impacts on K 12 student career paths, exhibit attendance, and surveys will be used to assess the impact of these outreach programs. Finally, training resources and software developed as a part of this project will be made widely available through the SeWHiP web site. MRI: Acquisition of a Computational Science and Engineering Parallel Cluster Proposal #: CNS 09 23050
PI(s): Jones, Jim; Jacher, Steven M.; Wang, Ke Gang; Zhang, Ming
Institution: Florida Institute of Technology
Melbourne, FL 32901 6975
Title: MRI/Acq.: Acq. of a Computational Science and Engineering Parallel Cluster
Project Proposed:
This project, acquiring a new computational science and engineering cluster, seeks to provide the faculty and students regular access to a modern, on campus parallel computer for their research. Servicing Mechanical and Aerospace Engineering, Physics and Space Science, Marine and Environmental Engineering, Oceanography, and Mathematical Sciences, the project aims to continue conducting research involving parallel computing in a wide variety of disciplines. Due to its age, an existing Beowulf Cluster acquired in 2001 has proved insufficient to conduct the current research and more powerful computing resources are sought. The proposed instrument consists of a 48 node cluster where each node has 2 quad core 2 GHz CPUs. The computing power of this machine will enable the large scale simulations needed in the application areas of ocean modeling, space weather, and material science. In these areas researchers have simulation codes, but need increased computing power to do the required runs. The new machine is expected to allow greater fidelity simulation on important application, such as the impact of space radiations on astronauts and electronic components on satellites. Other applications need the increased computing power and are beginning to parallelize their codes.

Broader Impacts: This project contributes to increase on campus access for students to develop their parallel programming skills. Building on past experience, careers in high performance and parallel computing will be encouraged. The students will be trained more in depth in the area and Parallel Computing will continue to be part of research at the institution. MRI: Development of Versatile Mobile Range Scanning System?Enabling Large scale High density 3D Data Acquisition for Cross Disciplinary Research Proposal #: CNS 09 23131
PI(s): Yang, Ruigang;
Crothers, George M.; Dinger, James S.; Seales, William B.; Weisenfluh, Gerald A.
Institution: University of Kentucky
Lexington, KY 40506 0057
Title: MRI/Dev.: Versatile Mobile Range Scanning System Enabling Large scale High Density 3D Data Acquisition for Cross Disciplinary Research

Project Proposed:
This project, developing a versatile mobile range system based on the principle of Light Detection and Ranging (LIDAR), enables:
New data driven approaches in 3D reconstruction and modeling, pattern recognition, and data mining;
Quantitative studies in geology (e.g., analysis of hill slope movement and characterization of outcrops;
Archaeological recording of fragile sites with limited access (e.g., those in caves).
The system consists of 3 LIDAR sensor heads, GPS and inertial measurement unit, high resolution digital cameras, and processing software to generate high resolution, accuracy colored point clouds that are geo referenced. Unlike typical laser scanning system, it supports both stationary and high speed mobile scanning. For example, when mounted on a vehicle, it should be able to generate over 50 points per meter given a speed of 100km/hr. In comparison, the typical density from an airborne LIDAR system is only 1 2 points per meter. This level of density at high scanning speeds allows rapid 3D scanning of large scenes not possible with either stationary systems (too slow) or airborne systems (too sparse). Through relatively simple hardware modification and more sophisticated processing algorithms, the state of the art in scanning range, density, and operational conditions is expected to advance as follows:
Increasing the scanning range over 6 times in the mobile mode, from the original 100 m (specified by the LIDAR sensor vendor) to greater than 600 m.
Improving the point density over 10 times by applying the principle of compressing sensing and information from the high resolution color image, without changing the hardware configuration.
Developing the ability to scan beyond line of sight with inexpensive self calibrating reflective mirrors, and develop real time vision and inertial based navigation to allow GPS free operations for mobile scan (critical for scanning in underground areas, where many archaeological sites reside).
Plans also include building a large scale 3D cityscape database with over 30 billions of raw data samples to cover the entire Lexington metro area. The database will be shared to enable research beyond graphics and vision, including, but not limited to, data compression, transmission, visualization, index and retrieval, and computational geometry.

Broader Impacts: The instrument is expected to be the first university owned mobile laser range sensing system, which might inspire and facilitate many exciting new research venues. The scanning and processing tasks will be carried out and documented by undergraduate students. The geological and archaeological studies will be used to attract local students, in particular those underrepresented from Appalachia. The system will have commercial values to provide on demand scanning for applications ranging from construction, city planning, law enforcement, survey and mapping, to 3D imaging. Commercial users will be charged a fee to maintain operation and sustain the enabling data beyond the life of the award. MRI Acquisition of Equipment for the Establishment of a Reconfigurable Computing Center at the University of Puerto Rico Proposal #: CNS 09 23152
PI(s): Arce Nazario, Rafael; Orozco, Edusmildo
Institution: University of Puerto Rico ? Rio Piedras

Title: MRI/Acq.: Acq.of Equipment for the Establishment of a Reconfigurable Computing Center at UPR RP

This project, acquiring a shared computational resource instrument with multiple high end general purpose processors and Field Programmable Gate Arrays (FPGA) co processors that form flagship platforms for Reconfigurable Computing (RC), supports interdisciplinary collaboration with research groups in Finite Field Applications, Computational Discrete Mathematics, Electronics Design Automation, and Developmental Biology. Focusing on research tools for assisting automatic partitioning of applications to multicore architectures and configurable platforms, the project services eight application fields, including hardware/software (HW/SW) co design tools, power optimization tools for embedded systems, Latin square orthogonality and hypercube cut number related problems, reconfigurable radar waveforms, and bioinformatics. The DMAGIC (DST Mapping using Algorithmic and Graph Interaction and Computation) methodology generates a high level efficient partition for a given pair of problem descriptions in Discrete Signal Transform (DST) and architectural descriptions. Hence, algorithms will be designed mapping the different research areas into the platform. Education and research training are given a high priority.

Broader Impacts: The instrument expedites scientific results, impacts the education and training of all users, and contributes to workforce training. Collaborations open new opportunities for undergraduate research, contributing not only to retain students, but also to direct them to graduate studies. This university, residing in an EPSCoR jurisdiction, services a large number of underrepresented minority students. MRI: Development of a Gesture Based Virtual Reality System for Research in Virtual Worlds Proposal #: CNS 09 23158
PI(s): Ilies, Horea T.;
Anderson, Amy; Kazerounian, Kazem; Marsh, Kerry L.; Nowak, Kristine
Institution: University of Connecticut
Title: MRI/Dev.: Dev. of a Gesture Based Virtual Reality System for Research in Virtual Worlds
Project Proposed:
This project, developing an integrated virtual environment system capable of allowing not only 3D visualization of data, but also interaction with data through natural hand and finger gestured based on a dual interface, exploits a multi touch interaction interface and a vision based hand gesture interface. Virtual reality environments rely on a collection of technologies that allow the user to go through a coherent and unified perceptual experience involving multiple senses, such as vision, touch, and sound, while interacting with 3 dimensional data. These immersive, highly visual, 3D environments currently offer a fairly high level of performance for spatially visualizing data. However, the corresponding machinery providing user interaction with these systems has not kept up the same pace of development with the visualization tools. At present some advanced commercial environments offer some user interaction capabilities achieved through wired wearable hardware (such as wired gloves and head mounted displays). This promotes, in turn, an unnatural and cumbersome interaction between the user and the virtual reality environments, curbing the acceptance of the technologies. The syntax and semantics of the hand and finger gestures developed interacts with geometric data, while the implementation relies on the Multi Touch Surface computing platform as well as on a newly developed gesture tracking and recognition system. The environment should lead to a potent open platform for interacting with virtual geometric data in an intuitive way, without the need for wearable hardware galvanizing the state of the art at the institution in Nano and Design Engineering, Psychology, Computer Science, Structural biology, as well as support research in the Center for Health, Intervention, and Prevention (CHIP). This project addresses the problem.
Broader Impacts:
This instrumentation will be open source and widely available with well documented set up procedures. The VR system will be networked with other VR sites, including VRAC at Iowa State, to maximize the impact and stimulate technology transfer. Moreover, the instrument contributes to
Stimulate critical avenues of interdisciplinary research involving engineering, biology, computer science,
psychology, and Human Computer Interaction (HCI),
Strengthen the potential for educational, student recruiting, and outreach activities, and
Perform targeted outreach to K 12 students, teachers, and school districts serving groups that have
traditionally been underrepresented in the engineering disciplines.
The project also advances the state of the art in the teaching and practice of engineering design, as well as other fields in which geometry plays and important role. Contributing to the development of a new generation of professionals that use capabilities of virtual reality tools to augment traditional disciplines for improved engineering design, this work should have a long lasting impact on the ability of scientists and engineers. MRI: Acquisition of Mobile, Distributed Instrumentation for Response Research (RESPOND R) Proposal #: CNS 09 23203
PI(s): Murphy, Robin R.; Ames, Aaron D.; Gutierrez Osuna, Ricardo; Song, Dezhen; Stoleru, Radu
Institution: Texas A & M
College Station, TX 77843 3000
Title: MRI/Acq.: Mobile, Distributed Instrumentation for Response Research (RESPOND R)

Project Proposed:
This project, acquiring equipment to create a mobile, modular, distributed instrument called RESPOND R, aims to develop an integrated, interoperable research instrument suitable for a comprehensive use of Cyber Physical Systems (CPS) technologies for various types of incidents: urban building and bridge collapse (earthquakes, hurricanes, explosives), radiological or chemical spills, and terrorism. CPS explores new technologies such as robots, wireless networks, miniature sensors, sensor networks, and other types for emergency preparedness, prevention, response, and recovery. The instrument services projects in multidisciplinary research in Emergency Informatics for disaster response. Emergency Informatics consists of a real time collection, processing, distribution, and visualization of information for emergency preservation, preparedness, response, and recovery. RESPOND R addresses the lack of an Integrated Emergency Informatics Instrument that can be configured to support scenario based experimentation and transported to exercises and responses. To capture the interaction of devices, the physical worlds, and human decision making, the project will use the Disaster City facility at the institution. CPS components in core areas (unmanned systems, imaging/chemical and radiological sensors, wireless communication networks, nodes and motes, and human performance) will be acquired. Addressing testing within real world conditions and scenarios, RESPOND R is expected to advance knowledge in the emerging, multidisciplinary fields of Emergency Informatics and CPS by providing access to a complete, large scale system that is located in a fidelity disaster testbed or can be transported to homeland security exercises or actual incidents. RESPOND R will be open to researchers outside the institution, with a projection that 22 universities will participate over 3 years.

Broader Impacts: Used by 18 faculty in 8 departments and 2 colleges (Engineering and Architecture), the instrument contributes to educate 10 20 graduate students each year through field components of courses and dissertation work. Leveraging initiatives for underrepresented group, RESPOND R will also be utilized for the training of 1000 responders and policy makers. In keeping with instruments used by ?storm chasers,? the instrument will be deployable for exercises and disasters where CPS components can reduce negative outcomes of events while gathering extremely valuable data and/or save lives. MRI: Acquisition of an Optical Motion Capture System for Human Centered Computing Research Proposal #: CNS 09 23238
PI(s): Sheng, Weihua; Cheng, Qi; Li, Xiaolin; Rahnavard, Nazanin
Institution: Oklahoma State University
Stillwater, OK 74078 1011
Title: MRI/Acq.: Acq.of an Optical Motion Capture System for Human Centered Computing Research
Project Proposed:
This project, acquiring an optical motion capture system, supports research for infrastructure free human context awareness and context based human intention recognition. The system relies on wearable sensing and computing devices to achieve human context awareness. The research aims to answer central questions in understanding the interplay between human and computing, mainly
Understanding human context (e.g., behavior, location) through embedded computing, and
Exploring the knowledge of human context to improve embedded computing applications.
The first research project investigates the development of infrastructure ?free human context awareness in GPS restricted environments. The second studies the use of the knowledge of human context to better understand human intentions and addresses the study in a human robot interaction (HRI) setup. Both projects require a motion capture system that can provide location ground truth, allow performance comparison, and facilitate system calibration. The work aims to
Develop a theoretical framework to achieve a stand alone human context awareness that requires zero infrastructure setup. (Some activities, such as virtual landmarks are expected to contribute in indoor human localization.)
Develop explicit human intention recognition based in the context of both location and activity that pushes forward research in human robot interaction
Lead to a better understanding of interactive, coupled relationship between human and computing.

Broader Impacts: The work impacts wearable computing, human computer interaction (HCI), and ubiquitous computing research. Modifications of the developed hardware and software might serve to track the locations and monitor the status (activity and health) of first responders, improving the efficiency of personnel safety in their operations. The system improves the research capabilities in Oklahoma, an EPSCoR state. Planned education and training activities contribute in preparing students in embedded computing, wireless communication, signal processing, and human behaviors. Outreach activities stimulate prospective students. Efforts will be made to involve Native American and female students. MRI: Acquisition of a High Performance Instrument for Interdisciplinary Computational Science and Engineering Proposal #: MRI 09 23256
PI(s): El Ghazawi, Tarek A.; Briscoe, William J.; Lang, Roger H.; Lee, Frank X.; Mittal, Rajat
Institution: George Washington University
Title: MRI/Acq.: High Performance Instrument for Interdisciplinary Computational Science and Engineering
Project Proposed:
This project, acquiring a Cray XT5 and adequate storage, upgrading the network of a computer cluster, and integrating these as one High Performance Computing (HPC) instrument, supports a team of interdisciplinary researchers from two schools and seven departments via the new Institute for Massively Parallel Applications and Computing Technologies (IMPACT). New programming models and processor technologies in understanding high performance computing systems productivity, as well as the impact of new technologies on science will be investigated. Programming models include UPC, Co Array FORTRAN, X10, and Parallel Matlab. Application requirements will drive the UPC 10 work and application specific PGAS compiler optimizations. Investigating new processor technologies will be pursued jointly with the NSF Industry/University Cooperative CHREC center. In computational fluid dynamics (CFD), the instrument will help in the design of micro aerial vehicles inspired by the dynamics of insect flight (previous work assumed rigid wings). The HPC instrument will help carry out detailed modeling to understand insect wing?s deformation. The instrument will also be used in employing CFD techniques to analyze swimming strokes for U.S. national teams using a novel immersed boundary method. Current simulations are carried out at low Reynolds numbers due to lack of HPC resources. In nuclear physics, the instrument will help unravel the structure of matter at its deepest level as governed by Quantum Chromodynamics (QCD). The latter is very difficult to solve without massively parallel computations on a discrete space time lattice with millions of degrees of freedom. The instrument will help expedite the linking of experimental and theoretical nuclear physics studies in fundamental nuclear reactions and account for the most recent experimental results in photo and electro production of mesons and hyperons. Helping the remote sensing group, it will also contribute to provide more accurate models of the soil, ground surface and vegetation that can relate the sensor responses to the bio physical variables on the ground.

Broader Impacts: This acquisition enables producing a new generation of students and postdoctoral fellows who can face the current changes in computing technology. The instrument will allow assigning
realistic computational problems that can integrate research and teaching promoting learning through discovery. Initiatives include an extensive outreach program. Open house events are planned and will include community colleges, high school and middle school students, and K 12 teachers to increase interest in science and technology. A concrete plan is devised to include women and minority to receive training, use the instrument, and engage in research collaboration with the faculty. The plan focuses on HBCU institutions in the area, and is open to women and minorities. Investigations will result in codes and data that will be shared with the community using a public license such as GNU/GPL. The website will provide access to papers, presentations, and freely distributed relevant software; conferences and workshops will also be organized. Furthermore, a graduate certificate in interdisciplinary HPC will be issued and more faculty in the HPC area will be hired. MRI/Acq: Proposal for support for the Consortium for Research Computing for the Sciences, Engineering and Technology CRCSET Proposal #: MRI 09 23282
PI(s): Flurchick, Kenneth
Li, Yaohang; Mohan, Ram; Tang, Guoqing
Institution: North Carolina Agricultural & Technical State University
Title: MRI/Acq.: Support for Consortium for Research Computing for Sciences, Engineering, and Technology CRCSET

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Project Proposed:
This project, acquiring, deploying, and maintaining a distributed High Performance Computing (HPC) system, aims to provide appropriate processing speed and power to pursue the needs required by various departments, programs, and research groups that have joined the institutional Consortium for Research Computing for the Sciences, Engineering, and Technology (CRCSET). These groups try to advance their respective research via the computational study of physical chemical properties and simulation of chemical/physical processes in molecules, nano materials and broad based bulk materials, physics based modeling in engineering disciplines, and other areas such as computational biology, finance, business, etc. Combined with advanced research training in computational science and engineering, the acquisition supports core projects and ancillary research projects, including:
Computational Science and Engineering Program (core program recently approved for a PhD)
All Electron Studies of Organic Molecular Crystals (OMCs) under Pressure
Polymer Composite Fabrication Process Modeling and Simulations
Ancillary Research Projects and Tools:
.Modeling C60 Reorientation of Various Solvents
.Modeling Material Deformation at Nano Length Scales
.Computational Modeling and Simulation of Bio Inspired Adaptive and Reconfigurable Systems
.New Research Tools for Collaborative Grid Computing and Visualization
Research and Education Support
Broader Impacts:
This work fosters research and research and education training for underrepresented groups at a historically black college and/or university (HBCU) and additionally provides computational resources to other HBCUs. Moreover, it aims to gain a better description at the atomic scale of the OMCs and to improve the understanding of the manufacturing processes for advanced composites. MRI: Acquisition of an RFID Testbed Using Renewable Energy for Object Identification and Habitat Monitoring Proposal #: CNS 09 23313
PI(s): Fu, Kevin;
Burleson, Wayne P.; Diao, Yanlei; Ganesan, Deepak K.; Ross, Charles
Institution: Five Colleges: Amherst, Hampshire, Mount Holyoke, Smith, U Mass
Amhers, MA 01002 2324
Title: MRI/Acq.: RFID Testbed Using Renewable Energy for Object
Identification and Habitat Monitoring

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Project Proposed:
This project, acquiring a Radio Frequency IDentification (RFID) testbed, aims to enable fundamental cross disciplinary cross layered research into design, implementation, and deployment of the next generation RFID technologies and applications. The equipment support research in the following topics:
Design of novel next generation Computational RFID (CRFID hardware that can harvest energy from alternative sources (WiFi and solar), and can operate with extremely limited energy,
Design of a CRFID operating system that can tolerate frequesnt interruptions pf power through lightweight checkpointing and predictive management,
Development of data cleaning and probabilistic interference techniques for processing streams of noisy and incomplete RFID data,
Design of accurate locationing techniques using re targetable mobile RFID readers on robots,
Deployment of an RFID based biological monitoring application involving crickets with small implanted tags that brings together the above research thrusts, and
Enhanced RFID security mechanisms and analysis of security vulnerabilities.
The effort integrates research and education in RFID systems, computer systems, data management, security, energy aware computing, and biology to advance knowledge and understanding of zero power pervasive devices. The project addresses the critical challenge of RFID reader placement by designing computational RFIDs that can harvest from alternative sources other readers, designs an operating system tailored to the needs of such computational RFID devices, and designs services to localize RFIDs and manage noisy and missing data. Moreover the security characteristics of RFIDs at different layers are analyzed and cryptographic solutions for highly constrained devices are developed Expected results include hardware designs, operating systems, algorithms, protocols, and models for designing such systems. The experimental and prototyping efforts should provide new insights into system design issues. Design of novel next generation Computational RFID (CRFID

Broader Impacts: The research directly impacts the design, development, and deployment of the next generation RFID applications and businesses including medical, environmental, energy, military, homeland security, and transportation areas. The equipment has direct impact on undergraduate research and student diversity. The RFID testbed will be integrated within the curriculum across Computer Science, Evolutionary Biology, and Electrical Engineering leading to a better understanding of the technology and its impact on the environment and energy. MRI: Acquisition of the ISTeC High Performance Computing Infrastructure for Science and Engineering Research Projects Proposal #: CNS 09 23386
PI(s): Siegel, Howard Jay
Burns, Patrick J.
Institution: Colorado State University
Title: MRI/Acq.: ISTeC High Performance Computing Infrastructure for Science and Eng Research

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Project Proposed:
This project, acquiring High Performance Computing (HPC) instrumentation, aims to support much larger and more complex problems in science and engineering (especially for data intensive applications), add greater physical fidelity to existing models, facilitate application of HPC to new areas of research and discovery, and support training to attract new researchers. Filling the current void in cyberinfrastructure (CI) resources, the project is expected to change the culture of discovery at the institution, elevating HPC to the level of central infrastructure. Advanced networking will facilitate the interaction of Tier 3 system with other systems, and enhance researchers? abilities to utilize Tiers 1 and 2 systems worldwide. The principal focus of the system will be data (and often compute ) intensive applications in NSF funded research areas, such as design of extreme ultraviolet lasers, weather forecasting, computational physics, climate change, atmospheric modeling, bioinformatics, network traffic analysis, robotics, computational electromagnetics, remote sensing, robust resource allocation, and magnetic materials. Tier 3 system will expand the domain and range of computations, allowing problems of greater fidelity including multi physics algorithms and much finer spatial and temporal resolutions. The work is carried out with CSU Pueblo, a Hispanic serving institution, as full partner.
Broader Impacts:
The existing Tier 3 will share cycles of different architectures for educational purposes: workshops, seminars, expertise via affinity groups, instructors, courses, and students. The institution, a large producer of STEM graduates, will integrate HPC into the STEM curricula and K 12 teacher training programs with emphasis on inclusion of women and minorities. Member companies of the ISTeC Industrial Advisory Council have expressed interest in participating in the activities. Existing courses support the activity through teams consisting of application experts from diverse disciplines (breadth), and mathematicians (algorithms) and computists (mapping applications and algorithms to parallel architectures) (depth). MRI: Development of a Next Generation Interactive Virtual Reality Display Environment for Science Proposal #: CNS 09 23393
PI(s): Laidlaw, David H.; Hesthaven, Jan S.; Karnadiakis, George E.; van Dam, Andries
Institution: Brown University
Providence, RI 02912 9002
Title: MRI/Dev.: Next Generation Interactive Virtual Reality Display Environment for Science

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).



Project Proposed:
This project, developing a world class interactive large field of view 95 megapixel immersive virtual reality environment, aims at creating a novel, demonstrably useful, rich, and expressive interaction, visualization, and analysis that truly leverage the human visual and motor systems in Virtual Reality (VR).
This work intends to help accelerate scientific work, research into innovative visualization methods for accelerating science in the future, and even leverage the fundamental advantages of immersive large field of view visualization and body centric human computer interaction. Two decades of research have established the value of immersive displays as a research tool in many scientific domains and has also identified a set of currently unmet needs that block application of such displays to new problems and domains. These needs encompass high display resolution, brightness, contrast, and size; fast, responsive tracking with high accuracy and low latency; ease of use in working with new kinds of data; and reliability. Although a few multi million dollar systems exist that may be able to address these needs, these few systems do not match the proposed display?s color gamut, small physical space requirement, and lower replication cost. The system is expected to support more natural and effective interaction with data than the current 3D point and click wand driven CAVEsTM by maximally utilizing as appropriate full body, motion captured user interactions and gestures. More display information will be made accessible to the human visual system with less user effort by matching, or exceeding the perceptual qualities of a modern LCD monitor. An immersive stereo display will be integrated with the perceptual resolution of a desktop display and superior brightness and contrast. Integration of software tools for creating virtual reality applications quickly will address ease of use and reliability. The new tools are expected to be simple, support a spectrum of displays, and provide rich support for gestural interaction. A monitoring process to identify potential problems among the interacting hardware and software components will be put in place to identify and address problems before instruments are delayed. Users of the system include planetary geologists, systems biologists, brain scientists, cell and molecular biologists, biologists studying animal motion (including flight), fluid dynamicists, bioengineers studying arterial hemodynamics, visual designers developing interactive techniques for scientists, digital literary artists, and visualization and interaction researchers. Within interaction research, experiments using the system are expected to establish the appropriate level of display technology (e.g., resolution, interactivity, or stereographic display) needed for different classes of scientific analysis. The techniques, monitoring system, and software environment will be distributed on SourceForge to respectively help accelerate scientific progress nationwide, for developing multi display applications, and for ensuring reliability.

Broader Impacts: While educating many students, the instrument is expected to enable new advances in all of the scientific disciplines of the users listed above, including a better understanding of the workings of cells and genes and proteins they contain (which could consequently improve quality of life broadly), behavior of fluids in arteries and around moving animals, animal locomotion (which could lead to improved biomimetic locomotive, floating, or flying vehicles), the wiring of the human brain, how it affects human capabilities, and how it can degrade; and Mars. The efforts are likely to produce a new generation of scientists who can better analyze research problems using scientific visualization, computer scientists more cognizant of scientists? analytical needs, and artists and designers who can accelerate the design process for immersive scientific visualization tools. MRI: Acquisition of the Cyber ShARE Collaborative Visualization System This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Proposal #: CNS 09 23442
PI(s): Romero, Rodrigo; Gonzalez, Virgilio; Hurtado, Jose M.; Konter, Jasper; Smith Konter, Bridget R.
Institution: University of Texas ? El Paso
El Paso, TX 79968 0587
Title: MRI/Acq.: Acq. of the Cyber ShARE Collaborative Visualization System

Project Proposed:
This project, acquiring a collaborative visualization system referred to as the Cyber ShARE Collaborative Visualization System (C2ViS) to present high resolution displays of scientific datasets for exploratory monitoring, supports interdisciplinary research and enables development of new visualization techniques for analyzing and integrating large datasets from the geosciences, environmental sciences, computational sciences, material sciences, biomedical engineering, and other domains. The tile display 32 monitor system will be driven by 32 workstations interconnected by a high speed network controlled by a single computer acting as a head node. With an overall 131 megapixels display resolution, images using the full resolution of the tiled system are physically sent to each machine via open source software. Supporting a variety of exploratory applications, C2ViS also supports both local and virtual organizations and serves as a resource for city and county government entities, as well as regional school districts.
Applications include the visualization of tomography models of the subsurface of the Earth, spatial surface datasets from geosciences and environmental science communities, and 3 D solid models for advanced material fabrication, tissue engineering, and biomodeling. In addition, the system will extend Cyber ShARE?s ability to support interdisciplinary projects, including those related to knowledge representation, trust management, and visualization of scientific data and results, CI based geological model and data fusion for integration of whole Earth models, and characterization of ecological and environmental phenomena through sensor network and data stream organization.

Broader Impacts: This project contributes to engage young faculty and attract quality researchers to this minority serving institution. A high resolution system allows researchers to view data from different perspectives and abstraction levels, enhancing analysis, exploration, and discovery. Furthermore, the system will be used for outreach and education programs to attract students to STEM fields. Middle high schools brought in for summer workshops students will be excited and motivated by interacting with complex graphical models. MRI: Development of an Anechoic Chamber and Instrumentation for Remote Sensing of Polar Regions and Transportation Interdisciplinary Research and Education Proposal #: CNS 09 23443
PI(s): Leuschen, Carlton J.; Braaten, David; Gogineni, S. Prasad; Seguin, Sarah; van der Veen, Cornelis
Institution: University of Kansas
Lawrence, KS 66045 7563
Title: MRI/Dev.: Anechoic Chamber & Instr. for Remote Sensing of Polar Regions & Transportation Interdisciplinary R&E
Project Proposed:
This project, developing an anechoic chamber at the Center for Remote Sensing of Ice Sheets (CReSIS) to measure ElectroMagnetic Interference and ElectroMagnetic Compatibility (EMI/EMC) and antenna characteristics, aims to improve sensor performance to meet environmental issues. The chamber offers an electromagnetically quiet room for detecting extremely weak emissions and accurate antenna pattern and impedance measurements. The facility covers a wide frequency range and attenuation of incident plane waves. Coupling effects on array scale models are necessary to be able to optimize critical signal processing algorithms. Optimizing antenna array performance requires radiation patterns and mutual coupling between elements. The chamber enables antenna pattern measurements using both far field and spherical near field scanning techniques over the entire frequency range. Thus, the work involves designing and developing a radar and sensor technology for observation of Greenland and Antarctic ice sheets (systems sensitive to sound) and image the coastal regions where rapid melting due to climate change is most evident. This ultra sensitive sounding and imaging radars enable radio frequency (RF) and microwave emission measurements 40 dB below the thermal noise and operates over a large frequency range from 100MHz up to 18 GHz, with attenuation of plane waves by about 100dB at the proposed chamber boundaries.

Broader Impacts: Beyond the potential contribution to ice sheet modeling, glaciology, and the overall international polar research community, the facility is likely to have significant impact on education. It enables new coursework in emerging and increasingly critical EMC/EMI field incorporating laboratory exercises into courses. Hence, the project provides educational and training opportunities for students at all levels in an EPSCoR state, provides business opportunities for businesses in the area, and clearly addresses international needs. MRI: Acquisition of Futuro: A Data Intensive and High Performance Computing Cluster for Integrated Research and Education Proposal #: MRI 09 23456
PI(s): Lei, Hansheng; Benacquista, Matthew J.; Figueroa, Andres; Iglesias, Juan R.; Mukherjee, Soma
Institution: University of Texas Brownsville
Title: MRI/Acq.: FUTURO: A Data Intensive and High Performance Computing Cluster for Integrated R & E

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Project Proposed:
This project, acquiring Futuro, a computer cluster for interdisciplinary research projects and Computer Science education programs, enables research activities in data mining, pattern discovery, genetic data analysis, experimental astronomical physical, collaborative filtering, theory of computation, high dimensional visualization, and other computational areas. Seeking to lay the foundation for a strong research and education integration centered on CS in two minority serving universities, UT Brownsville and UT Pan American, it enables the following potential transformative goals:
Terabyte scale data mining and pattern discovery in time series datasets obtained from heterogeneous sensor networks (addresses data analysis problems in Laser Interferometer Gravitational Wave Observatory (LIGO))
Genetic data analysis in complex human diseases to identify susceptibility factors enabling understanding of genetic causes of complex diseases such as schizophrenia (potential to lead to new therapeutic strategies)
Studying the dynamical systems and Stellar populations to model the behavior of black hole binaries in globular cluster and galactic nuclei (creates models of formation of stellar systems via intensive computation that can provide information for interpretation of results from operating gravitational wave detectors)
Exploring and creating computing effective, scalable, robust and intelligent learning algorithms for large recommender systems based on collaborative filtering by incorporating multispectral information (may lead to next generation recommender systems)
Visualizing high dimensional streaming data from heterogeneous sensors (potential to contribute in developing new data reduction methodologies that incorporate intelligent computation such as data mining and thus more advanced visualization systems with cross disciplinary utility)
Benchmarking and developing algorithms for approximating NP hard subgraph isomorphism problem with best possible practical performance (benefits applications such as image recognition and bioinformatics)
Futuro will serve as laboratory in which core research can be conducted in a collaborative fashion at a high level providing real world test applications while training students.

Broader Impacts: This project benefits many users from physics, bioinformatics, computational engineering, and environmental engineering in two minority serving universities. Futuro forms the nucleus for collaboration between computer scientists and researchers from other departments. The project will train students at the Rio Grande Valley, a historically underrepresented region with more than 90% Hispanics, in areas that are expected to have great national impact. The work provides experience in parallel programming and scientific computing, the CS curriculum will be enriched by the lab modules enabled by the cluster facility. MRI: Acquisition of an Integrated System for Advanced Visualization with Haptic Feedback Control A large scale visualization facility with integrated haptic devices is a critical need that will enable the realization of breakthrough science and engineering at the University of Texas at San Antonio (UTSA). Our proposed facility (Haptic Vis Wall) is uniquely configured to provide real time interaction with very large, complex data sets. Three primary research projects and ten other projects have been identified that require a large scale visualization facility so that researchers can develop in depth interpretation of their models and data and produce intellectual insight. In addition, due to the uniqueness of the equipment, sharing of the resource across UTSA, and the highly visible nature of the resource, it is foreseeable that collaborations across engineering, computer science, statistics and others will result. Also, the capability to include haptic devices into research and teaching will spawn new awareness and applications of this important resource into fundamental graduate and undergraduate courses and provide an engaging vehicle for addressing student retention and graduation rate issues.

Today s students often exhibit superior learning through graphically based instruction. This is especially true for underrepresented minority students that are typically the first generation to attend college and often lack access to the latest technology. As a result, this facility, which will be open to all UTSA faculty and students, can have a significant impact on educational and research accomplishments. Further benefits include a dramatic increase in publicity of computational science, engineering, computer programming, haptics and controls, and computer graphics to administrators, faculty, students and prospective students. The facility will be centrally located on campus for both teaching and research. MRI: Development of Reconfigurable DWDM Multi Mode Switching Platform for Supporting Telesurgery Telemedicine and Heterogeneous Applications Proposal #: MRI 09 23481
PI(s): Chen, Yuhua
Institution: University of Houston
Houston, TX 77204 2015
Title: MRI/Dev.: Reconfigurable DWDM Multi Mode Switching Platform for Supporting Telesurgery Telemedicine and Heterogeneous Applications

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Project Proposed:
This project, developing a dense wavelength division multiplexing (DWDM) multi mode switching platform instrument, supports telesurgery, telemedicine, and heterogeneous applications by allowing different types of messages within an application to choose from three switching modes: (electronic packet switching (EPS), optical burst switching (OBS), and optical circuit switching (OCS). While using the EPS mode to transfer short robotic control messages, applications such as telesurgery are expected to benefit from the approach by using the OBS mode or the OCS mode to transfer the ultra high bandwidth multi view 3 D high definition (HD) video. The multi mode switching platform will be developed using high performance reconfigurable field programmable gate array (FPGA) system along with optical switching nodes. The platform developed can be made available remotely to networking researchers and to the GENI infrastructure. The DWDM multi mode switching platform enabled by the router architecture allows the three switching paradigms mentioned above (i.e., EPS, OBS, and OCS) to be supported on the same router platform, providing the greatest flexibility to applications. Each DWDM channel in an optical fiber can be individually reconfigured to a different switching mode based on the dynamic traffic load. Additionally, the approach allows individual application or individual type of messages within an application to be transported using the best suited switching mode to achieve the best performance.

Broader Impacts:
This work can change the way quality medical services that can be delivered. Telesurgery and telemedicine allow medical experts and resources to be shared remotely by residents who do not have local access to such resources. While fostering research opportunities and collaborations in router designs, network management, and application designs, the platform instrument under development can provide research infrastructure to networking researchers and potentially unify the two often divided research communities: the electronic packet switching and the optical switching. The female investigator serves as a mentor in WELCOME (Women for Engineering Learning Community for Maximizing Excellence), actively recruiting and retaining women in engineering. She is hosting two minority Harmony Science Academy students. MRI: Development of a Next Generation Multimodal Data Management Human Sensing Instrument for Trustworthy Research Collaboration and Quality of Life Improvement Proposal #: CNS 09 23494
PI(s): Makedon, Fillia S.; Athitsos, Vassilis; Huang, Heng; Le, Zhengyi; Popa, Dan O.
Institution: University of Texas Arlington
Title: MRI/Dev.: Next Generation Multimodal Data Management Human Sensing Instrument for
Trustworthy Research Collaboration and Quality of Life Improvement
Project Proposed:
This project, developing an instrument that serves as an interactive personal care and human activity monitoring center, aims to keep a person with high quality life and safe at home as long as possible. The instrument enables privacy preserving and secure data sharing through wireless connection with remote users in an assistive living environment. Providing mental and emotional support, the zooscopion (zScope) can connect devices, humans, objects, and the environment. It can connect to other assistive living projects, making them interoperable and can deliver a Digital Library of sanitized research data and cases with high educational and training value. zScope combines and correlates many types of data and extracts events of interest that indicate changes, risks, etc. It can analyze facial expressions to detect pain, environmental data, house data (such as door opening, telephone sounds, vacuum cleaner, etc.), human performance metrics (e.g., hand strength), both in continuous and discrete format. Data are modeled and assembled in meaningful ways to predict and prevent physical and digital problems (e.g., respectively, falls and intrusions). Privacy and security are being made part of the data modeling at the design phase. The instrument will take sensor data, human body measurements, camera data when requested, known pattern of behavior from other cases, brain scans, and clinical information, aiming to provide high resolution displays of longitudinal as well as episodic events. It outputs a visual interactive display of patterns and significant human behavioral events valuable in assistive environments, setting where to use non invasive monitoring technologies, helping recognize behavioral biomarkers that will be connected to other types of health indicators that may come from brain imaging, genetic analysis, clinical results, or psychological evaluations. It will work with the next generation of data that include behavioral, clinical, body motion, etc., and have low latency tracking.
Broader Impacts:
This work enables human centric type of experiments and provides novel new ways of interaction, visualization, and secure collaboration. Developing smarter living environments for the aged opens new ways to education with immersive compelling projects that provide a better understanding of the role of science and engineering when combining health data (genomic information) to behavior, predict trends, and provide indicators of how medication and clinical assessments connect to longitudinal behavior. Long term goals include behavioral markers for assessing the confluence of environment, drugs, and human psychology. The instrument is also expected to respond to queries regarding emerging needs for new analysis of collected information. It includes training and educational modules with search and browsing tools and a recommender facility to support decision making and use stored strategies. Moreover, utilizing existing outreach programs, the project will support local minority students and high school students A new generation of scientists that can work together across domain silos towards human centered goals might be in the making! MRI: Acquisition of Tesla Hardware for Speech Recognition Proposal #: CNS 09 23511
PI(s): Janin, Adam
Institution: International Computer Science Institute
Title: MRI/Acq.: Acquisition of Tesla Hardware for Speech Recognition
Project Proposed:
This project, acquiring an nVidia Tesla architecture system, facilitates development parallel code algorithms for automatic speech recognition (ASR). The Tesla Hardware for Speech Recognition consists of a cluster of 10 nVidia Tesla rack mounted units with associated infrastructure. New parallel codes must be developed to improve the accuracy of speech recognition, since computer systems now obey ?Core?s Law? (where the number of cores on a chip doubles once every two years). The system provides a large scale multi core general purpose computing environment that enables the development of scalable parallel code algorithms. The instrument provides a testbed in which to investigate future scalable algorithms that are expected to facilitate continued leadership in the commercial and industrial markets in terms of advanced and novel algorithms. Thus, envisioning the state of computing in 5 to 10 years when multi core architectures prevail, the project aims to first
Secure appropriate computing resources for research in speech recognition and then
Eclipse the current 2 and 4 core basic desktop systems.
The institute performs extensive training for machine learning and speech recognition in realistic settings with challenging acoustic properties and natural, human to human communication. Applications run from hands free access to disabled users and natural speech driven interfaces for the non computer literate to automatic meeting assistants and browsers, in which meetings are recorded in real time and tools are provided that allow access to content both during and after the meeting. The following two relevant methods improve accuracy:
Multi stream methods that involve combinations at many levels within the system, including multiple features, multiple machine learning estimators, and multiple word streams combinations and
Increasing the size of the training set.
Careful integration of data can improve the accuracy even when the training data does not exactly match the conditions of actual application. Both methods require increasing computational power hard to fulfill with current conventional hardware.

Broader Impacts: The acquisition contributes to continue attracting young researchers and aiding in their training. The improved computational capability facilitates the demonstration of speech research to local high school students. The BFOIT Foundation for Opportunities in Information Technology aims to attract more women and underrepresented minorities in computer science and engineering. MRI Development: Heterogeneous, Autonomic Wireless Control Networks for Scalable Cyber Physical Systems Proposal #: CNS 09 23518 Institution: University of Denver
PI(s): Voyles, Richard M.; Denver, CO 80208 0000 Mangharam, Rahul; Anaraki Siavash Pourkamali; Rutherford, Matthew J.; Valavanis, Kimon P.
Title: MRI/Dev.: Heterogeneous, Autonomic Wireless Control Networks for Scalable Cyber Physical Systems
Project Proposed:
This collaborative project, creating an instrument consisting of a new class of heterogeneous wireless sensor actuator controller platforms, facilitates a wide range of experimental research on Networked CyberPhysical Systems (CPS). A key aim is to arrive at standardization for hardware and software interfaces over the platform categories that will support protocols for time and safety critical applications. Involving four universities (U Denver, Notre Dame U., U Penn, and UT Arlington), three categories of Networked CPS research platforms are developed across a wide range of hardware and software based runtime re configuration. The goals also include developing standardized hardware and software interfaces across these platforms so that nodes may be plug n play, evolve parametrically and programmatically at runtime, and maintain timeliness and reliability as connected objects for control and actuation. Existing computational node prototypes from Penn and U Denver will be refined and harmonized to provide a suite of interoperable nodes. These nodes will have dual radios for the data plane and a passive analog radio for fine grained hardware based global time synchronization to add determination. An Embedded Virtual Machine (EVM), a powerful distributed runtime system where virtual components and their properties are maintained across node boundaries, is introduced to maintain a set of functional invariants, such as control law and para functional invariants such as timeliness constraints, fault tolerance and safety standard across a set of controllers given the spatio temporal changes in the physical network. The EVM software allows tightly coupled communication and runtime control across the different hardware categories. Programming mechanisms treat the set of physical sensors, actuators, and controllers as a single virtual component and allow tasks to be assigned at runtime since the links, nodes, and topology of wireless systems are inherently unreliable. The system is expected to lower the barriers for research into reconfigurable computing across hardware, software, and virtual autonomic computing structures, heterogeneous sensor network timing, synchronization and task allocation strategies, and also serve as a springboard to applications in biomedical modeling, human surveillance and monitoring, and search and rescue robotics. Each node will interface to a suite of modular I/O devices with attendant sensors and actuators. Recent research activity on future wireless sensor networks and applications has been limited to open loop sensing and monitoring giving rise to predominantly event based, asynchronous platforms and systems software. Not much research has been devoted to heterogeneous wireless sensor networks that integrate across a range of computational and communication capabilities. When networks are integrated with higher rate sensors (e.g., video surveillance), actuators with timeliness and safety constraints (e.g., real time control), and networks requiring significant distributed in network processing (e.g., video analytics and autonomous systems), investigators have to go beyond the platforms for low rate sensors and applications for which time stamping is sufficient. Consequently, heterogeneous wireless sensor networks that integrate computational and communication capabilities are necessary.

Broader Impacts:
The project, involving four institutions, provides a range of interoperable control nodes to develop applications from the MEMS/NEMS (Micro/Nano ElectroMechanical Systems) scale to the macro scale, develops building blocks for wireless control networks with applications in search and rescue, industrial automation, medical devices and vehicular control. Students are involved in developing the instrumentation. MRI: Development of a Scalable Energy Efficient Datacenter (SEED) Proposal #: CNS 09 23523
PI(s): Papen, George Fainman, Y.; Vahdat, Amin M.
Institution: University of California San Diego
La Jolla, CA 92093 0934
Title: MRI/Dev.: Development of a Scalable Energy Efficient Datacenter (SEED)

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Project Proposed:
This project, building a Scalable Energy Efficient Datacenter (SEED), develops an integrated solution that encompasses physical layer hardware, protocols, and topologies that can provide the expected size and performance scaling for future data centers while minimizing the cost and energy per switched bit. The work creates the knowledge base required for the development of next generation scalable, energy efficient datacenters. Unique features of this instrument include novel statistical multiplexing modules to reduce connection complexity, a circuit switched optical interconnection fabric, and the ability to accommodate novel protocols, components and subsystems in a realistic system environment. With a design based entirely on commodity components and a non blocking and scalable switch, the baseline configuration of the SEED instrument will connect more than 250 servers, each operating at 10 Gb/s. The fully configured instrument is a hybrid electrical packet/optically circuit switched network designed to efficiently route large data flows into a circuit switched optical core utilizing an optical switch from a previously funded MRI, Quartzite. The instrument supports several newly established multidisciplinary projects including the ERC Center for Integrated Access Networks (CIAN), the MRI GreenLight project, the Center of Interdisciplinary Science for Art, Architecture, and Archaeology (CISA3), and projects at the San Diego Supercomputer center. Specifically, SEED is expected to create the technology base for an order of magnitude improvement in both the cost and energy per switched bit. This will be accomplished by the development of new protocols and topologies, measuring and optimizing application dependent traffic patterns, providing critical system driven specifications of a technology roadmap for the development of novel photonic technologies, and acting as a platform for training the next generation network engineers that are equally versed in both optical and electrical networks. The following four issues are associated with the SEED instrument.
Design of flow scheduling techniques for fat trees that fit both electrical and hybrid systems,
Algorithms for fault tolerance (components in large scale communication switches fail),
Optimal Wavelength Division Multiplexing (WDM) design (uses multiple lasers and transmits several wavelengths of light (lambdas) simultaneously over a single optical fiber)
Technology road map based on findings on performance metrics pertaining to building, testing, and operating the initial optical aggregation, transmission, and switching hardware to inform the Center for Integrated Access Networks (CIAN) ERC.

Broader Impacts: The engine of the 21st century economy, the creation of wealth through information processing, utilizes data centers as its cornerstones. Hence, technologies that can enable larger and more energy efficient information processing will affect many, if not every, aspect of modern life. Access to efficient remote processing should dramatically reduce the amount of physical transport and avoid the expense and human costs of unnecessary commuting, minimize environmental impact from infrastructure and pollution, substantially reduce our dependence on energy imports, improve educational opportunities, enhance the distribution of medical services, and increase overall national security. Thus, the infrastructure to carry these services constitutes a precious national resource, perhaps as precious as the air, rail, and road transportation. Indeed, it should enable this country to better compete globally. MRI: Development of ASSIST: Affordable System for Solar Irrdiance and Tracking This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

This project, building and testing an Affordable System for Solar Irradiance Sensing and Tracking (ASSIST), proposes a tiered architecture where a small number of expensive and highly calibrated observatories get complimented by a larger number of inexpensive, but uncalibrated, ASSIST nodes.
Integrating economic stand alone wireless global irradiance sensors with a new dome sensor that avoids having costly moving parts and automatic solar trackers (ASTs), ASSIST nodes should adapt to the vagaries of wireless communication channel, as well as possible failures of many nodes in the ensemble. The work responds to a major obstacle in developing policies that promote and take advantage of existing solar technologies, that of lack of reliable data for ground solar irradiance (direct normal and global irradiance). Despite well defined and easily calculated radiation reaching outer layers of the atmosphere, solar irradiance reaching ground level (where thermal and photovoltaic solar collectors operate) depends strongly on localized and complex atmospheric conditions. Hence, distributed, embedded environmental sensor systems now enable scientists and engineers to observe environmental systems with previously unattainable spatio temporal resolution. The vision of sensor systems coupled with smart networking, integrated with visualization tools by an overarching cyberinfrastructure is shared by disciplines actively engaged in solar irradiance monitoring all over the world, and is likely to be realized when such systems are developed ahead of the observatory efforts. The system, developed and tested in the heart of California s Central Valley, is coupled with well characterized infrastructure rich solar observatories already deployed. ASSIST aims to serve as a model sensor and information technology system for directly and quantitatively observing the effects of cloud cover, aerosol content, and the presence of participating gases in the lower atmosphere (water vapor, carbon dioxide) and in the stratosphere (ozone), all of which can reduce the availability of direct isolation at ground level to a small fraction of the solar irradiance that reaches the upper atmosphere. From the operational standpoint, the balancing of supply and demand peaks in the electrical grid requires detailed consideration of the availability of solar power as US embraces a more renewable profile of energy utilization. Thus, forecasting the available insolation enables information technology for the success of any policy to include power to the power grid. Engaging students and researchers, this end to end sensor system supporting the observatory scale science in solar systems science provides a well characterized, science driven design test bed in a minority serving university

Broader Impacts: This project enables engineers and scientists to quantify DNI data at spatial and temporal scales currently unavailable. The work develops distributed instruments that are self configurable, without the need of expensive and difficult to maintain mobile parts, and significantly less expensive than current instruments in solar observation technology. The system will be utilized for student experiences; it provides access to important data; and its findings may be adopted by other observatories. In addition to an expected major impact on environmental, CS, electrical, and mechanical research and education directives, the project services student in a minority serving institution. SHF: EAGER: Transactional Processors: Exploiting Hardware Transaction Processing for Reliable Computing With the semiconductor technology entering the nano scale era, CMOS devices are facing a dramatic increase in vulnerability to transient faults such as soft errors induced by energetic particle strikes. Such soft errors have become a major challenge in designing next generation microprocessors. While techniques for optimizing reliability at low levels can be accurate, they incur significantly high hardware overheads and costly manufacturing processes. The objective of the proposed EAGER proposal is to explore a new flexible processor architecture for highly effective reliable computing by exploiting the semantics of hardware transaction processing. Instead of simply augmenting existing processors for an attainable reliability, the PI proposes to exploit the semantics of transaction processing from database management systems and recent transactional memories for the design and implementation of the transactional processor architecture, where the reliable computing is an inherent property. The transactional processor aims to provide highly effective and flexible transaction level verification and native supports for recovery from detected errors. The PI will explore the design space of hardware transaction processing and transaction based reliable computing, as well as new programming language constructs to extend current programming languages for writing programs efficiently in transactions.

The success of this project may result in design of low cost reliable computing platforms based on hardware transaction processing. In addition, the proposed activities will provide a unique channel to attract students from under represented groups and minorities into science and engineering. The PI plans to take advantage of several college and university wide outreach programs to interact with high school students and teachers to motivate them in computer science. NeTS FIND: An Internet Architecture for User Controlled Routes Proposal Number: 0627166
PI: Xiaowei Yang
Institution: University of California, Irvine
Title: NeTS FIND: An Internet Architecture for User Controlled Routes


Abstract

This project is developing an Internet architecture that enables users or their end systems to select the paths their packets take through the network. Such user controlled routes are desirable for both economic and technical reasons. Economically, user controlled routes are a key factor in maintaining the competitiveness of the ISP marketplace, just as long distance carrier selection has had a lasting impact on telephony. Technically, user controlled routes provide a fundamental means for improving the performance and reliability of network communications. This is because they allow end systems to use multiple diverse paths concurrently and reduce the dependence on a single network path that has undesirable characteristics. The key technical difficulties associated with user controlled routes lie in the areas of scalability and robustness. This project directly addresses these issues. It involves the development of the following components: 1) a scalable inter domain routing protocol that distributes policy constrained domain level maps of the network to users and allows them to explicitly formulate domain level routes; 2) an end system protocol stack that allows various applications to take advantage of path diversity; 3) algorithms and protocols that prevent users from rapidly switching paths in a fashion that jeopardizes the overall routing stability; 4) techniques that mitigate the security threats associated with user controlled routes (e.g. source address spoofing, multi path denial of service attacks).

If successful, this work will have a positive economic impact that fosters ISP competition, result in measurable technical improvements in terms of reliability and performance of network communications, and contribute to education. Algorithms, protocols, and software developed in this project will be disseminated through research publications and a project web site. CPS: Small: Control of Distributed Cyber Physical Systems under Partial Information and Limited Communication CPS: Small: Control of Distributed Cyber Physical Systems under Partial Information and Limited Communication

The objective of this research is the development of novel control architectures and computationally efficient controller design algorithms for distributed cyber physical systems with decentralized information infrastructures and limited communication capabilities. Active safety in Intelligent Transportation Systems will be the focus cyber physical application. For the successful development and deployment of cooperative active safety systems, it is critical to develop theory and techniques to design algorithms with guaranteed safety properties and predictable behavior. The approach is to develop a new methodology for the design of communicating distributed hybrid controllers by integrating in a novel manner discrete event controller design and hybrid controller design and optimization.

The methodology to be developed will exploit problem decomposition and will have significant technological impact for a large class of cyber physical systems that share features of modularity in system representation, partial information, and limited communication. The focus on distributed control strategies with limited communication among agents is addressing an important gap in existing control theories for cyber physical systems. The approach will mitigate the computational limitations of existing approaches to control design for hybrid systems.

Given the focus on cooperative active safety in Intelligent Transportation Systems, the results of this effort will have significant societal impact in terms of increased traffic safety and reduced number and severity of accidents. The broader impacts of this proposal also include involvement of high school and undergraduate students and curriculum development by incorporating results of research into existing courses on cyber physical systems. CPS:Small: Collaborative Research: Distributed Coordination of Agents for Air Traffic Flow Management CPS: Small: Collaborative Research: Distributed Coordination of Agents For Air Traffic Flow Management

This objective of this proposal is to improve the management of the air traffic system, a cyber physical system where the need for a tight connection between the computational algorithms and the physical system is critical to safe, reliable and efficient performance.
The approach is based on an adaptive multiagent coordination algorithm with a particular emphasis on the systematic selection of the agents, their actions and the agents reward functions.

The intellectual merit lies in addressing the agent coordination problem in a physical setting by shifting the focus from ``how to learn to ``what to learn.
This paradigm shift allows a separation between the learning algorithms used by agents, and the reward functions used to tie those learning systems into system performance. By exploring agent reward functions that implicitly model agent interactions based on feedback from the real world, this work aims to build cyber physical systems where an agent that learns to optimize its own reward leads to the optimization of the system objective function.

The broader impact is in providing new air traffic flow management algorithms that will significantly reduce air traffic congestion. The potential impact cannot only be measured in currency ($41B loss in 2007) but in terms of improved experience by all travelers, providing a significant benefit to society. In addition, the PIs will use this project to train graduate and undergraduate students (i) by developing new courses in multiagent learning for transportation systems; and (ii) by providing summer internship opportunities at NASA Ames Research Center. CPS: Small: Collaborative Research: Methods and Tools for the Verification of Cyber Physical Systems CPS:Small:Collaborative Research: Methods and Tools for the Verification of Cyber Physical Systems

The objective of this research is to investigate and develop methods and tools for the analysis and verification of cyber physical systems. The approach is to augment the methods and tools that have been developed at the University of Utah and the University of South Florida for modeling and verification of asynchronous and analog/mixed signal circuits to address challenges in cyber physical system verification.

This research will develop a unified framework with methods and tools which include an integrated formalism to comprehensively model discrete/continuous, functional/timing, synchronous/asynchronous, and deterministic/stochastic behavior. These tools will also include algorithms to analyze behavior and verify that it satisfies the correctness requirements on functionality, timing, and robustness. Finally, they will include abstraction and compositional reasoning approaches to enable large systems to be analyzed and verified efficiently.

Since cyber physical systems are becoming ubiquitous, improvements in such systems such as higher reliability, better fault tolerance, improved performance, and lower design costs will have tremendous positive impact on society. Results from this research will be transferred to the cyber physical systems community and other application domains by both publishing papers in related conferences and journals as well as by freely distributing tools via the Internet. Both graduate and undergraduate students will be engaged in this multi institutional research where they will be exposed to the latest research in formal and probabilistic analysis. Early involvement of undergraduate students may help encourage them to attend graduate school. This research project will also recruit underrepresented and female students to allow it to reach broader audiences. CPS:Medium:Collaborative Research: Infrastructure and Technology Innovations for Medical Device Coordination The objective of this research is to develop a framework for the development
and deployment of next generation medical systems consisting of integrated and
cooperating medical devices. The approach is to design and implement an
open source medical device coordination framework and a model based component
oriented programming methodology for the device coordination, supported by a
formal framework for reasoning about device behaviors and clinical workflows.

The intellectual merit of the project lies in the formal foundations of the
framework that will enable rapid development, verification, and certification
of medical systems and their device components, as well as the clinical
scenarios they implement. The model based approach will supply evidence for the regulatory approval process, while run time monitoring components embedded
into the system will enable black box recording capabilities for the forensic
analysis of system failures. The open source distribution of tools supporting
the framework will enhance its adoption and technology transfer.

A rigorous framework for integrating and coordinating multiple medical devices
will enhance the implementation of complicated clinical scenarios and reduce
medical errors in the cases that involve such scenarios.

Furthermore, it will speed up and simplify the process of regulatory approval for coordination enabled medical devices, while the formal reasoning framework will improve the confidence in the design process and in the approval decisions.

Overall, the framework will help reduce costs and improve the quality of the
health care. CPS: Small: Design of Networked Control Systems for Chemical Processes CPS: Small: Design of Networked Control Systems for Chemical Processes

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The objective of the proposed research program is to develop, for the first time, the theory and methods needed for the design of networked control systems for chemical processes and demonstrate their application and effectiveness in the context of process systems of industrial importance.

The proposed approach to achieving this objective involves the development of a novel mathematical framework based on nonlinear asynchronous systems to model the sensor and actuator network behavior accounting explicitly for the effect of asynchronous and delayed measurements, network communication and actuation. Within the proposed asynchronous systems framework, novel control methods will be developed for the design of nonlinear networked control systems that improve closed loop stability, performance and robustness. The controller design methods will be based on nonlinear and predictive control theory and will have provable closed loop properties.

The development and implementation of networked control methods which take advantage of sensor and actuator networks is expected to significantly improve the operation and performance of chemical processes, increase process safety and reliability, and minimize the negative economic impact of process failures, thereby impacting directly the US economy. The integration of the research results into advanced level classes in process control and the writing of a new book on ``Networked Process Control will benefit students and researchers in the field. The development of software, short courses and workshops and the on going interaction of the PIs with an industrial consortium will be the means for transferring the results of this research into the industrial sector. Furthermore, the involvement of a diverse group of undergraduate and graduate students in the research will be pursued. CPS:Medium:Collaborative Research: Physical Modeling and Software Synthesis for Self Reconfigurable Sensors in River Environments The objective of this research is the transformation from static sensing into mobile, actuated sensing in dynamic environments, with a focus on sensing in tidally forced rivers. The approach is to develop inverse modeling techniques to sense the environment, coordination algorithms to distribute sensors spatially, and software that uses the sensed environmental data to enable these coordination algorithms to adapt to new sensed conditions.

This work relies on the concurrent sensing of the environment and actuation of those sensors based on sensed data. Sensing the environment is approached as a two layer optimization problem. Since mobile sensors in dynamic environments may move even when not actuated, sensor coordination and actuation algorithms must maintain connectivity for the sensors while ensuring those sensors are appropriately located. The algorithms and software developed consider the time scales of the sensed environment, as well as the motion capabilities of the mobile sensors. This closes the loop from sensing of the environment to actuation of the devices that perform that sensing.

This work is addresses a challenging problem: the management of clean water resources. Tidally forced rivers are critical elements in the water supply for millions of Californians. By involving students from underrepresented groups, this research provides a valuable opportunity for students to develop an interest in engineering and to learn first hand about the role of science and engineering in addressing environmental issues. CPS Small: Control of Surgical Robots: Network Layer to Tissue Contact CPS Small: Control of Surgical Robots: Network Layer to Tissue Contact

Research Objectives: This proposed CPS project aims to enable intelligent telesurgery in which a surgeon, or a distributed team of surgeons, can work on tiny regions in the body with minimal access. The University of Washington will expand an existing open surgical robot testbed, and create a robust infrastructure for cyber physical systems with which to extend traditional real time control and teleoperation concepts by adding three new interfaces to the system: networking, intelligent robotics, and novel non linear controllers.

Intellectual Merit: This project aims to break new ground beyond teleoperation by adding advanced robotic functions. Equally robust and flexible networking, high level interfaces, and novel controllers will be added to the existing sytsem. The resulting system will be an open architecture and a substrate upon which many cyber physical system ideas and algorithms will be tested under realistic conditions. The platforms proven physical robustness will permit rigorous evaluation of results and the open interfaces will encourage collaboration and sharing of results.

Broader Impacts: We expect the results to enable new research in multiple ways. First, the collaborators such as Johns Hopkins, U.C. Santa Cruz, and several foreign institutions will be able to remotely connect to new high level interfaces provided by this project. Second, for the first time a robust and completely open surgical telerobot will be available for research so that CPS researchers do not need to be limited to isolated toy problems but instead be able to prototype advanced surgical robotics techniques and evaluate them in realistic contexts including animal procedures. CPS:Medium:Collaborative Research:Physical modeling and software The objective of this research is the transformation from static sensing into mobile, actuated sensing in dynamic environments, with a focus on sensing in tidally forced rivers. The approach is to develop inverse modeling techniques to sense the environment, coordination algorithms to distribute sensors spatially, and software that uses the sensed environmental data to enable these coordination algorithms to adapt to new sensed conditions.

This work relies on the concurrent sensing of the environment and actuation of those sensors based on sensed data. Sensing the environment is approached as a two layer optimization problem. Since mobile sensors in dynamic environments may move even when not actuated, sensor coordination and actuation algorithms must maintain connectivity for the sensors while ensuring those sensors are appropriately located. The algorithms and software developed consider the time scales of the sensed environment, as well as the motion capabilities of the mobile sensors. This closes the loop from sensing of the environment to actuation of the devices that perform that sensing.

This work is addresses a challenging problem: the management of clean water resources. Tidally forced rivers are critical elements in the water supply for millions of Californians. By involving students from underrepresented groups, this research provides a valuable opportunity for students to develop an interest in engineering and to learn first hand about the role of science and engineering in addressing environmental issues. CPS: Small: Dynamically Managing the Real time Fabric of a Wireless Sensor Actuator Network CPS: Small: Dynamically Managing the Real time Fabric of a Wireless Sensor Actuator Network

The objective of this research is to develop algorithms for wireless sensor actuator networks (WSAN) that allow control applications and network servers to work together in maximizing control application performance subject to hard real time service constraints. The approach is a model based approach in which the WSAN is unfolded into a real time fabric that captures the interaction between the network s cyber processes and the application s physical processes.

The project s approach faces a number of challenges when they are applied to wireless control systems. This project addresses these challenges by 1) using network calculus concepts to pose a network utility maximization (NUM) problem that maximizes overall application performance subject to network capacity constraints, 2) using event triggered message passing schemes to reduce communication overhead, 3) using nonlinear analysis methods to more precisely characterize the problem s utility functions, and 4) using anytime control concepts to assure robustness over wide variations in network connectivity.

The project s impact will be broadened through interactions with industrial partner, EmNet LLC. The company will use this project s algorithms on its CSOnet system. CSOnet is a WSAN controlling combined sewer overflows (CSO), an environmental problem faced by nearly 800 cities in the United States. The project s impact will also be broadened through educational outreach activities that develop a graduate level course on formal methods in cyber physical systems. The project s impact will be broadened further through collaborations with colleagues working on networked control systems under the European Union s WIDE project. CPS: Medium: Quantitative Analysis and Design of Control Networks The objective of this research is to develop the scientific foundation for the quantitative analysis and design of control networks. Control networks are wireless substrates for industrial automation control, such as the WirelessHART and Honeywell s OneWireless, and have fundamental differences over their sensor network counterparts as they also include actuation and the physical dynamics. The approach of the project focuses on understanding cross cutting interfaces between computing systems, control systems, sensor networks, and wireless communications using time triggered architectures.

The intellectual merit of this research is based on using time triggered communication and computation as a unifying abstraction for understanding control networks. Time triggered architectures enable the natural integration of communication, computation, and physical aspects of control networks as switched control systems. The time triggered abstraction will serve for addressing the following interrelated themes: Optimal Schedules via Quantitative Automata, Quantitative Analysis and Design of Control Networks: Wireless Protocols for Optimal Control: Quantitative Trust Management for Control Networks.

Various components of this research will be integrated into the PIs RAVEN control network which is compatible with both WirelessHART and OneWireless specifications. This provides a direct path for this proposal to have immediate industrial impact. In order to increase the broader impact of this project, this project will launch the creation of a Masters program in Embedded Systems, one of the first in the nation. The principle that guides the curriculum development of this novel program is a unified systems view of computing, communication, and control systems. CPS: Medium: Learning for Control of Synthetic and Cyborg Insects in Uncertain Dynamic Environments The objective of this research is to enable operation of synthetic and
cyborg insects in complicated environments, such as outdoors or in a
collapsed building. As the mobile platforms and environment have
significant uncertainty, learning and adaptation capabilities are
critical. The approach consists of three main thrusts to enable the
desired learning and adaptation: (i) Development of algorithms to
efficiently learn optimal control policies and dynamics models through
sharing the learning and adaptation between various instantiations of
platforms and environments. (ii) Creation of control learning
algorithms which can be run on low cost, low power mobile platforms.
(iii) Development of algorithms for online improvement of policy
performance in a minimal number of real world trials.

The proposed research will advance learning and adaptation
capabilities of practical cyberphysical systems. The proposed
approach will be generally applicable and lead to a new class of
learning and adapting systems that are able to leverage shared
properties between multiple tasks to significantly speed up
learning and adaptation.

Success in this research project will bring society closer to solving
the grand challenge of teams of mobile, disposable, search and rescue
robots which can robustly locomote through uncertain and novel
environments, finding survivors in disaster situations, while removing
risk from rescuers. This project will provide interdisciplinary
training through research and classwork for undergraduate and graduate
students in creating systems which intimately couple the cyber and
physical aspects in robotic and living mobile platforms. Through the
SUPERB summer program, under represented students in engineering will
experience research in learning and robotics. CPS: Medium: Learning to Sense Robustly and Act Effectively The physical environment of a cyber physical system is unboundedly complex, changing continuously in time and space. An embodied cyber physical system, embedded in the physical world, will receive a high bandwidth stream of sensory information, and may have multiple effectors with continuous control signals. In addition to dynamic change in the world, the properties of the cyber physical system itself ? its sensors and effectors ? change over time. How can it cope with this complexity? The hypothesis behind this proposal is that a successful cyber physical system will need to be a learning agent, learning the properties of its sensors, effectors, and environment from its own experience, and adapting over time. Inspired by human developmental learning, the assertion is that foundational concepts such as Space, Object, Action, etc., are essential for such a learning agent to abstract and control the complexity of its world. To bridge the gap between continuous interaction with the physical environment, and discrete symbolic descriptions that support effective planning, the agent will need multiple representations for these foundational domains, linked by abstraction relations. To achieve this, the team is developing the Object Semantic Hierarchy (OSH), which shows how a learning agent can create a hierarchy of representations for objects it interacts with. The OSH shows how the ?object abstraction? factors the uncertainty in the sensor stream into object models and object trajectories. These object models then support the creation of action models, abstracting from low level motor signals. To ensure generality across cyber physical systems, these methods make only very generic assumptions about the nature of the sensors, effectors, and environment. However, to provide a physical test bed for rapid evaluation and refinement of our methods, the team has designed a model laboratory robotic system to be built from off the shelf components, including a stereo camera, a pan tilt translate base, and a manipulator arm. For dissemination and replication of research results, the core system will be affordable and easily duplicated at other labs. There are plans to distribute the plans, the control software, and the software for experiments, to encourage other labs to replicate and extend the work. The same system will serve as a platform for an open ended set of undergraduate laboratory tasks, ranging from classroom exercises, to term projects, to independent study projects. There is a preliminary design for a very inexpensive version of the model cyberphysical system that can be constructed from servo motors and pan tilt webcams, for use in collaborating high schools and middle schools, to communicate the breadth and excitement of STEM research. CPS: Small: Image Guided Autonomous Optical Manipulation of Cell Groups CPS: Small: Image Guided Autonomous Optical Manipulation of Cell Groups

The objective of this research is to create computational foundation, methods, and tools for efficient and autonomous optical micromanipulation using microsphere ensembles as grippers. The envisioned system will utilize a holographic optical tweezer, which uses multiple focused optical traps to position microspheres in three dimensional space. The proposed approach will focus on the following areas. First, it will provide an experimentally validated optical tweezers based workstation for concurrent manipulation of multiple cells. Second, it will provide algorithms for on line monitoring of workspace to support autonomous manipulation. Finally, it will provide real time image guided motion planning strategies for transporting microspheres ensembles.

The proposed work will lead to a new way of autonomously manipulating difficult to trap or sensitive objects using microspheres ensembles as reconfigurable grippers. The proposed work will also lead to fundamental advances in several cyber physical systems areas by providing new approaches to micromanipulations, fast and accurate algorithms with known uncertainty bounds for on line monitoring of moving microscale objects, and real time motion planning algorithms to transport particle ensembles.

The ability to quickly and accurately manipulate individual cells with minimal training will enable researchers to conduct basic research at the cellular scale. Control over cell cell interactions will enable unprecedented insights into cell signaling pathways and open up new avenues for medical diagnosis and treatment. The proposed integration of research with education will train students with a strong background in emerging robotics technologies and the inner workings of cells. These students will be in a unique position to rapidly develop and deploy specialized robotics technologies. CPS: Medium: Vehicular Cyber Physical Systems The objective of this research is to develop technologies to improve
the efficiency and safety of the road transportation infrastructure.
The approach is to develop location based vehicular services combining
on board automotive computers, in car devices, mobile phones, and
roadside monitoring/surveillance systems. The resulting vehicular
Cyber Physical Systems (CPS) can reduce travel times with smart
routing, save fuel and reduce carbon emissions by determining greener
routes and commute times, improve safety by detecting road hazards,
change driving behavior using smart tolling, and enable
measurement based insurance plans that incentivize good driving.

This research develops distributed algorithms for predictive travel
delay modeling, feedback based routing, and road hazard assessment.
It develops privacy preserving protocols for capturing and analyzing
data and using it for tasks such as congestion aware tolling. It also
develops a secure macro tasking software run time substrate to ensure
that algorithms can be programmed centrally without explicitly
programming each node separately, while ensuring that it is safe to
run third party code. The research focuses on re usable methods that
can benefit multiple vehicular services, and investigates which
lessons learned from this vehicular CPS effort generalize to other
situations.

Road transportation is a grand challenge problem for modern society,
which this research can help overcome. Automobile vendors, component
developers, and municipal authorities have all shown interest in deployment.
The education plan includes outreach to local K 12 students and a new
undergraduate course on transportation from a CPS perspective, which
will involve term projects using the data collected in the project CPS: Small: MPSoC based Control and Scheduling Co design for Battery Powered Cyber Physical Systems CPS: Small: MPSoC based Control and Scheduling Co design for Battery Powered Cyber Physical Systems

The objective of this research is to develop new scientific and engineering principles, algorithms and models for the design of battery powered cyber physical systems whose computational substrates include high performance multiprocessor systems on chip. The approach is to design control tasks that guarantee performance and meet criteria for battery operation time. Task schedulers are co designed to balance the computing load across the multiple processors, and to control the physical plant together with the control tasks. The controller and scheduler will be integrated with battery management algorithms through a systems theory approach so that the methods are provably correct with justfiable performance.

Intellectual Merit: The program will create progress in digital and hybrid control theory that keeps up with the recent trend of using multiprocessor systems on chips for control and robotic applications. The mechanism for the migration of control tasks between multiple processors will respect physical and thermal performance. A novel battery dynamic discharge model is developed, which may be applied to context when the discharge current of batteries cannot be predicted by existing static battery models.

Broader Impacts: Collaborations with industrial partners have been set up. The program offers multidisciplinary training in cyber physical systems. A teaching and outreach lab is in place to host K 12 student teams that participate in robot competitions, and has become an Explorer Post for Boy Scout of America. CPS:Medium:Collaborative Research:Monitoring Human Performance with Wearable Accelerometers This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The objective of this research is to develop a cyber physical system composed of accelerometers and novel machine learning algorithms to analyze data in the context of a set of driving health care applications. The approach is to develop novel machine learning algorithms for temporal segmentation, classification, and detection of subtle elements of human motion. These techniques will allow quantification of human motion and improved full time monitoring and assessment of medical conditions using a lightweight wearable system. The scientific contribution of this research is in advancing machine learning and human sensing in support of improved medical diagnoses and treatment monitoring by (i) modeling human activity and symptoms through sensor data analysis, (ii) integrating and fusing information from several accelerometers to monitor in real time, (iii) validating the efficacy of the automated detection through assessments applying the state of the art in diagnostic evaluation, (iv) developing novel machine learning methods for temporal segmentation, classification, and discovery of multiple temporal patterns that discriminate between temporal signals, and (v) providing quality measures to characterize subtle human motion. These algorithms will advance machine learning in the area of unsupervised and semisupervised learning. The driving applications for this research are job coaching for people with cognitive disabilities, tele rehabilitation for knee osteo arthritis, assessing variability in balance and gait as an indicator of health of older adults, and measures for assessing Parkinson s patients. This research is highly interdisciplinary and will train graduate students for careers in developing technological innovations in health and monitoring systems. CPS: Medium: Active Heterogeneous Sensing for Fall Detection and Fall Risk Assessment This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The objective of this research is to study active sensing and adaptive fusion using vision and acoustic sensors for continuous, reliable fall detection and assessment of fall risk in dynamic and unstructured home environments. The approach is to incorporate active vision with infrared light sources and camera controls, an acoustic array that identifies the sound characteristics and location, and sensor fusion based on the Choquet integral and hierarchical fuzzy logic systems that supports uncertain heterogeneous sensor data at varying time scales, qualitative data, and risk factors.
The project will advance the state of the art in (1) active vision sensing for human activity recognition in dynamic and unpredictable environments, (2) acoustic sensing in unstructured environments, (3) adaptive sensor fusion and decision making using heterogeneous sensor data in dynamic and unpredictable environments, (4) automatic fall detection and fall risk assessment using non wearable sensors, and (5) algorithms for cyber physical systems that address the interplay of anomaly detection (falls) and risk factors affecting the likelihood of an anomaly event.
The project will impact the health care and quality of life for older adults. New approaches will assist health care providers to identify potential health problems early, offering a model for eldercare technology that keeps seniors independent while reducing health care costs. The project will train the next generation of researchers to handle real, cyber physical systems. Students will be mentored, and research outcomes will be integrated into the classroom. Novel outreach activities are planned to reach the elderly community and the general public CPS: Small: Control Subject to Human Behavioral Disturbances CPS: Small: Control Subject to Human Behavioral Disturbances

The objective of this research is to develop an integrated methodology for control system design in situations where disturbances primarily result from routine human behavior, as, for example, in future artificial pancreas systems where meals and exercise are the main disturbances affecting blood glucose concentration. The approach is to recognize that human behavioral disturbances (i) are generally random but cannot be treated as zero mean white noise processes and (ii) occur with statistical regularity but cannot be treated as periodic due to natural variation in human behavior. This emerging class of problems requires (i) the derivation of new mathematical representations of disturbances for specific applications and (ii) the formulation of new stochastic control models and algorithms that exploit statistical regularity in the disturbance process.

The intellectual merit of the proposed research stems from the fact that it explicitly recognizes a new class of disturbances, human behavioral disturbances, seeking to develop an integrated approach to statistically characterizing and responding to future perturbations, adapting gracefully to uncertainty about the future. The anticipated research outcomes will be relevant in diverse fields, including stochastic hybrid control and human automation interaction.

As a broader implication, the proposed research will enable the design of future field deployable artificial pancreas systems, potentially improving the lives of 1.5 million Americans suffering from Type 1 diabetes. With help from the two graduate students funded by the project, the principle investigator will supervise a ?Capstone? design course, exposing undergraduates to various aspects of control under human behavioral disturbances. CPS:Medium: Tightly Integrated Perception and Planning in Intelligent Robotics This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The objective of this research is to develop truly intelligent, automated driving through a new paradigm that tightly integrates probabilistic perception and deterministic planning in a formal, verifiable framework. The interdisciplinary approach utilizes three interlinked tasks. Representations develops new techniques for constructing and maintaining representations of a dynamic environment to facilitate higher level planning. Anticipation and Motion Planning develops methods to anticipate changes in the environment and use them as part of the planning process. Verifiable Task Planning develops theory and techniques for providing probabilistic guarantees for high level behaviors. Ingrained in the approach is the synergy between theory and experiment using an in house, fully equipped vehicle.
The recent Urban Challenge showed the current brittleness of autonomous driving, where small perception mistakes would propagate into planners, causing near misses and small accidents; Fundamentally, there is a mismatch between probabilistic perception and deterministic planning, leading to reactive rather than intelligent behaviors. The proposed research directly addresses this by developing a single, unified theory of perception and planning for intelligent cyber physical systems.
Near term, this research could be used to develop advanced safety systems in cars. The elderly and physically impaired would benefit from inexpensive, advanced automation in cars. Far term, the advanced intelligence could lead to automated vehicles for applications such as cooperative search and rescue. The research program will educate students through interdisciplinary courses in computer science and mechanical engineering, and experiential learning projects. Results will be disseminated to the community including under represented colleges and universities. CPS: Small: Control Design for CyberPhysical Systems Using Slow Computing CPS:Small:Control Design for CyberPhysical Systems Using Slow Computing

The objective of this research is to develop principles and tools for the
design of control systems using highly distributed, but slow, computational
elements. The approach of this research is to build an architecture that
uses highly parallelized, simple computational elements incorporating
nonlinearities, time delay and asynchronous computation as integral design
elements. Tools for the design of non deterministic protocols will be
developed and demonstrated using an existing multi vehicle testbed at
Caltech.

The motivation for using slow computing is to develop new feedback control
architectures for applications where computational power is extremely
limited. Examples of such systems are those where the energy usage of the
system must remain small, either due to the source of power available
(e.g. batteries or solar cells) or the physical size of the device
(e.g. microscale and nanoscale robots). A longer term application area is
in the design of control systems using novel computing substrates, such as
biological circuits. A critical element in both cases is the tight coupling
between the dynamics of the underlying process and the temporal properties
of the algorithm that is controlling it.

The implementation plan for this project involves students from multiple
disciplines (including bioengineering, computer science, electrical
engineering and mechanical engineering) as well as at multiple experience
levels (sophomores through PhD students) working together on a set of
interlinked research problems. The project is centered in the Control and
Dynamical Systems department at Caltech, which has a strong record of
recruiting women and underrepresented minority students into its programs. CPS:Medium:Cyber Enabled Efficient Energy Management of Structures (CEEMS) This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The objective of this research is the development of methods for the control of energy flow in buildings, as enabled by cyber infrastructure. The approach is inherently interdisciplinary, bringing together electrical and mechanical engineers alongside computer scientists to advance the state of the art in simulation, design, specification and control of buildings with multiple forms of energy systems, including generation and storage. A significant novelty of this project lies in a fundamental view of a building as a set of overlapping, interacting networks. These networks include the thermal network of the physical building, the energy distribution network, the sensing and control network, as well as the human network, which in the past have been considered only separately. This work thus seeks to develop methods for simulating, optimizing, modeling, and control of complex, heterogeneous networks, with specific application to energy efficient buildings. The advent of maturing distributed and renewable energy sources for on site cooling, heating, and power production and the concomitant developments in the areas of cyberphysical and microgrid systems present an enormous opportunity to substantially increase energy efficiency and reduce energy related emissions in the commercial building energy sector. In addition, there is a direct impact of the proposed work in training students with backgrounds in the unique blend of engineering and computer science that is needed for the study of cyber enabled energy efficient management of structures, as well as planned interactions at the undergraduate and K 12 level. CPS: Small: Mathematical, Computational, and Perceptual Foundations for Interactive Cyber Physical Systems CPS:Small:Mathematical, Computational, and Perceptual Foundations for Interactive Cyber Physical Systems

The objective of this research is to create interfaces that enable people with impaired sensory motor function to control interactive cyber physical systems such as artificial limbs, wheelchairs, automobiles, and aircraft. The approach is based on the premise that performance can be significantly enhanced merely by warping the perceptual feedback provided to the human user. A systematic way to design this feedback will be developed by addressing a number of underlying mathematical and computational challenges.

The intellectual merit lies in the way that perceptual feedback is constructed. Local performance criteria like stability and collision avoidance are encoded by potential functions, and gradients of these functions are used to warp the display. Global performance criteria like optimal navigation are encoded by conditional probabilities on a language of motion primitives, and metric embeddings of these probabilities are used to warp the display. Together, these two types of feedback facilitate improved safety and performance while still allowing the user to retain full control over the system.

If successful, this research could improve the lives of people suffering from debilitating physical conditions such as amputation or stroke and also could protect people like drivers or pilots that are impaired by transient conditions such as fatigue, boredom, or substance abuse. Undergraduate and graduate engineering students will benefit through involvement in research projects, and K 12 students and teachers will benefit through participation in exhibits presented at the Engineering Open House, an event hosted annually by the College of Engineering at the University of Illinois. CPS:Small:Collaborative Research:Establishing Integrity in Dynamic Networks of Cyber Physical Devices CPS:Small:Collaborative Research: Establishing Integrity in Dynamic Networks of Cyber Physical Devices


The objective of this research is to develop energy efficient integrity
establishment techniques for dynamic networks of cyber physical devices.
In such dynamic networks, devices connect opportunistically and perform
general purpose computations on behalf of other devices. However, some
devices may be malicious in intent and affect the integrity of computation.
The approach is to develop new trust establishment mechanisms for dynamic
networks. Existing trusted computing mechanisms are not directly applicable
to cyber physical devices because they are resource intensive and require
devices to have special purpose hardware.

This project is addressing these problems along three research prongs.
The first is a comprehensive study of the resource bottlenecks in current
trust establishment protocols. Second, the insights from this study are
being used to develop resource aware attestation protocols for cyber
physical devices that are equipped with trusted hardware. Third, the
project is developing new trust establishment protocols for cyber physical
devices that may lack trusted hardware. A key outcome of the project is
an improved understanding of the tradeoffs needed to balance the concerns
of security and resource awareness in dynamic networks.

Dynamic networks allow cyber physical devices to form a highly distributed,
cloud like infrastructure for computations involving the physical world. The
trust establishment mechanisms developed in this project encourage devices to
participate in dynamic networks, thereby unleashing the full potential of
dynamic networks. This project includes development of dynamic networking
applications, such as distributed gaming and social networking, in
undergraduate curricula and course projects, thereby fostering the
participation of this key demographic. CPS: Small: Sensor Network Information Flow Dynamics The objective of this research is to develop numerical techniques for solving partial differential equations (PDE) that govern information flow in dense wireless networks. Despite the analogy of information flow in these networks to physical phenomena such as thermodynamics and fluid mechanics, many physical and protocol imposed constraints make information flow PDEs unique and different from the observed PDEs in physical phenomena. The approach is to develop a systematic method where a unified framework is capable of optimizing a broad class of objective functions on the information flow in a network of a massive number of nodes. The objective function is defined depending on desired property of the geometric paths of information. This leads to PDEs whose form varies depending on the optimization objective. Finally, numerical techniques will be developed to solve the PDEs in a network setting and in a distributed manner.

The intellectual merits of this project are: developing mathematical tools that address a broad range of design objectives in large scale wireless sensor networks under a unified framework; initiating a new field on numerical analysis of information flow in dense wireless networks; and developing design tools for networking problems such as transport capacity, routing, and load balancing.

The broader impacts of this research are: helping the development of next generation wireless networks; encouraging involvement of undergraduate students and underrepresented groups, and incorporating the research results into graduate level courses. Additionally, the research is interdisciplinary, bringing together sensor networking, theoretical physics, partial differential equations, and numerical optimization. CPS:Small: Transforming a City s Transportation Infrastructure through an Embedded Pervasive Communication Network CPS: Small: Transforming a City s Transportation Infrastructure through an Embedded Pervasive Communication Network

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The objective of this inter disciplinary research is to develop new technologies to transform the streets of a city into a hybrid transportation/communication system, called the Intelligent Road (iRoad), where autonomous wireless devices are co located with traffic signals to form a wireless network that fuses real time transportation data from all over the city to make a wide range of new applications possible. The approach is to build new capacities of quantitative bandwidth distribution, rate/delay assurance, and location dependent security on a pervasive wireless platform through distributed queue management, adaptive rate control, and multi layered trust. These new capacities lead to transformative changes in the way the transportation monitoring and control functions are designed and operated.

Many technical challenges faced by the iRoad system are open problems. New theories/protocols developed in this project will support sophisticated bandwidth management, quality of service, multi layered trust, and information fusion in a demanding environment where critical transportation functions are implemented. Solving these fundamental problems advances the state of the art in both wireless technologies and transportation engineering. The research outcome is likely to be broadly applicable in other wireless systems.

The economic and societal impact of the iRoad system is tremendous at a time when the country is modernizing its ailing transportation infrastructure. It provides a pervasive communication infrastructure and engineering framework to build new applications such as real time traffic map, online best route query, intelligent fuel efficient vehicles, etc. The research results will be disseminated through course materials, academic publication, industry connection, and presentations at the local transportation department. CPS: Small: Collaborative Research: Foundations of Cyber Physical Networks CPS: Small: Collaborative Research: Foundations of Cyber Physical Networks
The objective of this research is to investigate the foundations,
methodologies, algorithms and implementations of cyberphysical
networks in the context of medical applications. The approach is to
design, implement and study Carenet, a medical care network, by
investigating three critical issues in the design and construction
of cyberphysical networks: (1) rare event detection and
multidimensional analysis in cyberphysical data streams, (2)
reliable and trusted data analysis with cyberphysical networks,
including veracity analysis for object consolidation and redundancy
elimination, entity resolution and information integration, and
feedback interaction between cyber and physical networks, and (3)
spatiotemporal data analysis including spatiotemporal cluster
analysis, sequential pattern mining, and evolution of cyberphysical
networks.
Intellectual merit: This project focuses on several most pressing
issues in large scale cyberphysical networks, and develops
foundations, principles, methods, and technologies of cyberphysical
networks. It will deepen our understanding of the foundations,
develop effective and scalable methods for mining such networks,
enrich our understanding of cyberphysical systems, and benefit many
mission critical applications. The study will enrich the principles
and technologies of both cyberphysical systems and information
network mining.
Broader impacts: The project will integrate multiple disciplines,
including networked cyberphysical systems, data mining, and
information network technology, and advance these frontiers. It will
turn raw data into useful knowledge and facilitate strategically
important applications, including the analysis of patient networks,
combat networks, and traffic networks. Moreover, the project
systematically generates new knowledge and contains a comprehensive
education and training plan to promote diversity, publicity, and
outreach. CPS: Small: Compositionality and Reconfiguration for Distributed Hybrid Systems CPS: Small: Compositionality and Reconfiguration for Distributed Hybrid Systems

The objective of this research is to address fundamental challenges in the verification and analysis of reconfigurable distributed hybrid control systems. These occur frequently whenever control decisions for a continuous plant depend on the actions and state of other participants. They are not supported by verification technology today. The approach advocated here is to develop strictly compositional proof based verification techniques to close this analytic gap in cyber physical system design and to overcome scalability issues. This project develops techniques using symbolic invariants for differential equations to address the analytic gap between nonlinear applications and present verification techniques for linear dynamics.

This project aims at transformative research changing the scope of systems that can be analyzed. The proposed research develops a compositional proof based approach to hybrid systems verification in contrast to the dominant automata based verification approaches. It represents a major improvement addressing the challenges of composition, reconfiguration, and nonlinearity in system models.

The proposed research has significant applications in the verification of safety critical properties in next generation cyber physical systems. This includes distributed car control, robotic swarms, and unmanned aerial vehicle cooperation schemes to full collision avoidance protocols for multiple aircraft. Analysis tools for distributed hybrid systems have a broad range of applications of varying degrees of safety criticality, validation cost, and operative risk. Analytic techniques that find bugs or ensure correct functioning can save lives and money, and therefore are likely to have substantial economic and societal impact. CPS: Small: Programming Environment and Architecture for Situational Awareness and Response CPS: Small: Programming Environment and Architecture for Situational Awareness and Response

The objective of this research is to investigate and implement a software architecture to improve productivity in the development of rapidly deployable, robust, real time situational awareness and response applications. The approach is based on a modular cross layered architecture that combines a data centric descriptive programming model with an overlay based communication model. The cross layer architecture will promote an efficient implementation. Simultaneously, the data centric programming model and overlay based communication model will promote a robust implementation that can take advantage of heterogeneous resources and respond to different failures.

There is currently no high level software architecture that meets the stringent requirements of many situational awareness and response applications. The proposed project will fill this void by developing a novel data centric programming model that spans devices with varying computational and communication capabilities. Similarly, the overlay communication model will extend existing work by integrating network resources with the programming model. This cross layer design will promote the implementation of efficient and robust applications.

This research will benefit society by providing emergency responders with software tools that present key information in a timely fashion. This, in turn, will increase safety and reduce economic and human loss during emergencies. The productivity gains in deploying sensors and mobile devices will benefit other domains, such as field research using sensor networks. Software will be released under an open source license to promote the use by government agencies, research institutions, and individuals. Products of this research, including the software, will be used in courses at the University
of North Carolina. CPS:Small:Non Volatile Computing for Embedded Cyber Physical Systems The objective of this research is to develop non volatile computing devices, which allow the power source to be cut off at any time, and yet resume regular operation without loss of information when the power comes back. The approach is to replace all critical memory components with non volatile units so that computing state is maintained over power interruptions. The advancement in new Flash memory devices makes this approach feasible by enabling low voltage program/erase (P/E) around ±2V and a long (projected >1016) cycling endurance to be integrated into CMOS technology.

This research effort seeks to establish a new paradigm of computing where non volatile memory units are used pervasively to enhance reliability against power source instability, energy efficiency, and security. The non volatile computing devices are especially useful for embedded cyber physical systems enabling long running computations and data collection even with unreliable power sources. The technologies developed from this project can also benefit conventional architecture in its power optimization and internal security code generation. The project is a close collaboration between computer architecture and CMOS technology development groups, where all levels in the design hierarchy will be visited for system and technology evaluation.

This project integrates its research efforts with education by developing an undergraduate and Master curriculum that spans over the vertical design hierarchy in microprocessors. This vertical education will better prepare future work force in tackling tremendous design challenges spanning many layers of microprocessors. The results from this project will be made widely available to both industry and academia. CPS: Small: RUI: CPS Foundations in Computation and Communication This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5)

The objective of this research is considering security and timing as primary concerns, re envisioning computer architecture and network algorithms to provide a robust foundation for CPS. The approach is rethinking the hardware and software divide, providing true process concurrency and isolation. Extending these benefits to the communication network so integral to CPS, multicast and security innovations that consider CPS constraints will be proposed.

This project will provide computational and communication foundations for CPS through the following tasks. (1) An open source hardware design will be created. Abandoning the error prone paradigm of shared memory communication, Precision Timed (PRET) processors for dataflow computations will be extended. (2) The hardware/software interface will be investigated specifically for PRET architectures. (3) A routing algorithm considering the CPS constraints will be investigated. The constraints include efficiency, adaptability, scalability, simplicity, and security. (4) Distributed Source Coding for CPS applications will be studied with focus on challenges from small packet sizes in these applications.

This project will engage the community and students in multiple grades and institutions, through the following undertakings. (1) A package for education and research in CPS will be assembled. This package and the material from this project in the form of tutorials, publications, and curriculum will be available to other institutions. (2) New courses will be created integrating research results into education. (3) A diverse group of students including women and minorities will be recruited. (4) Two applications will be implemented in the fields of medical devices and emergency response. CPS: Medium: Collaborative Research: The Foundations of Implicit and Explicit Communication in Cyberphysical Systems Proposal Title: CPS:Medium:Collaborative Research: The Foundations of Implicit
and Explicit Communication in Cyberphysical Systems
Institution: University of California Berkeley
Abstract Date: 07/30/09
The objective of this research is to develop the theoretical
foundations for understanding implicit and explicit
communication within cyber physical systems. The approach is
two fold: (a) developing new information theoretic tools to
reveal the essential nature of implicit communication in a
manner analogous to (and compatible with) classical network
information theory; (b) viewing the wireless ecosystem itself
as a cyber physical system in which spectrum is the physical
substrate that is manipulated by heterogeneous interacting
cyber systems that must be certified to meet safety and
performance objectives.
The intellectual merit of this project comes from the
transformative technical approaches being developed. The key to
understanding implicit communication is a conceptual
breakthrough in attacking the unsolved 40 year old Witsenhausen
counterexample by using an approximate optimality paradigm
combined with new ideas from sphere packing and cognitive radio
channels. These techniques open up radically new mathematical
avenues to attack distributed control problems that have long
been considered fundamentally intractable. They guide the
development of nonlinear control strategies that are provably
orders of magnitude better than the best linear strategies. The
keys to understanding explicit communication in cyber physical
systems are new approaches to active learning, detection, and
estimation in distributed environments that combine worst case
and probabilistic elements.
Beyond the many diverse applications (the Internet, the smart
grid, intelligent transportation, etc.) of heterogeneous
cyber physical systems themselves, this research reaches out to
wireless policy: allowing the principled formulation of
government regulations for next generation networks. Graduate
students (including female ones) and postdoctoral scholars will
be trained and research results incorporated into both the
undergraduate and graduate curricula.
NATIONAL SCIENCE FOUNDATION
Proposal Abstract
Proposal:0932410 PI Name:Sahai, Anant
Printed from CPS:Medium: LoCal A Network Architecture for Localized Electrical Energy Reduction, Generation and Sharing This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).


The objective of this research is to understand how pervasive information changes energy production, distribution and use. The design of a more scalable and flexible electric infrastructure, encouraging efficient use, integrating local generation, and managing demand through awareness of energy availability and use over time, is investigated. The approach is to develop a cyber overlay on the energy distribution system in its physical manifestations: machine rooms, buildings, neighborhoods, isolated generation islands and regional grids. A scaled series of experimental energy networks will be constructed, to demonstrate monitoring, negotiation protocols, control algorithms and Intelligent Power Switches integrating information and energy flows in a datacenter, building, renewable energy: farm , and off grid village. These will be generalized and validated through larger scale simulations. The proposal?s intellectual merit is in understanding broadly how information enables energy efficiencies: through intelligent matching of loads to sources, via various levels of aggregation, and by managing how and when energy is delivered to demand, adapted in time and form to available supply. Bi directional information exchange is integrated everywhere that power is transferred. Broader impacts include training diverse students, such as undergraduates and underrepresented groups, in a new interdisciplinary curriculum in information and energy technologies. Societal impact is achieved by demonstrating dramatic reductions in the carbon footprint of energy and its overall usage, greater penetration of renewables while avoiding additional fossil fuel plants, and shaping a new culture of energy awareness and management. The evolution of Computer Science will be accelerated to meet the challenges of cyber physical information processing. CPS:Small:Collaborative Research:Localization and System Services for SpatioTemporal Actions in Cyber Physical Systems CPS: Small: Collaborative Research: Localization and System Services for SpatioTemporal Actions in Cyber Physical Systems

The objective of this research is to develop models, methods and tools for capturing and processing of events and actions in cyber physical systems (CPS) in a manner that does not violate the underlying physics or computational logic. The project approach uses a novel notion of cyber physical objects (CPO) to capture the mobility and localization of computation in cyber physical systems using recent advances in geolocation and the Internet infrastructure and supports novel methods for spatiotemporal resource discovery.

Project innovations include a model for computing spatiotemporal relationships among events of interests in the physical and logical parts of a CPS, and its use in a novel cyberspatial reference model. Using this model the project builds a framework for locating cyber physical application services and an operating environment for these services. The project plan includes an experimental platform to demonstrate capabilities for building new OS services for CPS applications including collaborative control applications drawn from the intermodal transportation system.

The project will enable design and analysis of societal scale applications such as the transportation and electrical power grid that also include a governance structure. It will directly contribute to educating an engineering talent pool by offering curricular training that range from degree programs in embedded systems to seminars and technology transfer opportunities coordinated through the CalIT2 institute at UCSD and the Institute for Sensing Systems (ISS) at OSU. The team will collaborate with the non profit Milwaukee Institute to explore policies and mechanisms for enterprise governance systems. CPS: Medium: Image Guided Robot Assisted Medical Interventions The goal of this project is to develop a novel cyber physical system (CPS) for performing multimodal image guided robot assisted minimally invasive surgeries (MIS). The approach is based on: (1) novel quantitative analysis of multi contrast data, (2) control that uses this information to maneuver conformable robotic manipulators, while adjusting on the fly scanning parameters to acquire additional information, and (3) human information/machine interfacing for comprehensive appreciation of the physical environment.

The intellectual merit arises from the development of: (1) a CPS that relies on real and real time data, minimizing parametric and abstracted assumptions, extracts and matures information from a dynamic physical system (patient and robot) by combining management of data collection (at the physical sensor site) and data analysis (at the cyber site), (2) smart sensing , to control data acquisition based on disruptive or situation altering events, (3) control coordination by interlacing sensing, control and perception, and the incorporation of steerable tools.

The societal impact arises from contributions to a leap in MIS: from keyhole visualization (i.e., laparoscopy) to in situ real time image guidance, thereby enabling a wider range of MIS. This will directly benefit patients and their families (faster recovery/reduced trauma). Economic impact arises from the cost effectiveness of MIS to the health care system, faster patient return to the workplace, and technology commercialization. The project will integrate research and education, diversity and outreach, by enhancing current and introducing new research intensive courses in Cyber physical Systems, Medical Imaging and Medical Robotics, and dissemination via trans institutional collaborations, a comprehensive web site, multimedia web seminars, and distribution to high schools. CPS: Small: A Real Time Cognitive Operating System The objective of this research is to develop a real time operating system for a virtual humanoid avatar that will model human behaviors such as visual tracking and other sensori motor tasks in natural environments. This approach has become possible to test because of the development of theoretical tools in inverse reinforcement learning (IRL) that allow the acquisition of reward functions from detailed measurements of human behavior, together with technical developments in virtual environments and behavioral monitoring that allow such measurements to be obtained.

The central idea is that complex behaviors can be decomposed into sub tasks that can be considered more or less independently. An embodied agent learns a policy for actions required by each sub task, given the state information from sensori motor measurements, in order to maximize total reward. The reward functions implied by human data can be computed and compared to those of an avatar model using the newly developed IRL technique, constituting an exacting test of the system.

The broadest impact of the project would provide a formal template for further investigations of human mental function. Modular RL models of human behavior would allow realistic humanoid avatars to be used in training for emergency situations, conversation, computer games, and classroom tutoring. Monitoring behavior in patients with diseases that exhibit unusual eye movements (e.g., Tourettes, Schizophrenia, ADHD) and unusual body movement patterns (e.g., Parkinsons), should lead to new diagnostic methods. In addition the regular use of the laboratory in undergraduate courses and outreach programs promotes diversity. CPS:Medium:Collaborative Research:Infrastructure and Technology Innovations for Medical Device Coordination The objective of this research is to develop a framework for the development
and deployment of next generation medical systems consisting of integrated and
cooperating medical devices. The approach is to design and implement an
open source medical device coordination framework and a model based component
oriented programming methodology for the device coordination, supported by a
formal framework for reasoning about device behaviors and clinical workflows.

The intellectual merit of the project lies in the formal foundations of the
framework that will enable rapid development, verification, and certification
of medical systems and their device components, as well as the clinical
scenarios they implement. The model based approach will supply evidence for
the regulatory approval process, while run time monitoring components embedded
into the system will enable black box recording capabilities for the forensic
analysis of system failures. The open source distribution of tools supporting
the framework will enhance its adoption and technology transfer.

A rigorous framework for integrating and coordinating multiple medical devices
will enhance the implementation of complicated clinical scenarios and reduce
medical errors in the cases that involve such scenarios. Furthermore, it will
speed up and simplify the process of regulatory approval for coordination enabled medical devices, while the formal reasoning framework will improve the confidence in the design process and in the approval decisions.

Overall, the framework will help reduce costs and improve the quality of the
health care. CPS:Small: A Unified Distributed Spatiotemporal Signal Processing Framework for Structural Health Monitoring CPS: CPS:Small: A Unified Distributed Spatiotemporal Signal Processing Framework for Structural Health Monitoring

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The objective of this research is to meet the urgent global need for improved safety and reduced maintenance costs of important infrastructures by developing a unified signal processing framework coupling spatiotemporal sensing data with physics based and data driven models. The approach is structured along the following thrusts: investigating the feasibility of statistical modeling of dynamic structures to address the spatiotemporal correlation of sensing data; developing efficient distributed damage detection and localization algorithms; investigating network enhancement through strategic sensor placement; addressing optimal sensor collaboration for recursive localized structural state estimation and prediction.

Intellectual merit: This innovative unified framework approach has the potential of being more reliable and efficient with better scalability compared to the current state of the art in structural health monitoring. The proposed research is also practical as it allows analysis of real world data that accounts for structural properties, environmental noise, and loss of integrity over sensors. Probabilistic representation of significant damages allows more informative risk assessment.

Broader impacts: The outcome of this project will provide an important step toward safety and reliability albeit increasing complexity in dynamic systems. New models and algorithms developed in this project are generic and can contribute in many other areas and applications that involve distributed recursive state estimation, distributed change detection and data fusion. This project will serve as an excellent educational platform to educate and train the next generation CPS researchers and engineers. Under represented groups such as female students and Native American students will be supported in this project, at both the graduate and undergraduate levels. CPS:Small:Cyber physical system challenges in man machine interfaces: context dependent control of smart artificial hands through enhanced touch perception and mechatronic reflexes CPS: Small: Cyber physical system challenges in man machine interfaces: context dependent control of smart artificial hands through enhanced touch perception and mechatronic reflexes

The objective of this research is to integrate user control with automated reflexes in the human machine interface. The approach, taking inspiration from biology, analyzes control switching issues in brain computer interfaces. A nonhuman primate will perform a manual task while movement and touch related brain signals are recorded. While a robotic hand replays the movements, electronic signals will be recorded from touch sensors on the robot?s fingers, then mapped to touch based brain signals, and used to give the subject tactile sensation via direct cortical stimulation. Context dependent transfers of authority between the subject and reflex like controls will be developed based on relationships between sensor signals and command signals.

Issues of mixed authority and context awareness have general applicability in human machine systems. This research advances methods for providing tactile feedback from a remote manipulator, dividing control appropriate to human and machine capabilities, and transferring authority in a smooth, context dependent manner. These principles are essential to any cyber physical system requiring robustness in the face of uncertainty, control delays, or limited information flow.

The resulting transformative methods of human machine communication and control will have applications for robotics (space, underwater, military, rescue, surgery, assistive, prosthetic), haptics, biomechanics, and neuroscience. Underrepresented undergraduates will be recruited from competitive university programs at Arizona State University and Mexico?s Tec de Monterrey University. Outreach projects will engage the public and underrepresented school aged children through interactive lab tours, instructional modules, and public lectures on robotics, human machine systems, and social and ethical implications of neuroprostheses. CPS: Small: Community based Sense & Respond Theory and Applications CPS: Small: Community based Sense & Respond Theory and Applications

The objective of this research is to address a fundamental question in cyber physical systems: What is the ideal structure of systems that detect critical events ? such as earthquakes ? by using data from large numbers of sensors held and managed by ordinary people in the community? The approach is to develop theory about widely distributed sense and respond systems, using dynamic ? and possibly unreliable ? networks using sensors and responders installed and managed by ordinary citizens, and to apply the theory to problems important to society, such as responding to earthquakes.

Intellectual Merit: This research develops theory and prototype implementations of community based sense and respond systems that enable people help one another in societal crises. The number of participants in such systems may change rapidly; some participants may be unreliable and some may even deliberately attack systems; and the structures of networks change as crises unfold. Such systems must function in rare critical situations; so designs, analyses and tests of these systems must give confidence that they will function when the crisis hits. The proposed research will show how to design systems with organic growth, unreliable components and connections, security against rogue components, and methods of demonstrating reliability.

Broader Impact: People want to help one another in a crisis. Cheap sensors, mobile phones, and laptops enable people to use information technology to help. This research empowers ordinary citizens collaborate to overcome crises. The researchers collaborate with the US Geological Service, Southern California Edison, and Microsoft, and will host 3,000 students at a seismic facility CPS:Medium: CitiSense Adaptive Services for Community Driven Behavioral and Environmental Monitoring to Induce Change This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The objective of this research project is to achieve fundamental advances in software technology that will enable building cyber physical systems to allow citizens to see the environmental and health impacts of their daily activities through a citizen driven body worn mobile phone based commodity sensing platform. The approach is to create aspect oriented extensions to a publish subscribe architecture, called Open Rich Services (ORS), to provide a highly extensible and adaptive infrastructure. As one example, ORS will enable highly adaptive power management that not only adapts to current device conditions, but also the nature of the data, the data?s application, and the presence and status of other sensors in the area. In this way, ORS will enable additional research advances in power management, algorithms, security and privacy during the project. A test bed called CitiSense will be built, enabling in the world user and system studies for evaluating the approach and providing a glimpse of a future enhanced by cyber physical systems.

The research in this proposal will lead to fundamental advances in modularity techniques for composable adaptive systems, adaptive power management, cryptographic methods for open systems, interaction design for the mobile context, and statistical inference under multiple sources of noise.

The scientific and engineering advances achieved through this proposal will advance our national capability to develop cyber physical systems operating under decentralized control and severe resource constraints. The students trained under this project will become part of a new generation of researchers and practitioners prepared to advance the state of cyber physical systems for the coming decades. CPS: Small: Real time, Simulation based Planning and Asynchronous Coordination for Cyber Physical Systems CPS:Small: Real time, Simulation based Planning and Asynchronous Coordination for Cyber Physical Systems

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The objective of this research is to investigate how to replace human decision making with computational intelligence at a scale not possible before and in applications such as manufacturing, transportation, power systems and bio sensors. The approach is to build upon recent contributions in algorithmic motion planning, sensor networks and other fields so as to identify general solutions for planning and coordination in networks of cyber physical systems.

The intellectual merit of the project lies in defining a planning framework, which integrates simulation to utilize its predictive capabilities, and focuses on safety issues in real time planning problems. The framework is extended to asynchronous coordination by utilizing distributed constraint optimization protocols and dealing with inconsistent state estimates among networked agents. Thus, the project addresses the frequent lack of well behaved mathematical models for complex systems, the challenges of dynamic and partially observable environments, and the difficulties in synchronizing and maintaining a unified, global world state estimate for multiple devices over a large scale network.

The broader impact involves the development and dissemination of new algorithms and open source software. Research outcomes will be integrated to teaching efforts and undergraduate students will be involved in research. Underrepresented groups will be encouraged to participate, along with students from the Davidson Academy of Nevada, a free public high school for gifted students. At a societal level, this project will contribute towards achieving flexible manufacturing floors, automating the transportation infrastructure, autonomously delivering drugs to patients and mitigating cascading failures of the power network. Collaboration with domain experts will assist in realizing this impact. CPS: Small: Random Matrix Recursions and Estimation and Control over Lossy Networks This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Many of the future applications of systems and control that will pertain to cyber physical systems are those related to problems of (possibly) distributed estimation and control of multiple agents (both sensors and actuators) over networks. Examples include areas such as distributed sensor networks, control of distributed autonomous agents, collision avoidance, distributed power systems, etc. Central to the study of such systems is the study of the behavior of random Lyapunov and Riccati recursions (the analogy is to traditional LTI systems where deterministic Lyapunov and Riccati recursions and equations play a prominent role). Unfortunately, to date, the tools for analyzing such systems are woefully lacking, ostensibly because the recursions are both nonlinear and random, and hence intractable if one wants to analyze them exactly. The methodology proposed in this work is to exploit tools from the theory of large random matrices to find the asymptotic eigendistribution of the matrices in the random Riccati recursions when the number of states in the system, n, is large. In many cases, the eigendistribution contains sufficient information about the overall behavior of the system. Stability can be inferred from the eigenanalysis. The mean of the eigenvalues is simply related to the mean of the trace (i.e., the mean square error of the system), whereas the support set of the eigendistribution says something about best and worst case performances of the system. Furthermore, a general philosophy of this approach is to identify and exhibit the universal behavior of the system, provided such a behavior does exist. Here, universal means behavior that does not depend on the microscopic details of the system (where losses occur, what the exact topology of the network or underlying distributions are), but rather on some simple macroscopic properties. A main idea of the approach is to replace a high dimensional matrix valued nonlinear and stochastic recursion by a scalar valued deterministic functional recursion (involving an appropriate transform of the eigendistribution), which is much more amenable to analysis and computation.

The project will include course development and the recruitment of women and minority students to research. It will also make use of undergraduate
and underrepresented minority student researchers through Caltech s SURF and MURF programs. CIF: SMALL: Explorations and Insights into Adaptive Networks, Animal Flocking Behavior, and Swarm Intelligence Abstract



Since the early 1990s, some useful optimization algorithms have emerged from the social and biological sciences in their studies of animal flock behavior and swarm intelligence. It has been observed, for example, that while individual agents in an animal colony are not capable of complex behavior, the combined coordination among multiple agents leads to the manifestation of regular patterns of behavior. Several algorithms have been developed to model the movement of animal flocks. These investigations are proving useful in modeling and understanding complex phenomena and in developing applications in areas ranging from biology to nanotechnology. The research involves investigating interconnections between these studies on swarm intelligence in the biological and social sciences, and more recent studies on adaptive networks in system theory.



Adaptive networks consist of isotropic nodes spread over a geographic domain. The nodes sense the environment and attempt to understand a phenomenon of interest based on their noisy observations and without any node taking a central control role. The nodes cooperate with each other through local interactions and adapt their states, and the network topology, in response to data collected at the nodes and data received from their neighbors. Information arriving at a node propagates throughout the network by means of a diffusive process. The diffusion of information results in a form of collective intelligence as is evidenced by improved learning and convergence behavior relative to non cooperative networks. A closer study of the dynamics of swarm behavior and adaptive networks can suggest alternative techniques for designing adaptive networks and for understanding flock behavior, with potential impact on the applications that can be motivated from these developments. AF:EAGER: Combinatorial Geometry, Partitioning, and Algorithms Divide and conquer and prune and search are two fundamental and ubiquitous paradigms in the design of algorithms. The first refers to the process of (i) splitting a given problem into smaller sub problems, (ii) solving each of these subproblems, and then (iii) combining these solutions to obtain the solution to the original problem. The second is a way of searching among possible solutions to a given problem whereby (a) the possible solutions are split into several groups, then (b) all but one of the groups is somehow eliminated, and finally (c) the search continues, now confined to the one remaining group. It is noteworthy that in both approaches, sets are ``split into smaller ones step (i) in divide and conquer and step (a) in prune and search and that in addition, many efficient and beautiful algorithms are based on one of these approaches. A main goal of this project is the development of some unusual, new, splitting tools that may be used in these paradigms. They will be sought from within an unexpected domain geometric partitioning theorems. Topological methods have been applied to obtain facts like the ham sandwich theorem, but there has not been much work on their algorithmic aspects, and what results we do have suggest that they would not be very useful as splitting tools for other algorithms. However some recent partitioning results of the investigator encourage the search for more tools of this kind.

Therefore this work will continue to seek new geometric partitioning results that can give novel, useful, splitting tools. Simultaneously the project will address a specific set of concrete, stubborn computational problems that arise frequently, and naturally, but have so far resisted efficient solutions. The goal is to better understand the complexity of these important and interesting problems, and to apply the new tools to obtain effective algorithms. Part of the intellectual merit of the project rests on the unusual approach to develop new algorithmic tools; in addition there is chance to make progress on a set of prevalent, hard, computational problems. Broader impacts reside in the potential to strengthen connections between geometry, combinatorics, and computation. TC: EAGER: Modularization Supporting Extensibility for an Industrial strength Theorem Prover The ACL2 theorem prover has an established user community in industry,
government, and academia. ACL2 supports industrial scale verification
projects by combining automation and controllability, but its
extensibility is limited: its large (10 MB), complex code base
requires, for soundness, that it be entrusted solely to its two
authors. The PIs propose to modify ACL2 radically, opening up the
system by making it more modular, thus enabling trusted development by
untrusted users while maintaining proof security. Key challenges are
to expose the functionality of system components, and to separate out
a trusted core from code that need not be trusted for correct
functionality, such as code implementing heuristics, I/O, theory
management, and interactive proof development and debugging. Code
refactoring is already a hard problem, especially for a system with
the complexity of ACL2, but in this project there is also the
challenge of making the resulting system modifiable in a way that does
not compromise logical soundness.

Expected results include an ACL2 system that can be modified soundly
by users according to specific needs. In particular, research on
teasing apart inherently sequential output from reasoning code should
support research on parallel reasoning algorithms taking advantage of
modern multi core machines, leading to formally verified parallel
implementations. More generally, the system will provide a platform
that promotes research in heuristics for automating reasoning. It
will also facilitate the customization of ACL2 for use in the
undergraduate classroom. The resulting system will be freely
distributed on the Internet. TC: A U.S. France Collaborative Symposium of Young Engineering Scientists (YESS 2009) In what has become an annual event, the French Government and The National Science Foundation jointly sponsor symposia, at the French Embassy, that bring together young investigators from both countries on a topic of current importance and that involves scientific challenges. For 2009, the topic was digital identity management, an important subtopic under the Trustworthy Computing Program?s emphasis area on ?fundamentals?. In bringing together these young scientists, it is anticipated that the field will be advanced by getting researchers together on problems that benefit from consideration of culture differences between the two countries (as is the case with the study of privacy issues),

A very broad view of identity management is being considered, including biometrics and other methods for authentication, forensics to determine if after the fact an authentication violation took place, cryptographic and other methods to achieve anonymity, and metrics to determine the effectiveness of methods individually and in combination.

The PI created a program that consists of invited talks, talks by young investigators, talks by Government research agencies in which important research areas for both countries are declared, and panels.

The funds requested would be used to reimburse the expenses of graduate and undergraduate students. The workshop is to take place on July 7 10, 2009 in Washington, DC.

Our systems are experiencing an increasing number of attacks from all kinds of attackers: underground criminals, nation states, terrorists, etc. Knowing who is on your system is vital to discouraging attacks but also to achieve accountability in the face of an attack. III/EAGER: Towards Workflows as First Class Citizens in Cyberinfrastructure: Designing Shared Repositories Scientific computing has entered a new era of scale and sharing with the arrival of cyberinfrastructure for computational experimentation. A key emerging concept is scientific workflows, which provide a declarative representation of scientific applications as complex compositions of software components and the dataflow among them. Workflow systems manage their execution in distributed resources, track provenance of analysis products, and enable rapid reproducibility of results. In current cyberinfrastructure, there are well understood mechanisms for sharing data, instruments, and computing resources. This is not the case for sharing workflows, though there is an emerging movement for sharing analysis processes in the scientific community.

This project explores computational mechanisms for sharing workflows as a key missing element of cyberinfrastructure for scientific research, with three major research foci: (1) Elicitation of new requirements that workflow sharing poses over current techniques to share software tools and libraries; (2) Understanding how shared workflow catalogs should be designed, as the existing data catalogs are a successful model, and software components require different representations and access functions; and (3) Studying what sharing paradigms might be appropriate for scientific communities.

Expected results from this work include: use cases for workflow sharing and reuse that motivate this research area, a comparison between software reuse and workflow reuse requirements, a specification of a workflow catalog defining expected functions and services, and an investigation of social issues that arise in building this new kind of shared resource in scientific communities. Results are available at the project Web site (http://workflow sharing.isi.edu). III/EAGER: TwitterStand: Separating the Wheat from the Chaff in Breaking News Twitter is an electronic medium that allows a large user populace to communicate with each other simultaneously. Inherent to Twitter is an asymmetrical relationship between friends and followers thereby provides an interesting social network like structure among the users of Twitter. Twitter messages, called tweets, are restricted to 140 characters and thus are usually very focused. Twitter is becoming the medium of choice for keeping abreast of rapidly breaking news. This project explores the use of Twitter to build a news processing system from Twitter tweets. The result is analogous to a distributed news wire service. The difference is that the identities of the contributors/reporters are not known in advance and there may be many of them. The tweets are not sent according to a schedule. The tweets occur as news is happening and are noisy while usually arriving at a high throughput rate.

The goal of this exploratory research project is to find effective methods for making Twitter a useful news gathering mechanism. Challenges addressed in this project include: removing the noise; determining tweet clusters of interest bearing in mind that the methods must be online; and determining the relevant location associated with the tweets.

The broad impact of this research is to make it easier to disseminate late breaking news and enhancing the distributed news gathering and reporting process. Web site (http://www.cs.umd.edu/~hjs/hjscat.html) reports results of this and related research. III/EAGER: Temporal Relationships Among Clusters in Data Streams (TRACDS) State of the art data stream clustering algorithms developed by the data mining community do not utilize the temporal order of events and therefore in the resulting clustering all temporal information is lost. This is quite strange as one of the salient features of data streams is temporal ordering of events. In this project we develop a technique to efficiently incorporate temporal ordering into the clustering process and prove its usefulness on large, high throughput data streams. Temporal ordering is introduced into the data stream clustering process by dynamically constructing an evolving Markov Chain where the states represent clusters. Our approach is based on the previously developed Extensible Markov Model (EMM). The results of this project will provide a framework upon which important stream mining applications such as anomaly detection and prediction of future events are easily implemented.

By showing that state of the art data steam clustering algorithms can incorporate temporal order information efficiently, this project will have a broad impact on many areas where temporal order is essential. As examples, NOAA Hurricane Data and NASA satellite data will be used throughout this project. Results, including open source software will be distributed via the project Web site (http://lyle.smu.edu/ida/tracds). NeTS: Efficient and Localized Broadcasting in Ad Hoc Wireless Networks NeTS NR: Efficient and Localized Broadcasting in Ad Hoc Wireless Networks

Jie Wu, Florida Atlantic University

Award 0434533

Abstract

Collective communication represents a set of important communication functions that involve multiple senders and receivers. Broadcasting is one of the fundamental operations and has extensive applications, including the route discovery process in reactive routing protocols, naming and addressing, and dense mode multicasting. Due to the broadcast nature of wireless communication, blind flooding of the broadcast message may cause serious contention and collision, resulting in the broadcast storm problem. This project studies the challenge of efficient and localized broadcasting in ad hoc wireless networks by offering a generic framework that can capture many existing localized broadcast algorithms and, in addition, some efficient solutions can be derived from this framework. This research has six thrusts: (1) Provide a more generic framework for deterministic and localized broadcasting in ad hoc networks, including constructing consistent views. (2) Derive cost effective broadcast schemes from the framework. (3) Reduce excessive broadcast redundancy through energy efficient design,. (4) Explore the use of broadcasting as a basic building block to support other types of collective communication. (5) Ensure broadcast coverage with controlled redundant transmission without solely relying on ACK/NACK. (6) Integrate different components and fine tune the system through an empirical study based on a set of well defined quantitative performance metrics. The new framework can easily integrate other objectives such as energy efficiency and reliability. The results of this research will provide guidelines for efficient and localized algorithms for a wide range of applications. This research will also exploit and contribute to theoretical studies in graph theory and distributed algorithms. RI: EAGER: Robust Opportunistic Fitting of Partial Body Models This project addresses the problem of fitting an articulated body model to a person in an image. The task is challenging due to large variation in appearance caused by body pose, clothing, illumination, viewpoint and background clutter. Unlike current methods that try to fit a full body model to every image, this approach uses opportunistic search within a space of partial body models to find only those body parts that are currently visible and detected with high confidence. Not trying to fit occluded or poorly visible parts reduces the chances of making a mistake, so subsequent processes can rely on receiving a high quality partial model solution. A stochastic search technique employing high level subroutines to propose candidate body configurations searches for the globally optimal solution in terms of number and configuration of visible body parts, removing the need for a close initial estimate and allowing more thorough exploration of the solution space. The proposed partial body configurations also provide top down guidance for image segmentation of individual body parts, yielding better delineation of body shape than simple parameterized models or bottom up segmentation. An implementation of the approach is being compared against existing work using publicly available datasets. Robust segmentation of torso and limbs from still images provides a natural representation of the human body that can have broad impact on tasks such as human activity recognition and markerless body tracking within interactive smart spaces. TC: EAGER: Measuring Architectures for Resilient Security (MARS) A system has resilient security if it retains a degree of secure functioning despite the compromise of some components. Since vulnerable components will long be in widespread use, resilient security is what counts against sophisticated adversaries with persistent footholds in American systems.

Resiliency infrastructures can help secure application components that may have many intrinsic weaknesses. They can structure systems so the risk of successful attack can be meaningfully measured.

Resiliency is more achievable than previously, because of recent architectural changes. One is virtualization , allowing many virtual machines to execute on a physical platform. Some virtual machines may serve as resiliency infrastructure nodes, controlling adjacent application nodes. Second, software attestation and appraisal, supported by Trusted Platform Modules and secure virtualization, allow a component to appraise the software state of remote peers.

We add three architectural ideas. Emulsification means breaking application functionality into small pieces, implemented as separate virtual machines. Second, their interactions can be monitored and secured by infrastructure nodes. Monitoring includes auditing, filtering , and modifying messages among application components. Third, data provenance uses annotations prepared by infrastructure nodes and stored with data objects.

Game theory applies to attacks that must succeed against several components, spread between the infrastructure level and the application level. Networks with randomized components force the adversary to use probabilistic strategies with low probability of defeating all of a sequence of components.

Broader impacts: Our society depends on information systems riddled with vulnerabilities. New architectures will reduce the severity of this problem , and provide measurements of risk. SHF: EAGER: Closing the gap in Controller Synthesis Automatic controller synthesis algorithms hold the promise of producing correct by construction systems, obviating the need for costly post facto verification. However, there is currently a gap between theoretical foundations of controller synthesis and their practical implementations on hardware and software platforms. This project addresses challenges in closing the gap in control synthesis. In particular, we consider two fundamental problems. First, we consider the problem of implementation complexity of controllers. While theoretical results have focused on the optimal memory requirements for controllers, in practice, a controller implementation may have several other optimality requirements such as the size of the implementing (combinational and sequential) circuit, the complexity and frequency of computing the control action, and the sensing and actuation bandwidth. Accordingly, we study algorithms for the construction of optimal controllers under these metrics. Second, we consider the problem of fault tolerance in controllers, in which we consider effects of (possibly stochastic) errors in controller implementations. While traditional fault tolerance techniques such as error correcting codes and redundancy can be applied directly, our thesis is that a closer interaction of fault tolerance with controller synthesis algorithms can lead to fault tolerant designs at costs lower than traditional techniques. For example, by distinguishing the importance of signals to the control objective, the costs associated with error correction and redundancy can be decreased while having a negligible effect on the control objective. The research performed in this project is prerequisite for a more widespread adoption of correct by construction techniques.

The tools and techniques developed in this project have the potential to significantly enhance our ability to produce robust cyber physical systems, thus affecting several large scale application areas beyond the computer science and control engineering domains. Practically, the results of the research will lead to better controller synthesis tools. Theoretically, the research will bring together cross cutting techniques, ranging from theoretical foundations in logics and algorithms for control, to optimization techniques, real time systems, and hardware and software synthesis. In addition, by fostering collaboration between software foundations, control foundations, and hardware synthesis foundations, the project will train graduate and undergraduate students in the emerging and important domain of formal techniques for cyber physical systems. MRI R2: Acquisition: A Hybrid High Performance GPU/CPU System This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
Proposal #: 09 58854
PI(s): Wu, Jie, Biswas, Saroj K., Klein, Michael L., Rivin, Igor, Shi, Yuan
Institution: Temple University
Title: MRI R2: Acquisition: A Hybrid High Performance GPU/CPU System
Project Proposed:
This project, acquiring a hybrid high performance GPU (graphics processing unit)/CPU system, enables broader heterogeneous computing by deploying multiple types of computing nodes and allowing each to perform the tasks to which it is best suited in traditional CPU based, GPU based, and hybrid GPU/CPU applications. Satisfying two major research environments in computer and information sciences and scientific computing, the instrument
Serves faculty and students conducting research using existing parallel application software and developing tools for parallel programming and parallel applications of the system,
Serves the center for High Performance Computing and Networking at the institution and the greater Philadelphia region offering services not only to other departments on campus, but also to other institutions in the local community, such as area high schools and local colleges/universities ,
Educates and trains providing a computing environment to various science courses for students to gain hands on experience and offers research training sessions to the local research community; and
Fosters collaboration by supporting joint research with local colleges/universities, in collaboration with other schools in the state, to develop, test, and apply advanced tools for designing and executing parallel programs.
Among others, the Center services Chemistry, Computer and Information Sciences, Electrical and Computer Engineering, Mathematics, Physics, and Pulmonary, Critical Care Medicine and Physical Therapy. The instrument specifically supports research projects with broad impact in molecular self assembly, microvessel networks, spatio temporal data analysis, large scale system simulation, effective uniformization, fault tolerant computing, etc. and augments existing applications using the GPU as an accelerator enabling some problems to run entirely on the GPU.
Broader Impacts:
The instrument greatly enhances the current computing facilities at the institution. With its GPU/CPU component, this instrument is the first of its kind of high performance computing facility in the greater Philadelphia region, an area with a high degree of diversity. This university draws a substantial portion of its students from one of the largest African American populations in the country. Enabling education, training, and collaboration, the instrument contributes to foster economic growth in an area with a high concentration of high tech and IT related industries that currently has no high performance computing and networking center of this magnitude. The region is now able to go beyond minimal services and promote and support collaboration and cooperation across sectors (e.g., higher education and local industry). MRI R2 Consortium: Development of Dynamic Network System (DYNES) This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

This project will develop and deploy the Dynamic Network System (DYNES), a nationwide cyber instrument spanning 39 US universities and 16 regional networks. DYNES will support large, long distance scientific data flows in the LHC, other leading programs in data intensive science (such as LIGO, Virtual Observatory, and other large scale sky surveys), and the broader scientific community.

By integrating existing and emerging protocols and software for dynamic circuit provisioning and scheduling, in depth end to end network path and end system monitoring, and higher level services for management on a national scale, DYNES will allocate and schedule channels with bandwidth guarantees to several classes of prioritized data flows with known bandwidth requirements, and to the largest high priority data flows, enabling scientists to utilize and share network resources effectively. DYNES is dimensioned to support many data transfers which require aggregate network throughputs between sites of 1 20 Gbps, rising to the 40 100 Gbps range. This capacity will enhance researchers? ability to distribute, process, access, and collaboratively analyze 1 to 100 TB datasets at university based Tier2 and Tier3 centers now, and PB scale datasets once the LHC begins operation.

DYNES is based on a ?hybrid? packet and circuit architecture composed of Internet2 s ION service and extensions over regional and state networks to US campuses. It will connect with transoceanic (IRNC, USLHCNet), European (GÉANT), Asian (SINET3) and Latin American (RNP and ANSP) Research and Education networks. It will build on existing key open source software components that have already been individually field tested and hardened: DCN Software Suite (OSCARS / DRAGON), perfSONAR, UltraLight Linux kernel, FDT, FDT/dCache, FDT/Hadoop, and PLaNeTs.

The DYNES team will partner with the LHC and astrophysics communities, OSG, and Worldwide LHC Computing Grid (WLCG) to deliver these capabilities to the LHC experiment as well as others such as LIGO, VO and eVLBI programs, broadening existing Grid computing systems by promoting the network to a reliable, high performance, actively managed component.

Future science programs in HEP, astrophysics and gravity wave physics, and other data intensive disciplines, will be facilitated by DYNES? technologies and worldwide network partnerships. Working with CHEPREO and similar education and outreach efforts targeting under served communities both in the US and overseas, DYNES will reach a wide variety of students at collaborating institutes including underrepresented groups and minorities. This will lower the barriers, and enable individual graduate students, undergrads, postdocs and faculty to use DYNES to achieve high throughput in support of their research in many data intensive fields. MRI R2: Development of the Next Generation CAVE Virtual Environment (NG CAVE) This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
Proposal #: 09 59053
PI(s): Johnson, Andrew E.; Brown, Maxine, Leigh, Jason, Peterka, Tom
Institution: University of Illinois Chicago
Title: MRI/Dev.: Dev. of the Next Generation CAVE Virtual Environment (NG CAVE)
Project Proposed:
This project, developing the Next Generation CAVE (NG CAVE), supports 15 research projects from local institutions. These projects in multiple domains (Astronomy, Astrophysics, Art, Bioengineering, Earth Science, High Performance Computing, Homeland Security, Neuroscience, Rehabilitation, etc.) are poised to use NG CAVE for their large visualization needs. Just as cyberinfrastructure provides better access to greater volumes and varieties of data, from data storage systems, online instrumentation, and/or major computational resources like the TeraGrid and future Petascale facility, advanced visualization instruments serve as the eyepieces of a telescope or microscope, enabling researchers to view their data in cyberspaces and to better manage the increased scale and complexity of accessing and analyzing the data. NG CAVE is such an eyepiece, providing researchers with powerful and easy to use information rich instrumentation in support of cyberinfrastructure enabled scientific discovery. It provides users with the ability to see 3D content at nearly 106 Megapixels.
For the Electronic Visualization Laboratory (EVL) at the institution, NG CAVE represents the culmination of decades of experience and expertise developing immersive environments, from the room sized CAVE virtual environment in 1992, to the office sized ImmersaDesk in 1994, to the GeoWall in 2000, and the more recent ultra high resolution LamdaVision tiled display wall and autostereoscopic Varrier tiled display wall. Each new generation of visualization instrumentation has provided scientific communities with one or more advanced features (higher resolution, unencumbered stereoscopic viewing, multi Gigabit connectivity, and intuitive user interfaces), better coupling worldwide scientific virtual organizations, and better integrating scientific workplaces with globally distributed cyberinfrastructure.
NG CAVE provides an alternative approach to constructing CAVEs by using new near seamless flat LCD technology augmented with micropolarization, rather than traditional projection technologies. The net effect is a new CAVE that has 3D acuity to match human vision, can be scaled near seamless to even greater resolution, is affordable compared to projection based approaches, requires little maintenance, can be used for both 2D and 3D stereoscopic viewing, and can support multiple simultaneous viewers. The instrument also opens new opportunities in computer science research at the intersection of large scale data visualization, human computer interaction, virtual reality, and high speed networking,
Broader Impacts:
This project provides state of the art equipment, opportunities, and supervision to enhance undergraduate and graduate research and education. The new NG CAVE supports 10 classes in Computer Science, Art and Design, and Biomedical Science departments. It provides scientific communities with highly integrated virtual reality collaboration environments; it enables working with industry to commercialize new technologies for the advancement of science and engineering and to continue ongoing partnerships with many of the world s best domain scientists and computer scientists in academia and industry, who readily become early adopters of new instrumentation and who provide students with summer internships and jobs upon graduation. Thus, this instrument enables US to maintain its leadership position in high performance computing and contributes in the advancement of complex global issues (e.g., environment, health, homeland security, economy, etc.), which, in turn, benefit society as a whole. MRI R2: Acquisition of Data Analysis and Visualization Cyber Infrastructure for Computational Science and Engineering Applications(DAVinCI) This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

We will acquire and operate a new facility, Data Analysis and Visualization Cyber Infrastructure for Computational Science and Engineering Applications (DAVinCI), that will use hybrid a design that integrates high throughput serial and tightly coupled parallel computing as well as General Purpose Computing on Graphics Processing Units (better known as GPGPUs). DAVinCI is specifically designed for large computations requiring fast I/O on large datasets and will integrate storage and visualization to support a wide array of data intensive science and engineering applications. In addition, we will establish a center for 3D stereo visualization with active tracking for analysis of large simulations and datasets in order to fulfill researchers? critical need to rapidly view and analyze the results of large computations.

DAVinCI will be used by researchers tackling a broad range of science and engineering problems including earth, environmental science and energy research, natural hazards and physical infrastructure research, bioscience and bioengineering research, and physics, space physics and astronomy research.

DAVinCI will also serve to interface high performance computing at Rice with national resources, such as TeraGrid. The machine and visualization center will be available to the entire Rice community and their local and national collaborators. It will therefore promote quantitative research in the social sciences and other divisions of Rice and beyond.

DAVinCI will enhance the research training of hundreds of undergraduate and graduate students and post doctoral fellows in science and engineering. It will directly impact the educational experience for all students as high performance computing and computational problem solving are included by Rice faculty in courses at all levels.

DAVinCI will be a primary research tool for some undergraduate research projects in Rice?s Century Scholars program, and also a critical tool used by the many undergraduates working on NSF funded research. Furthermore, a diverse group of graduate students as well as many undergraduate students from the NSF sponsored Rice Alliance for Graduate Education and the Professoriate (AGEP) program, established to promote the advancement of under represented minorities in the STEM fields, will benefit from use of these new computational resources in their training. A special summer program to familiarize students with high performance computing and to teach computational problem solving skills will be organized. The availability of this powerful facility to new faculty will further support the Rice NSF ADVANCE program?s efforts to recruit top women faculty to Rice in science and engineering. MRI R2: Acquisition of an Integrated Instrument for Computational Research and Education This proposal will be awarded using funds made available by the American Recovery and Reinvestment Act of 2009 (Public Law 111 5), and meets the requirements established in Section 2 of the White House Memorandum entitled, Ensuring Responsible Spending of Recovery Act Funds, dated March 20, 2009. As the cognizant Program Officer, I also affirm that the proposal does not support projects described in Section 1604 of Division A of the Recovery Act.

Proposal #: 09 59124
PI(s): Apon, Amy W.; Cothren, Jackson, D.; El Shenawee, Magda O.; Pulay, Peter; Spearot, Douglas E.
Institution: University of Arkansas
Title: MRI R2/Acq.: Acq. of an Integrated Instrument for Computational Research and Education
Project Proposed:
This project, acquiring an integrated supercomputing platform for distributed and shared memory parallel applications that are computationally, data, and storage intensive, aims to service a broad range of projects in computational science and engineering. The work potentially expands the capability to do computational research in Arkansas by an order of magnitude.
The supercomputer requested would be used principally in the areas of computer science, computational chemistry, computational biomagnetics, mechanical engineering, geosciences, and spatial science. It also benefits researchers in integrated nanoscience, astrophysics, bioinformatics, large scale optimization, video analysis, and network topology.
Broader Impacts:
The proposed cluster will be the primary high end computing platform for research and educational activities for the University and for the state. The geosciences projects have substantial commercial potential and particular significance for the economic development in Arkansas. The proposal includes a significant outreach and an educational component targeted to undergraduate students. In addition, the project contains training activities targeted to faculty who teach undergraduate students at many institutions of the eleven four year institutions of higher education in Arkansas. III:EAGER:Collaborative Research:A Collaborative Scientific Workflow Composition Tool Supporting Scientific Collaboration The goal of this research aims to produce a general purpose but domain customizable collaborative scientific workflow tool for accelerating scientific discovery, particularly facilitating large scale and cross disciplinary research projects that are collaborative in nature and require intensive user interaction from multiple distributed domain scientists. As a natural extension to the existing single user oriented scientific workflow management tools by providing direct system support for scientific collaboration, this project seeks to pave a way toward a next generation tool supporting scientific collaboration over the Internet. The expected tool can be easily expanded to support data centric collaborative information management in any intelligence community.

To achieve this broader impact, this research seeks to establish a set of foundational models and techniques supporting collaborative scientific workflow composition and management; and based on them, construct an Internet based collaborative scientific workflow tool framed with an open source initiative. The initial focus of the Early Concept Grants Exploratory Research(EAGER) project is on rapidly creating a prototype tool as a proof of concept, equipped with basic dataflow oriented scientific workflow models and collaboration patterns integrated in an agile service oriented architecture. To avoid reinventing the wheel, the efforts will be concentrated on transforming and extending Taverna, a known open source scientific workflow management tool, into a collaborative version. Evaluations and validations will be conducted through partnerships with multiple collaborative scientific communities.

For further information, see the project website at http://www.CollaborativeScientificWorkflows.org/ MRI R2 : Acquisition of a Heterogenous Terascale Shared Campus Computing Facility This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
Proposal #: 09 59382
PI(s): Fisher, Robert T.; Cowles, Geoffrey; W., Gottlieb; Sigal, Khanna, Gaurav; Wang, Cheng
Institution: University of Massachusetts Dartmouth
North Dartmouth, MA 02727 2300
Title: MRI R2: Acquisition of a Heterogeneous Terascale Shared Campus Computing Facility
Project Proposed :
This project, acquiring a heterogeneous terascale parallel computer cluster incorporating graphics processing units (GPUs), services an inter and multi disciplinary group of mathematicians and computational scientists and engineers, and their undergraduate and graduate students. This shared campus research instrument leverages an existing small scale cluster without GPUs at UMass Dartmouth resulting in a combined system of more than four times the size.
Specifically, the proposed cluster contains 32 8 way nodes (256 cores) plus one NVIDIA Tesla GPU per node for a total of 32 GPUs. The cluster is not intended to be a general computing resource, but rather highly specialized instrument necessary for the development of high performance algorithms for scientific computation by the group of researchers in the proposal. Researchers will take advantage of the incorporation of the GPUs in each node to develop hybrid parallel and many core architecture based computing. The investigators will abstract out these GPU accelerated solvers into an open source GPU Scientific Library (GPUSL), which will be cleanly separated from the overlying simulation codes through a set of well defined interfaces.
Faculty members from the Physics, Mathematics, Marine Science and Civil Engineering departments at UMass Dartmouth have formed a scientific computing group that will become the core management and user group of the shared computational resources. The research foci of the group range from Evolution of Giant Molecular Clouds (using the FLASH code) to Coastal Ocean Modeling (using the FVCOM code) to the Simulation of an Incompressible Turbulent Convective Fluid. Activities of this group include research collaborations, weekly team meetings for the sharing of expertise on numerical algorithms, co advising of students, and a development of a multi disciplinary doctoral program.
The cluster will be housed in the Campus Information Technology Services (CITS) data center at UMass Dartmouth, and directly supported by the CITS Enterprise Systems Administration Team. This team will be responsible for preventive maintenance session backup and recovery of user disk space. The access to the machine and its allocation will be handled by a Research User Group Committee which will be responsible for soliciting proposals from faculty members seeking computer time and will also be responsible for balancing the research needs of the faculty, graduate, and undergraduate students in the user base.
Broader Impacts:
The availability of this cluster allows for the training of the next generation of high performance computing experts, and supports a new interdisciplinary Ph.D. program in computational science, currently in development. Graduate students benefit from the opportunity to develop cutting edge parallel and GPU accelerated algorithms for a variety of challenging physical problems, in a multi disciplinary setting. The cluster also directly benefits undergraduate education, as part of the NSF funded Computational Science for Undergraduate Mathematics Students (CSUMS) project, which aims to prepare undergraduate students for, and engage them in, a sustained research experience in computational mathematics. The approach is to recruit promising high school seniors, then guide freshman and sophomores students through research topics in computational science, and culminate by mentoring seniors through carefully selected research projects.
In addition, UMass Dartmouth has a successful tradition of broadening higher education access to underrepresented students through its College Now Steps Toward Abstract Reasoning and Thinking (START) program. Over its 40 year history, the program has played a significant role in broadening STEM access to disadvantaged students to women and underrepresented minorities; in particular, approximately 25% 30% of START students in recent semesters are underrepresented minority persons of color. For the scientific community, all computer codes developed on this enabling cluster, including the GPU Software Library, will be released via a dedicated website, immediately impacting more than 2500 researchers worldwide who use the community codes FLASH and FVCOM. MRI R2: Development of a User Centered Network Measurement Platform This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
Proposal #: 09 59441
PI(s): Wills, Craig
Claypool, Mark L.; Doyle, James K.; Ward, Mathew O.
Institution: Worcester Polytechnic Institute
Title: MRI/Acq.: Dev. of a User Centered Network Measurement Platform
Project Proposed:
This project, developing a network measurement approach centered on where users live and on their specific interactions with the Internet, supports a wide range of research projects. The instrumentation provides a means to test applicable techniques from research projects on a large scale through vetted experiments and will be accessible to all users via two integrated measurement approaches that
Execute within the protected sandbox of a user Web browser to provide low impediments encouraging broad user participation while still yielding useful Internet data.
Perform more extensive application oriented performance tests from a user s machine and is not constrained by the sandbox.
The former makes use of entertainment applications in the form of games and videos to help engage users and increase participation. The application oriented nature of the latter tests yields performance data that can be uploaded for sharing among network researchers while providing feedback to users on network applications of interest. This feedback serves as an incentive for user participation. Thus, the platform provides information about the network applications that interest home users and conditions in which these applications operate. Researchers can use the platform tests from their own measurement points (MPs) and then calibrate the characteristics of their MPs with others. The entertainment orientation of the sandboxed platform serves provides an environment to decide how best to employ the use of videos and games for measurement. The open research area offers a training opportunity for students.
Broader Impacts:
This project promotes teaching and training through integration with systems courses and the Interactive Media and Games Development (IMGD) program. The work also aims to broaden participation in science and engineering by underrepresented groups through IMGD. Moreover, the available data sets are likely to provide a foundation for scientific discovery. MRI R2: Acquisition of a GPU cluster for solving n body systems in science and engineering This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Graphics Processors (GPUs) are potentially a cost effective and low power vehicle for science and engineering research that requires high performance computation. The primary challenge to the use of GPUs more broadly is the difficulty in programming. Dr. Walker and a team of colleagues representing five different scientific and engineering disciplines propose to pursue research topics in each of the disciplines. By selecting important research topics which require a fundamentally similar computational algorithm for a class of problems labelled n body problems , the project offers opportunity for meaningful interdisciplinary collaboration across scientific domains that are normally quite distinct. Since, solutions to this class of problem are particularly well suited to GPUs, there is likelihood of advances in multiple areas of scientific interest at a fraction of alternative costs and power. Therefore NSF s Office of Cyberinfrastructure (OCI) is supporting the acquisition of the instrument. MRI R2 Consortium: Acquisition of hardware for data visualization and exploratory analysis This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Visualization is increasingly essential for some forms of scientific and engineering research, particularly to explore change over time, or to examine complex interactions in models. While most scientific visualization occurs in two dimensions, selective use of three dimensional immersive technology for modeling and simulation projects is proposed by Drs. Joiner, Chang and Morreale at Kean University in collaboration with Dr. Manson of the Richard Stockton College of New Jersey. The specific research projects include topics in biochemistry, environmental studies, biology, meteorology, applied mathematics, psychology and computer science. There is a strong STEM education component to the proposed activities since Kean and Stockton are first and foremost, teaching universities. Nearly half of Kean s undergraduate population is minority and both schools have a significant percent of their student population who are first generation college students. The award from NSF s Office of Cyberinfrastructure supports acquisition of advanced scientific visualization tools, with Kean receiving the most significant grant. MRI R2: Development of a Second Generation Applications Driven Wireless Sensor Networking Instrument This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
Proposal #: 09 59584
PI(s): Noubir, Guevara
Basagni, Stefano; Desnoyers, Peter; Vona, Marsette A.
Institution: Northeastern University
Title: MRI R2/Dev.: Second Generation Applications Driven Wireless Sensor Networking Instrument
Project Proposed:
This project, developing a multi purpose wireless sensor networking instrument, supports the specific experimental research needs of cross layer protocols for heterogeneous sensor networks and key applications such as search and rescue by swarms of robots, buildings structure health monitoring, and hand and patient motion tracking. The instrument enables
Research and education for developing and experimenting with protocols and algorithms for a future generation of wireless sensor networks,
Cross cutting research and education in application areas of key interest to the society and the institution.
This work enhances the support of existing mechanisms of today s wireless sensor networks such as security, energy efficiency, and reliability by using a more capable second systems on chips at an order of magnitude lower cost and within a significantly smaller package than today s solution. Additionally, inexistent capabilities that enable research are designed and integrated in the target multi disciplinary application areas; these include:
Directional antennas for localization, and interference cancellation, using a combination of low cost mechanical and electronic beam forming techniques (outperforming purely electronic smart antennas). This improves communication capacity and robustness against unintentional and malicious interference. It will be fitted on mobile robots and combined with ultra sound transceivers for faster localization of transmission sources in search and rescue missions.
Ultralow power with multi radio support including wakeup radios, enabling asymmetric communication architectures, and allowing deployed sensor nodes to last for over a decade without battery changes.
Nodes hardware, software, and network design architected for ease of composability to quickly integrate specific hardware components of new applications such as wideband reduces personnel in the development microelectronic mechanical systems (MEMS) ultrasound transceivers, MEMS accelerometers, flash storage, and also interfaces with a variety of robots and off the shelf components (e.g., miniSD GPS).
Broader Impacts:
This project fosters a wider use of wireless sensor networks in application areas of national importance, namely, emergency preparedness and health science. All the instrument components are open sourced including the schematics, printed circuit board (PCB) layouts, and software with adequate documentation, thus enabling other institutions to easily extend and build copies of the instrument at low cost. The work fosters the development of the next generation application driven wireless sensor networks widening the use of such networks both in research and education. Several education kits and evaluation systems will be made available to other academic groups. Dissemination workshops will also be organized for application and module developers. The instrument will be used in several multi disciplinary courses bridging CS, EE, ME, and CE. MRI R2: Acquisition for the Critical Infrastructure Security and Assessment Laboratory (CISAL) This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

A strong need for research activities on critical infrastructure protection motivated the establishment of a research facility at Jacksonville State University (JSU) whose primary purpose is to investigate, develop, implement, and assess security technologies for critical infrastructure protection. This cross disciplinary project undertaken by the Mathematical, Computer and Information Sciences (MCIS) and the Engineering and Technology departments represents a novel approach of establishing a cybersecurity focused research laboratory facility that emulates multiple configurations of real industrial control systems.

The project builds upon the existing STEM knowledge base in the areas of network security, risk assessment, security compliance, industrial controls, and digital forensic analysis to enhance existing and develop new security technologies in an area of national need. The lessons learned and the experiences gained in designing and utilizing the facility will provide a blueprint for the design of similar facilities in medium to large sized institutions of higher education. A key component of this project is the collaborative effort that will take place between JSU, area community colleges, and industrial partners in maximizing the utilization of the proposed research facility through active collaborations.

The creation of new and the adoption of existing STEM knowledge nuggets, including the creative research techniques in security, networks, and industrial controls, will provide exemplary materials to institutions with diverse demographics. JSU, having been recently designated by the National
Security Agency (NSA) as a National Center of Academic Excellence in Information Assurance Education, is in a unique position of being able to contribute to broadening the participation of the underrepresented groups through the diversity of its student population. In addition, this project will target smaller institutions of learning, including institutions that traditionally serve underrepresented groups, as future collaborators.

The collaboration between JSU, area community colleges and industrial partners can serve as a model for future collaborations between institutions of higher learning and the industrial sector. The results of our research and project evaluation will be disseminated using a dedicated Web server ? a cost effective approach to closing the gap between those who have access to rich learning opportunities and those who do not. We will initiate a pilot workshop for community college instructors and industry personnel on the utilization and capability of the research facility. Since the reach of this project can extend beyond STEM disciplines, the project will also provide orientation workshops to faculty members of other disciplines to stimulate a cross pollination of research methods and ideas that will impact information security. MRI R2: Acquisition of High Performance Computer and Microarray Scanner for Interdisciplinary Research in Computer Science and Biology at St. Lawrence University This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
Proposal #: 09 59713
PI(s): Sharp, Richard
Dixon, Emily H.; Estevez, Ana Y.; Harcourt, Edwin; Olendzenski, Lorraine, C.
Institution: Saint Lawrence University
Title: MRI R2/Acq.: High Performance Computer and Microarray Scanner for Interdisciplinary Research in Computer Science and Biology at St. Lawrence University
Project Proposed:
This project, acquiring a microarray scanner and a high end compute server, enables the research programs of twelve faculty members from the departments of Computer Science, Biology, Statistics, and Psychology. Projects include image synthesis, comparative genome analysis in yeast, phylogenetic analysis of microbial communities and comparative analysis of bacterial genomes, microarray analysis of gene expression, and bird foraging studies. Both of these instruments will be the first of their kind on the St. Lawrence University (SLU) campus; the microarray scanner is the first in New York State north of Syracuse and the high end server, the first in any neighboring universities. The placement of a microarray scanner eliminates the need to utilize one in North Carolina that provides services at a discounted rate for faculty conducting research with undergraduate students. This currently useful but unwieldy procedure prevents faculty from expanding the number and types of microarrays used in their research, significantly limiting progress. A single high end server has been chosen, as opposed to a cluster of computers, due to the simpler programming model inherent in one machine versus many. Most of the St. Lawrence users employ software tools that are parallel, but not distributed, and hence would not benefit from a cluster. The large memory size in one machine is especially beneficial for those research projects that generate large hash tables, such as the BLAST algorithm for genome database querying, and photon mapping algorithms used in numerical simulations of light scatter.
The microarray scanner is housed in a common use equipment room close to faculty microbiology, biochemistry, genetics and cell biology labs where it will be easily accessible to all biology faculty. The scanner and space will be managed by the Department of Biology, but available for use by faculty and students in all science disciplines, as well as scientists/researchers from the broader North Country Region of New York State (Clarkson University, SUNY Potsdam, SUNY Canton, Paul Smith s College). The biology co PIs oversee day to day operations, training, scheduling of all users, and equipment maintenance (via a service contract support from manufacturers). The server will be housed in the SLU Information Technology (IT) Central Server Facility. Physical maintenance is conducted by IT s server manager and IT. User accounts are maintained, and training is conducted. A multidisciplinary Project Steering Committee oversees the use and management of the equipment. With the new microarray scanner on site, the volume of microarray processing will increase, as will the costs associated with running these experiments. These higher research costs will be covered through research support funds from the Biology Department. At the completion of the project period, SLU assumes all costs for the day to day operation and long term maintenance.
Broader Impacts:
This instrumentation strongly impacts SLU and strengthens SLU s undergraduate research culture. The new equipment will allow undergraduates in these disciplines to gain experience with key scientific technologies and relevant methodologies. However, the reach of the equipment is not limited to SLU alone; two research projects from neighboring Clarkson University have been identified, and will be a focal point for The Associated Colleges of the St. Lawrence Valley, an educational consortium between St. Lawrence University and three neighboring universities: Clarkson University, State University of New York, (SUNY) Potsdam, and SUNY Canton. This consortium was created to expand the number and variety of educational opportunities for faculty and students, to share resources, and to innovate through joint action. Emphasis is placed on the recruitment of students from underrepresented backgrounds. MRI R2: Acquisition of Shared Cluster and Database Computing Facilities at Wesleyan University This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
Proposal #: 09 59856
PI(s): Starr, Francis W.; Beveridge, David, L.; Weir, Michael, P.
Institution: Wesleyan University
Title: MRI R2/Acq: Shared Cluster and Database Computing Facilities at Wesleyan University
Project Proposed:
This project, acquiring a shared, state of the art, mid sized super computing cluster and database server, supports intensive research computing and corresponding initiatives at the institution. The current central facilities are no longer able to meet the intensive research computing and corresponding initiatives in education, as evidenced by a five fold increase in the number of pending jobs over a period of two years. The requested cluster roughly doubles the size of the central super computing facility, as well as adds new support for informatics based research that relies heavily on relational database mining.
Research programs in astrophysics, biomaterials, liquid state chemical physics, nanotechnology, quantum chemistry, molecular biophysics, neuroinformatics, and structural bioinformatics, all require high end computing resources and have become increasingly dependent on high performance computing facilities. All these areas offer courses that involve hands on computation. Currently, a large number of faculty, staff and students are active research users, and roughly 2/3 of those are students; additionally, 40 50 students per year make use of facilities for course related computational activities.
Based on consultations with the current primary users of the system and the systems administrator, the new proposed cluster consists of 40 50 CPU nodes, each with two Intel Xeon quad core processors, and containing 8 16 GB of storage. An infiniband interconnect will connect the processors to facilitate parallel applications. Data storage is provided for the user calculations via a RAID array of 30 40 TB. The dedicated database server has dual 4 or 6 core processors, a large memory footprint, and high performance local disks to store the database.
The system is housed in the renovated facility on the Information Technology Services (ITS) floor of the Wesleyan Science Tower, maintained by a designated ITS systems manager and made available for research and instructional needs over the current WESNET network. A faculty Computer Advisory Committee provides academic oversight, manages allocation requests, and coordinates with ITS. The institution continues to support the administration of the new system.
Broader Impacts:
The instrument services both the faculty research and student education. It enables new science to be produced in the course of diverse research projects and impacts the ongoing instruction in the University, serving as a learning tool to develop student scientific computing proficiency both through existing courses and participating in faculty led research. To further extend the educational and training goals, a new Scientific Computing and Informatics Center (SCIC) was developed with funding support from the University, aimed at facilitating the effective use of the new high performance computing facility by all faculty and students, and supporting the educational initiatives of courses utilizing computational resources. The SCIC also operates a summer program supporting undergraduate research and developing teaching modules (in collaboration with relevant faculty) that directly impacts many courses across the curriculum.
The cluster and database facility also serve as a focal point and resource for undergraduate students in computationally intensive courses of study, including a recently approved interdisciplinary Certificate Program in Informatics and Modeling . This interdisciplinary program functions effectively as a minor field of study alongside major fields. The program has two elements, a computational and an informatics foci. MRI R2: Development of a Software Traceability Instrument to Facilitate and Empower Traceability Research and Technology Transfer This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

This project, developing a software traceability instrument, enables research in diverse, much needed, software related research areas such as formal methods, human computer factors, software visualization, and project management. The work supports a critical research agenda of the software engineering community and facilitates technology transfer of traceability solutions to business and industry. Traceability refers to the ability to capture the relationships between different artifacts of a software intensive system development project, including code, requirements, design, specifications, external regulations, and software architectures. Many software engineering activities involve tracing between interdependent artifacts, for example, finding software components that implement a given system requirement in order to assure that an as built system meets its intended goals, or conversely, finding all the requirements that pertain to a code module to ensure that the system does not contain extraneous and potentially malicious features. The traceability instrument contains a library of reusable trace algorithms and utilities, a benchmarked repository of trace related datasets, tasks, metrics, and experimental results, a plug and play environment for conducting trace related experiments, and predefined experimental templates representing common types of empirical traceability experiments. The instrument facilitates the application of traceability solutions across a broad range of software engineering activities including requirements analysis, architectural design, maintenance, reverse engineering, and IV&V (independent verification and validation) or V&V activities.

Broader Impacts: Despite the criticality of software traceability, organizations have struggled to implement successful and cost effective traceability due to the complexity and error proneness of the task. Despite a compelling research agenda, traceability research has been impeded by the lack of shared instruments. The instrument will make state of the art traceability results available to enable the next generation of customized traceability environments and will provide support for conducting empirical research. Results will be disseminated broadly through outreach endeavors by the Center of Excellence for Software Traceability. The project will provide multiple opportunities for participation by underrepresented minorities and undergraduate students, and will create three fulltime job positions. The long term broader impacts of the project will be to improve software project productivity and to improve the reliability of software, as well as developing software engineering educational curriculum for training students and practitioners. MRI R2: Development of an Immersive Giga pixel Display This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
Proposal #: 09 59979
PI(s): Kaufman, Arie E., Mueller, Klaus, Qin, Hong, Samaras, Dimitrios, Varshney, Amitabh
Institution: SUNY at Stony Brook
Title: MRI R2: Development of an Immersive Giga pixel Display
Project Proposed:
This project, developing a next generation of immersive display instrument (called Reality Deck ), aims to explore and visualize data from many fields. To satisfy the need driven by the explosive growth of data size and environments already at the institution, the work builds on the existing experience with immersive environments (e.g., the Immersive Cabin a current generation device using projectors). This unique project generates a one of a kind exploration theater, using 308 high resolution 30 LCD display monitors, by contributing an environment whose visual resolution is at the limit of the human eye s acuity. Within this environment investigators can interact with the data/information displayed.
The instrument services many groups, including visual computing, virtual and augmented reality, human computer interfaces, computer vision and image processing, data mining, physics, scientific computing, chemistry, marine and atmospheric sciences, climate and weather modeling, material science, etc.
Collaborating scientist s applications will be ported to the RealityDeck, including applications in nanoelectronics, climate and weather modeling, biotoxin simulations, microtomography, astronomy, atmospheric science, G pixel camera for intelligence gathering, architectural design and disaster simulations, smart energy grid, and many others.
A unique assembly of displays, GPU cluster, sensors, communication/networking, computer vision and human computer interaction technologies, RealityDeck is an engineering feat with user studies to deliver a holistic system with a significant societal and research value. It is a one of a kind pioneering G pixel display approaching the limits of visual cognition that provides functionalities to a diverse user base. Its resolution at the eye s visual acuity and its field of view will always exceed that of a human user (wherever a human chooses to look), satisfying visual queries into the data in a very intuitive way. This visual interaction is tightly coupled with physical navigation.
This surround virtual environment consists of inertial sensors and six cameras mounted around the top corners of the RealityDeck room to allow interaction with the displays. The display system is driven by a cluster of about 80 high end computer nodes, each equipped with two high end GPUs. A small scale video wall has already been constructed as an experimental platform for the RealityDeck consisting of 9 high resolution 30 LCD panels in a 3×3 configuration.
Broader Impacts:
The instrument will be used for research, education, and outreach across many departments at the institution, the University of Maryland, and Brookhaven National Laboratory (BNL). It fosters collaborations across disciplines attracting faculty, researchers, and students. RealityDeck significantly enriches the quality of visual thinking and data exploration. It substantially enhances the infrastructure of research and education and has the potential to alter the way computer graphicists, engineers, and scientists work and/or conduct scientific discoveries. MRI R2: Development of an Instrument for Information Science and Computing in Neuroscience This award is funded under the American Recovery and Reinvestment Act of 2009(Public Law 111 5).
Proposal #: 09 59985
PI(s): Adjouadi, Malek; Barreto, Armando B.; Gaillard,William; Jayakar,Prasanna; Rishe, Naphtali D.
Institution: Florida International University
Title: MRI R2/Dev: Instrument for Information Science and Computing in Neuroscience
This project, developing an instrument for information processing and computing that enables cohesive study of the brain, involves the new concept of a 5 D brain processing platform while addressing the challenge of finding the best way to put together five dimensions to provide a complete picture of brain dynamics. This unified approach brings together fields of Computer Engineering, Computer Science, Electrical Engineering, and Bioengineering to create an instrument for precisely measuring and visualizing significant information and results across the five dimensions, three from spatial data, time, and an imaging modality that serves as the fifth dimension. Accomplishing this mission involves advanced designs, hardware software integration mechanisms, and novel interfaces that bring competing, and sometimes diverging, technologies into a unified brain research platform. Combining a Multi site Data Repository, Modality Integration and Computational, Visualization, and Operation Support Units, this new instrument is expected to bring:
New insights into brain structure, functional correlations and dynamics, both in its normal state and under specific pathological conditions,
Far improved mapping of patterns of brain activity,
Integration of multimodal technologies in order to augment their capabilities with new insights while consolidating high spatial resolution with high temporal resolution,
Database design and management augmented with mechanisms for fast user interaction and visualization to meet the challenge of managing complex spatio temporal datasets, posing complex queries, and establishing effective methods for data representation and visualization, and
Resolution of those paradigms that confront heavy computational requirements and compatibility problems that arise from the use of different recording modalities and diverse software platforms.
The project aims to collectively overcome the primary barriers in identifying the different factors that influence the functional organization of the brain and its underlying pathology. As an example, it delves in the epileptic seizure that can be mapped over time as it moves along specific fiber tracts that may enable better identification of specific areas of treatment and consequently protect functionally important parts of the brain during surgery that may lead to better and safer outcomes.
Broader Impacts:
Fostering an environment that supports cross disciplinary initiatives, joint collaboration and programs, the instrument establishes a research platform that enables academic institutions and hospitals to investigate multi site collaborative studies in accordance to systematically administered standardize protocols to a database of common assessments and measures. The project extends the breadth and depth of multidisciplinary efforts including new paradigms and findings. Moreover, the project advances the education, research, and training of many students in a minority serving university. III:EAGER:Collaborative Research: A Collaborative Scientific Workflow Composition Tool Supporting Scientific Collaboration The goal of this research aims to produce a general purpose but domain customizable collaborative scientific workflow tool for accelerating scientific discovery, particularly facilitating large scale and cross disciplinary research projects that are collaborative in nature and require intensive user interaction from multiple distributed domain scientists. As a natural extension to the existing single user oriented scientific workflow management tools by providing direct system support for scientific collaboration, this project seeks to pave a way toward a next generation tool supporting scientific collaboration over the Internet. The expected tool can be easily expanded to support data centric collaborative information management in any intelligence community.

To achieve this broader impact, this research seeks to establish a set of foundational models and techniques supporting collaborative scientific workflow composition and management; and based on them, construct an Internet based collaborative scientific workflow tool framed with an open source initiative. The initial focus of the Early Concept Grants Exploratory Research (EAGER) project is on rapidly creating a prototype tool as a proof of concept, equipped with basic dataflow oriented scientific workflow models and collaboration patterns integrated in an agile service oriented architecture. To avoid reinventing the wheel, the efforts will be concentrated on transforming and extending Taverna, a known open source scientific workflow management tool, into a collaborative version. Evaluations and validations will be conducted through partnerships with multiple collaborative scientific communities.

For further information, see the project website at
http://www.CollaborativeScientificWorkflows.org/ MRI R2: Development of Common Platform for Unifying Humanoids Research This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
Proposal #: 09 60061
PI(s): Kim, Youngmoo E., Gogotsi, Yury, Hong, Dennis H., Regli, William C., Schaal, Stefan
Institution: Drexel University
Title: Development of Common Platform for Unifying Humanoids Research
Project Proposed:
This project, developing and disseminating HUBO+, a new common humanoid research platform instrument, enables novel and previously infeasible capabilities for future research efforts while working with a common instrument. HUBO will be the first homogeneous, full sized humanoid to be used as a common research and education platform. Eight universities (Drexel, CMU, MIT, Ohio State, Penn, Purdue, Southern California, and VaTech), representing a critical mass of humanoids research within US, participate in this development of the world s first homogeneous full sized humanoid team. Building upon unique expertise, the work extends current capabilities, resulting in six identical units, facilitating the following potentially transformative advances in robotics:
A state of the art, standardized humanoid platform instrument with embedded capabilities for sensing, manipulation, and rapid locomotion, ideal for a broad range of future humanoids research
The ability, for the first time, to directly compare and across validate algorithms and methodologies and consistently benchmark results across research teams
Novel energy storage technology for mobile robotics incorporating supercapacitors for operations requiring high power density, far exceeding the capabilities of traditional battery only power sources
A widely distributed platform that motivates, recruits, and trains a broad range of students spanning multiple disciplines, including artificial intelligence, digital, signal processing, mechanics, and control
Humanoids, robots engineered to mimic human form and motion, open broad avenues of cross disciplinary research spanning multiple fields, such as mechanical control, artificial intelligence, and power systems. Common humanoids are rarely autonomous and are not ready for unconstrained interaction with humans. The most compelling demonstrations are meticulously pre programmed and painstakingly choreographed. A few common platforms have already advanced some research. Hence, having a consistent platform should facilitate rapid progress in areas needed for autonomy and natural interaction, including mobility, manipulation, robot vision, speech communication, and cognition and learning. However, although currently Japan and Korea are considered world leaders in design and construction of humanoids, best practices have not been developed for constructing multiple, identical humanoids. These conditions call for the making of an urgently needed benchmark providing evaluations and cross validation of results. With this development and the servicing of 6 humanoids, this project aims to create knowledge and best practices contributing to robotics research, possibly leading to the standardization needed for ubiquity.
Broader Impacts:
The instrument enables US researchers to develop expertise in the design and construction of humanoids, while the distribution of the work activities ensures the broad dissemination of the knowledge. Humanoids research, inherently interdisciplinary and integrative, inspires young students. The graduate and undergraduates students participating are likely to receive a world class training in robotics. Outreach partners, including several high profile museums will introduce people of all ages to the exciting technologies of robotics, particularly useful in recruiting K 12 students into science, engineering, mathematics, etc. A partnership with the Science Leadership Academy (SLA), a magnet school with more than 63% underrepresented students, assures their involvement. With SLA, the project initiates an annual program modeled on a NASA style experiment design competition, in which students use simulation tools to propose humanoids projects and activities. Selected winner(s) will have their proposed projects implemented on HUBO. MRI R2: Acquisition of a Heterogeneous Supercomputing Instrument for Transformative Interdisciplinary Research This award is funded under the American Recovery and Reinvestment Act of 2009(Public Law 111 5).
Proposal #: 09 60081
PI(s): Feng, Wuchun
Hilu; Khidir W.; King, Scott D.
Institution: Virginia Polytechnic Institute and State University
Title: MRI R2/Acq:Heterogeneous Supercomputing Instrument for Transformative Interdisciplinary Research
Project Proposed :
This project, acquiring a versatile heterogeneous supercomputing instrument, called HokieSpeed, in which each compute node consists of CPUs (central processing units) and GPUs (graphics processing units), aims to empower faculty, students, and staff across disciplines to tackle problems previously viewed as intractable or that required heroic efforts and significant domain expertise to solve. The instrument is expected to catalyze new approaches for conducting research via synergistic amalgamation of heterogeneous supercomputing and cyber enabled tools that enhance ease of use. In particular, it should allow end users to:
Make commonplace the ability to perform in situ visualization for rapid visual information synthesis and analysis, and
Control their level of immersion in the discovery process from being immersed (a la human in the loop intuitively making real time decisions via a large scale gigapixel display), to observing the instrument automatically collect, organize, and analyze data of visual analytics.
Recent trends have exposed the CPU as a jack of all computing trades, master of none, thus giving rise to heterogeneous computing instruments with multiple types of brains (e.g., CPUs and GPUs). Though conventional wisdom believed in 2007 that missing genes in 699 microbial genomes was computationally infeasible, the PI led a team of more than twenty interdisciplinary researchers from eight institutions around the world and developed software cybertool instruments to integrate a set of distributed supercomputing with more that twelve thousand CPUs to complete the task in ten weeks. Coupled with the existing cybertools and those available from NVIDIA the instrument should enable completion of the task in a day with only two researchers. HokieSpeed will have 8,192 CPU cores and at least 61,440 GPU cores in 40 square feet and deliver 35 times better peak performance, 70 times better power efficiency, and 14 times better space efficiency than the current system at the institution (circa 2003) while offering versatility by enabling discovery through a large scale gigapixel display and appropriately varying degrees of interaction between human and machine. Creating new forms of social research, and what if scenarios for evaluation and promising to deliver research advances across multiple disciplines, the results will be generated often in real time allowing immediate in situ visualization on a gigapixel display.
Broader Impacts:
Reaching a broad community, the instrument is expected to herald a new age in multi purpose and interdisciplinary computing instrumentation creating a computational ecosystem that simplifies access and use of the instrument to the masses. Students, including females and underrepresented minorities, will learn about energy efficient computing, and visualization and interaction at large scale. Among others, the project will also utilize programs that specifically service broader impacts: MAOP (Multicultural/Minority Academic Opportunities Program), CTech2 (Computers and Technology at VaTech), and CRA W s CREU (Collaborative Research Experience for Undergraduates). Additionally, it will extend an experimental GPU course and offer a workshop version for on line instruction. Other activities include K 12 outreach, and delivery by CREU of information technology (IT) to rural and economically disadvantaged areas via virtual computing and the VaTech STEM K 12 Outreach Initiative. MRI R2: Acquisition of a Networked AUV based Instrument for the Southern California Bight This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
Proposal #: 09 60163
PI(s): Sukhatme, Gaurav; Caron, David A.; Edwards, Katrina J.; Jones, Burton, H; Mitra, Urbashi
Institution: University of Southern California
Title: MRI R2: Acq. of a Networked AUV based Instrument for the Southern California Bight
Project Proposed:
This project, acquiring a networked instrument composed of two complementary Autonomous Underwater Vehicles (AUVs), supports an extensive program of research in robotics, underwater acoustic communication and networking, marine biology, oceanography, and biogeochemistry at the Center for Integrated Networked Aquatic PlatformS (CINAPS). These two AUVs add capability to those already in use by CINAPS. The work addresses two current key limitations: limited depth (depth no greater than 200m) and limited communication (only supports communication through the air, i.e., when the vehicles are not below the surface.) These instruments impact agendas in two main fields: research in the Southern California coastal marine ecosystem (in physical oceanography, geobiological oceanography, and microbial ecology), and research in robotics, communication, and networking.
The instrumentation consists of a Slocum Electric Glider and an EcoMapper EP (Expandable Payload) Autonomous Underwater Vehicle. The Slocum glider, for use up to 1000m, was developed by in the early 1990 s, and has become an integral tool of ocean science in this decade. It is driven entirely by a variable buoyancy system rather than active propulsion. The glider s wings convert the buoyancy dependent vertical motion into forward velocity. Slocum gliders are truly autonomous, requiring a surface vessel only for deployment and recovery. On board communications capabilities include a two way RF modem, Iridium satellite and an ARGOS locator. With its onboard instrumentation, combined with its mobility and long range capabilities, the glider provides continuous, near real time information, about the physics and biogeochemistry of the ocean. Designed specifically for water quality and bathymetry mapping applications, the Ecomapper for shallow use (up to 200m) is a unique AUV. Easily deployed by one person, it is able to perform a wide area survey without a surface workboat or associated personnel. The EP class EcoMapper allows for additional ports for customized sensor integration and space for a second low power CPU to support the additional sensors and software. This additional CPU enables mission adaptations to occur on the fly based on real time sensor readings that are necessary when trying to detect and track dynamically changing oceanic features a range of important processes and associated questions can only be studied with the coupling of deep AUV operations (up to 1000m) with shallow (up to 200m). These include (a) the toxic effect of harmful algal blooms at greater depth, (b) the temporal and spatial variations of the low oxygen interface (deep basins can become hypoxic or anoxic due to isolation), and (c) fluxes of particulate material from urban runoff to the deeper sea that are discontinuous in both time and space; the spatial extent of which can be best resolved with autonomous vehicles that can maintain a presence over several events. To enable these studies, new algorithms for multi AUV communication and control are necessary. The ability to coordinate vehicle trajectories and missions while submerged presents a current limitation. This is of particular concern in the Southern California Bight, the study site, a region with high maritime traffic where surfacing of the AUVs needs to be minimized. Used as a field instrument deployed in the Southern California Bight (SCB), the instrument is programmed and bench tested in the Robotic Embedded Systems Laboratory on the USC main campus in LA. The instrument manager is supported by the PI;s grants. USC covers all operational and maintenance costs.
Broader Impacts:
The instrumentation impacts faculty research and student education, contributing to enable new science in diverse research projects and impacting the ongoing instruction at the institution and serves as a learning tool to develop student scientific proficiency (through existing courses and participation in faculty led research). This acquisition plays an important role in the Southern California Bight Study planned for Spring 2010, coordinated by the Southern California Coastal Water Research Project (SCCWRP), a public agency focusing on the coastal ecosystems of Southern California. The PIs have close collaborative ties to this agency. Moreover, the networked and adaptive sensor systems provide important data relating to the climate and health of Southern California coastal ocean.
Undergraduate students are involved through REU site awards at USC. Since the PI runs the USC Computer Science department REU site for research in various areas of computer science, students in his lab often assist with data analysis and programming for the robotic boats. The EcoMapper vehicle, which is designed to be portable and deployable in shallow water, is particularly suitable for the next generation of REU students. Other PIs participate as faculty mentors in USC biochemistry REU site encouraging similar participation from undergraduate students in biology. MRI R2: Acquisition of a Configurable Supercomputer for Arctic Research This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

The University of Alaska Fairbanks campus will install a research supercomputer that will be a broadly accessible instrument for investigation of phenomena related to the Arctic as well as a set of other key research areas. The instrument will be utilized by researchers from a number of domains, including geophysical science, mathematics, biology, supercomputing research, and engineering. The importance of the Arctic for understanding global climate change and its many influences is a driving force in bringing together these researchers, to address common interests from different angles, to share results widely through publications and presentations, and to integrate research and instrument use in the classroom. For the scientific areas, research will be both normative and transformative. Transformative results are part of the instrument design proposal, by providing a rich and useful resource for collaboration, not just for individual research projects.

Advances are envisioned in geophysical phenomena, including increased understanding of high latitudes roles for climate change, the atmospheric sciences, the oceans, ice sheets, and sea ice. In biology and ecology, advances are envisioned in population genetics, marine ecosystems, and Boreal ecosystems. In the human sphere, advances are anticipated in human ecosystem interaction and the influence on human infrastructure, on integration of datasets for native language study, and understanding of environmental influences on air and water quality. Engineering concerns, ranging from analysis of next generation supercomputers to radio wave propagation, will see rapid progress.

This research equipment infrastructure enhancement will have broader impacts in a number of ways. As a minority serving institution in an EPSCoR state, UAF will leverage the instrument to provide opportunities for involvement of minorities and underrepresented groups in teaching, research and outreach. In addition, there is a significant community of researchers at UAF who utilize computational methods for discovery who will be enabled by this instrument. MRI R2: Acquisition of a high performance central computing facility at UCSB This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).
Proposal #: 09 60316
PI(s): Brown, Frank, L.; Fredrickson, Glenn, H.; Garcia Cervera, Carlos; Gilbert, John, R.; Van de Walle, Christian, G.
Institution: University of California Santa Barbara
Title: MRI R2: Acquisition of a High Performance Central Computing Facility at UCSB
Project Proposed:
This project, acquiring of a computational cluster to replace a six year old system, allows access to fast mid sized parallel computation to dozens of researchers and serves as the institution s primary resource for parallel computation. The system is structured for a variety of uses. Standard MPI computation is carried out with a tightly coupled cluster of quad core processors, while fat nodes with 256 GB of RAM as well as local high speed disk storage service jobs that require large shared memory. Researchers actively developing codes can take advantage of the unique performance characteristics of GPU (graphics processing) nodes. The system will have several NVidia Tesla nodes.
The users of the system are drawn from all five departments of the College of Engineering (Chemical Engineering, Computer Science, Electrical & Computer Engineering, Materials, and Mechanical Engineering), seven departments of the Division of Mathematical, Life and Physical Sciences (Chemistry & Biochemistry, Earth Science, Ecology Evolution & Marine Biology, Mathematics, Molecular Cellular & Developmental Biology, Physics, and Psychology), as well as the departments of Economics, Geography, and Media Arts & Technology from the Division of Humanities and Social Sciences. In addition, the system supports research in eight campus centers: Allosphere Research Facility, the California NanoSystems Institute, the Center for Polymers and Organic Solids, the Institute for Crustal Studies, the Kavli Institute for Theoretical Physics, the Materials Research Laboratory, the National Center for Ecological Analysis & Synthesis, and the Neuroscience Research Institute.
The new system will be housed in the same location as the previous one where the same successful administrative and maintenance procedures used for the past six years will be applied. The proposed cluster will be accessible via the UC Grid, a web portal interface that makes high performance computing resources easy to use from desktop machines (PCs or Macs). The acquired system will come with a three year warranty. Prior experience has shown that only a small number of nodes are expected to malfunction during the useful lifetime (> 3 years) of the cluster.
Broader Impacts:
The research enabled by the campus wide facility, interdisciplinary and collaborative in nature, is available to the broad research community. The large majority of users, roughly 75%, consists of postdocs, graduate students, and undergraduates (5%), allowing this award to accomplish NSF s longstanding goal of integrating research and education. Outreach to K 12 takes place via a new initiative, The School for Scientific Thought (SST), an extension to the Let s Explore Physical Sciences (LEAPS) Program. Under the SST Program UCSB science and engineering graduate students design and teach a course for an audience of high school students on Saturdays. In addition, many of the faculty associated with this proposal participate in the UC Leadership Excellence through the Advanced Degrees program that has increased the number of underrepresented students in science and engineering at UCSB. MRI R2: Acquisition of an Applied Computational Instrument This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111 5).

Building on the success of a previousMRI funded project, an interdisciplinary group of computer scientists, psychologists, biologists, chemists, and physicists at the University of Oregon is acquiring a large scale computational resource, the Applied Computational Instrument for Scientific Synthesis (ACISS), to support continued cutting edge scientific research in these areas. The ACISS hardware will consist of general purpose multicore computing nodes, high performance computing nodes augmented with GPGPU acceleration, a 400TB storage system, high bandwidth networking infrastructure and additional computing resources that will be incorporated into an existing visualization lab in the Department of Computer and Information Science. A key part of the proposed infrastructure is the unique opportunity to manage ACISS as a computational science cloud.

The ACISS infrastructure will allow an expanded the scope for the current projects in the areas of software tools for performance measurement, programming environments and languages for describing and executing complex simulations and scientific work flows, new algorithms for multiple sequence alignment and phylogenetic inference and undertake new projects in support of the domain sciences. Research projects that will benefit include: a) modeling neural networks in C. elegans to better understand the neural mechanisms responsible for chemotaxis and klinotaxis, and investigation of the evolution of genes involved in development and their role in speciation and phenotypic variation; b) development of neuroinformatic techniques used in brain imaging and analysis, integrating structural information from fMRI and other sources with EEG data; c) molecular modeling research, including the definition of new techniques for meso scale modeling and applying computational methods to understand phase transitions and nitrogen fixation; d) astrophysical simulations of turbulent plasma flows that influence the early stages of planet formation.

The ACISS infrastructure will provide the computational resources necessary for future multidisciplinary research. ACISS will establish a novel paradigm for computational science research and practice. The experience gained in early adoption of the ACISS cloud computing technologies will allow us to more rapidly apply this knowledge to create new scientific work flows, more productive research collaborations, and enhanced multidisciplinary education programs. Farther reaching, ACISS can be seen as a model for translational computational science, in which ACISS based services function as cyber incubators where new work flows for scientific research are prototyped. III EAGER Collaborative Research: Exploratory Research on the Annotated Biological Web The life science research community generates an abundance of data on genes, proteins, sequences, etc. These are captured in publicly available resources such as Entrez Gene, PDB and PubMed and in focused collections such as TAIR and OMIM. A number of ontologies such as GO, PO and UMLS are in use to increase interoperability. Records in these resources are typically annotated with controlled vocabulary (CV) terms from one or more ontologies. Records are often hyperlinked to those in other repositories, creating a richly curated biological Web of semantic knowledge.

The objective of this project is to develop tools to explore and mine this rich Web of annotated and hyperlinked entries so as to discover meaningful patterns.
The approach builds upon finding potentially meaningful and novel associations between pairs of CV terms cross multiple ontologies. The bridge of associations across ontologies reflects annotation practices across repositories. A variety of graph data mining and network analysis techniques are being explored to find complex patterns of groups of CV terms cross multiple ontologies. The intent is to identify biologically meaningful associations that yield nuggets of actionable knowledge to be made available to the scientist together with a set of golden publications that support the identified patterns.

The intellectual merit of the project is that it is unique in comparison to other bioinformatics data integration and analysis projects. Data is integrated from across numerous sources including genes, gene annotations, ontologies, and the literature. The exploratory nature (EAGER) of this research is both with respect to the biological and the computer science disciplines. From the biological viewpoint, a high level of speculation is associated with any discovered biological patterns. Discovered patterns night not necessarily meet criteria for experimental validation. The research methodology combines algorithmic and analytical techniques from multiple computer science sub disciplines. While specific technical innovations are expected, an inter related set of computer science challenges needs to be defined.
This research has the potential for broader impact since the methodology can be applied to any type of interlinked resources on the biological semantic Web as well as to any collection of hyperlinked resources. This research is a collaboration between the University of Maryland and the University of Iowa. For further information see the project web pages at the following URL:
http://www.umiacs.umd.edu/research/CLIP/RSEAGER2009/ RI: Medium: Collaborative Research: Game Theory Pragmatics The past decade has seen unprecedented interaction between artificial intelligence and game theory, with exciting intellectual problems, technical results, and potentially important applications. However, this thriving interaction has thus far not challenged in a fundamental way some of the basic assumptions of game theory, to include various forms of equilibrium as the fundamental strategic concept. Equilibrium specifies conditions under which the strategic choices of agents are in some sense stable. Equilibria are clever and beautiful constructs, but they embody strong idealized assumptions and as a result their applicability to complex, realistic games (i.e., formalizable social interactions) is limited. Arguably computer science can provide alternative modeling foundations, or at least significantly contribute to them.

This project explores several complementary directions, to include: alternatives to equilibrium as game solution criteria; replacing analysis of large, complex games with analysis of their abbreviated or approximate versions; using machine learning techniques to model the extent to which agent behavior is strategic, adaptive, or otherwise intelligent; investigating the role of strategic reasoning in controlled but rich environments, such as Computational Billiards, which involves continuous state and actions spaces as well as control uncertainty. One of the outreach and educational components of this project is organizing and participating in an annual Computational Billiards competition. Applications range from electronic commerce to social networks to peer to peer systems to online games, and in general all settings in which individual interests intertwine with computational elements. RI: Medium: Collaborative Research: Unlocking Biologically Inspired Computer Vision: A High Throughput Approach This project exploits advances in parallel computing hardware and a neuroscience informed perspective to design next generation computer vision algorithms that aim to match a human s ability to recognize objects. The human brain has superlative visual object recognition abilities humans can effortlessly identify and categorize tens of thousands of objects with high accuracy in a fraction of a second and a stronger connection between neuroscience and computer vision has driven new progress on machine algorithms. However, these models have not yet achieved robust, human level object recognition in part because the number of possible bio inspired model configurations is enormous. Powerful models hidden in this model class have yet to be systematically characterized and the correct biological model is not known.

To break through this barrier, this project will leverage newly available computational tools to undertake a systematic exploration of the bio inspired model class by using a high throughput approach in which millions of candidate models are generated and screened for desirable object recognition properties (Objective 1). To drive this systematic search, the project will create and employ a suite of benchmark vision tasks and performance report cards that operationally define what constitutes a good visual image representation for object recognition (Objective 2). The highest performing visual representations harvested from these ongoing high throughput searches will be used: for applications in other machine vision domains, to generate new experimental predictions, and to determine the underlying computing motifs that enable this high performance (Objective 3). Preliminary results show that this approach already yields algorithms that exceed state of the art performance in object recognition tasks and generalize to other visual tasks.

As the scale of available computational power continues to expand, this approach holds great potential to rapidly accelerate progress in computer vision, neuroscience, and cognitive science: it will create a large scale laboratory for testing neuroscience ideas within the domain of computer vision; it will generate new, testable computational hypotheses to guide neuroscience experiments; it will produce a new kind of multidimensional image challenge suite that will be a rallying point for computer models, neuronal population studies, and behavioral investigations; and it could unleash a host of new applications. CIF: Medium: Iterative Decoding Beyond Belief Propagation Error correcting codes are an integral part of modern day communications, computer and data storage systems and play a vital role in ensuring the integrity of data. At the heart of modern coding theory is the fact that the low density parity check codes can be efficiently decoded by the algorithm known as belief propagation (BP). The BP is an iterative algorithm which operates on a graphical representation of a code by sending coded bit likelihoods beliefs. The project establishes a new paradigm and develops tools for the design and analysis of decoding algorithms which are much simpler yet better than belief propagation. This novel paradigm provides a new angle in addressing a fundamental coding theory questions and a methodology for designing a class of decoding algorithms with provable performance and large flexibility in controlling complexity and speed.

Unlike BP decoders, these decoders do not propagate beliefs but a rather different kind of messages that reflect the local structure of the code graph. The methodology for designing such decoders involves identifying graphical structures on which traditional decoders fail, and deriving message passing rules that can correct a majority of these structures with minimal number of bits used in the messages. New and successively better decoding algorithms are built by adding more bits to the messages passed in a simpler decoder. The project develops a comprehensive framework to study decoders that achieve the best possible trade off between the complexity and performance in the low noise region. Also by increasing the number of bits to represent the input alphabet successively better approximations of the behavior of the decoders for continuous channels are obtained. SHF: Medium: Combining Speculation with Continuous Validation for Software Developers Unprecedented computational power is available from multi core processors and cloud computing. To date, this power has been used primarily to make programs run faster. However, in many cases the bottleneck to solving users problems is in the challenge of creating the software, not in the time to run it. This project will apply computational power to the real bottleneck, providing developers with new types of feedback. As a key broader impact, the research will enable developers to create software more quickly, more cheaply, and with higher quality.

The key technical idea is to inform developers, in advance, of the consequences of their likely actions. The development environment speculates about developer actions, evaluates the effect of each action (on compilation, tests, version control conflicts, etc.), and unobtrusively makes this information available to the developer. By knowing which choices are good and which are bad, developers can avoid bad choices that cost time or reduce quality. The project s intellectual merits include algorithms to quickly create and evaluate many possible developer actions, UI design for developer awareness, and evaluation of how increased awareness of contingent information, about possible actions, affects developers. This also leads toward an answer to the question: If developers had infinite processing power, what fundamental software engineering research problems would remain? AF: Medium: Collaborative Research: Solutions to Planar Optimization Problems The aim of this research is to develop new algorithms and algorithmic
techniques for solving fundamental optimization problems on planar
networks. Many optimization problems in networks are considered
computationally difficult; some are even difficult to solve
approximately. However, problems often become easier when the input
network is restricted to be planar, i.e. when it can be drawn on the
plane so that no edges cross each other. Such planar instances of
optimization problems arise in several application areas, including
logistics and route planning in road maps, image processing and
computer vision, and VLSI chip design.

The investigators plan to develop algorithms that achieve faster
running times or better approximations by exploiting the planarity of
the input networks. In addition, in order to address the use of
optimization in the discovery of some ground truth, the investigators
will develop algorithms not just for the traditional worst case input
model but also for models in which there is an unusually good planted
solution; for a model of this kind, the investigators expect to find
algorithms that produce even more accurate answers.

The research will likely uncover new computational techniques whose
applicability goes beyond planar networks. In the recent past, once a
technique has been developed and understood in the context of planar
networks, it has been generalized to apply to broader families of
networks.

In addition, new algorithms and techniques resulting from this
research might enable people to quickly compute better solutions to
problems arising in diverse application areas. For example, research
in this area has already had an impact in the computer vision
community. Further research has the potential to be useful, for
example, in the design of networks, the planning of routes in road
maps, the processing of images. AF: Medium: New Directions in Coding Theory and Pseudorandomness The probabilistic method is a powerful tool to establish the existence
of diverse objects of importance in several applications. For example,
Shannon s famous theorem asserts that a random codebook can be used
for reliably transmitting information at optimal rates on a noisy
channel. It is well known that a random graph is typically Ramsey
and has no large clique or independent set, and a random sparse graph
is very likely to be an expander with excellent connectivity. Yet, in
applications it is important to explicitly construct such an object
with a certified guarantee of the desired property. Obtaining such
constructions of comparable strength to what is guaranteed by the
probabilistic method is typically much harder and often unknown.

Pseudorandomness is a broad area that deals with efficiently
generating objects that exhibit the desirable properties of
random like objects despite being constructed either explicitly or
with limited randomness. Such pseudorandom constructions are important
in the study of error correcting codes, complexity theory,
combinatorics, cryptography, and high dimensional geometry. Research
in recent years has addressed some of these challenges and led to
powerful constructions of error correcting codes, expander graphs,
randomness extractors, Ramsey graphs, compressed sensing matrices,
etc. Despite the seemingly different definitions and motivations for
the study of these objects, much of this progress was based on
insights uncovering intimate connections between them, leading to a
rich theory with a common pool of broadly useful techniques.

This progress notwithstanding, explicit constructions with optimal
parameters typically remain open, and the area is full of exciting new
directions motivated by emerging applications. This project will
involve a comprehensive collection of interconnected research
activities focusing on the theory of error correcting codes and
pseudorandomness. The directions pursued will include strengthening
the existing connections between various pseudorandom constructs and
discovering new computational applications thereof, and investigating
the pseudorandom properties of codes and related objects that have
important structural characteristics often needed in applications
(such as linearity or sparsity). Topics in coding theory inspired by
complexity theory such as list decoding and locally testable codes,
and codes for poorly understood noise models such as deletion channels
will be studied. Another important goal of the project is to bridge
the gap between worst case and probabilistic noise models via codes
for channels with natural computational restrictions.

The research will use ideas from computer science in setting new
directions for research in coding theory as well as discovering new
constructions of codes and decoding algorithms, thereby enhancing the
connection between the computer science and information theory
communities. The discovery of new coding schemes has potential direct
applications in communication and storage of data. Many of the
questions to be addressed, therefore, have a natural practical
connection alongside their fundamental theoretical appeal. On the
education front, the project will engage several graduate students and
provide a stimulating research environment for them, and help with the
planned writing of a goal oriented textbook on coding theory. AF: Medium: Algorithmic Research in Game Theory, Networks, and Biology Over the past two decades there have been many examples of
results and insights in the Theory of Algorithms and Complexity that are motivated by,
and inform, important research fronts in other sciences: Quantum Computing,
Markov chain Monte Carlo, and Algorithmic Game Theory are
examples. This project is about computational research
whose purpose is to shed light to a broad front of central problems in
Game Theory and Economics, Networking, and Biology.
Aspects of this work involve devising algorithms for approximating
Nash equilibria, or identifying computational impediments to the task;
understanding better the power and limitations of learning and
distributed algorithms for computing equilibria; exploring from the
algorithmic standpoint the refinements of the Nash equilibrium concept, as
well as of equilibrium selection processes, which have been
developed by researchers in Game Theory over the past fifty years;
coming up new, compelling, and computationally motivated solution concepts;
identifying complexity theoretic impediments to the central problem of characterizing
the optimal multi object auction, as well as developing algorithms and lower
bounds for the problem of computing price equilibria in both the
most general setting and in novel special cases of markets.
Further goals of this project in the realm of Networking include
studying a recently identified class of models known as ``networks of games;
determining the extent to which sophisticated pricing schemes
can improve the efficiency of selfish routing; and the analyzing two new
genres of network models: One for social
networks, and one for financial markets. In Biology, a recent result
established that natural selection under recombination optimizes not fitness,
as it had been assumed for decades, but a novel metric which we call ``mixability;
this project shall follow on this work by devising techniques for proving mathematically
this effect, as well as gauging rigorously its impact on Evolution, and and by exploring
the relationship between recombination, mixability, and genetic modularity.

This project is likely to produce new insights into both Computation
and the target disciplines (Game Theory, Networking, Theory of Evolution),
as well as new mathematical techniques. In Algorithmic Game Theory,
it seeks to make progress on some of the deepest and most looked at problems, but also to
identify new exciting research fronts and directions in this field.
In the Theory of Evolution, the effort is to shed light on some of
the most fundamental and old questions of this important discipline.
The insights from this work will be used in the development of graduate and
undergraduate courses. A substantial part of this project aims at
understanding and improving the global information environment (the
Internet, the worldwide web, and the digital social networks they
enable), which is one of Humankind s most valuable resources. TC: Medium: Collaborative Research: Experience Based Access Management (EBAM) for Hospital Information Technology Insufficient attention has been given to enterprise Identity and Access Management (IAM) as a process that needs to be carried out on a continuing basis in the presence of change and evolution. In particular, there is little formal support for how IAM can exploit experience the enterprise collects over time. This project is developing a lifecycle model of IAM called Experience Based Access Management (EBAM) that provides a set of models, techniques, and tools to reconcile differences between the ideal access model, as judged by high level enterprise, professional, and legal standards, and the enforced access control, specific to the operational IAM system. The principal component of an EBAM support system is an expected access model that is used to represent differences between the ideal and enforced models based on information collected from access logs and other operational information. The project is developing and validating an approach to the expected model based on using probabilistic information to inform the design of access rules. The project focuses on EBAM for hospital information systems since these are an especially important class of enterprise systems that present diverse and interesting challenges but also provide potential insight into similar issues in other types of enterprise IAM systems. The team consists of specialists in cyber security, biomedical informatics, and a physician who serves as chief medical information officer within a major hospital system. The project will demonstrate how analysis of clinical experience can address gaps between ideal and enforced access control models in a representative hospital. TC: Medium: Dissemination and Analysis of Private Network Data The goal of this research project is to enable statistical analysis and knowledge discovery on networks without violating the privacy of participating entities. Network data sets record the structure of computer, communication, social, or organizational networks, but they often contain highly sensitive information about individuals. The availability of network data is crucial for analyzing, modeling, and predicting the behavior of networks.

The team s approach is based on model based generation of synthetic data, in which a model of the network is released under strong privacy conditions and samples from that model are studied directly by analysts. Output perturbation techniques are used to privately compute the parameters of popular network models. The resulting noisy model parameters are released, satisfying a strong, quantifiable privacy guarantee, but still preserving key properties of the networks. Analysts can use the released models to sample individual networks or to reason about properties of the implied ensemble of networks.

By synthesizing versions of networks that would otherwise remain hidden, this research can advance the study of topics such as disease transmission, network resiliency, and fraud detection. The project will result in publicly available privacy tools, a repository for derived models and sample networks, and contributions to workforce development in the field of information assurance. The experimental research is linked to educational efforts including undergraduate involvement in research through a Research Experience for Undergraduates site, as well as interdisciplinary seminars.

For further information see the project web site at the URL:
http://dbgroup.cs.umass.edu/private network data RI: Medium: Quantifying Causality in Distributed Spatial Temporal Brain Networks A key hurdle in studies of brain function is to be able to measure not only what signals are correlated with one another, but also how they are causally related. Correlation quantifies linear dependence, while causality is capable of distinguishing which brain area is leading the correlated counterparts; causality puts an arrow into correlation. Causality is a difficult problem in data analysis and here a novel measure of conditional statistical dependence to evaluate causality is proposed. The ultimate practical goal is to elucidate the principles of cognitive processing and provide online cognitive feedback to human subjects performing complex tasks.

The objective of this project is to use a recently developed paradigm for electroencephalogram (EEG) quantification based on periodic visual stimulation to improve the signal to noise ratio of visual stimulation on a pre determined EEG frequency band (here around 10 Hz). The goal is to develop advanced signal processing techniques based on instantaneous frequency (Hilbert transform) to quantify the instantaneous amplitude of a visual stimulus in 32 channels over the scalp.

A recently developed measure of local statistical dependence in the joint space called correntropy will be utilized to evaluate the dependency among instantaneous amplitude time series collected over the scalp. The maximum value of correntropy is a measure of statistical dependence, which is the first step towards causality. To achieve a causality measure, conditional dependence will be evaluated by extending correntropy to conditional correntropy, first for triplets of variables and them to subspaces of arbitrary dimensions. Correntropy is a nonparametric measure of dependence; hence, the new method will be compared to linear and nonlinear Granger causality methods implemented in reproducing kernel Hilbert spaces.

These algorithms will be tested on data collected from human subjects in a study of affective visual perception. The goal is to study and quantify the re entry hypothesis of emotional perception that re entrant modulation originating from higher order cortices is responsible for enhanced activation in the occipital cortex when emotionally arousing stimuli are perceived. The signal processing and statistical methods developed here will provide a way to identify dependent EEG channels and causal relationships amongst them during the presentation of the stimulus, effectively tracing the flow of neural activity from the stimulated visual areas to frontal areas and back to the visual cortex. TC: Medium: Collaborative Research: Pay as you Go Security and Privacy for Integrated Transportation Payment Systems Pay as you Go investigates security and privacy for Integrated Transportation Payment Systems (ITPS). The research addresses integrated payments for trains, subways, buses, ferries, and recharging of electric cars, as well as toll collection for roads, bridges, and tunnels. Multi disciplinary aspects include novel cryptographic protocols and lightweight implementations of privacy preserving payment systems. Challenges include providing security and privacy in a low cost, usable, and reliable manner. Payments for transportation differ significantly from general purpose e commerce in terms of both security and engineering constraints. For instance, transportation operators wish to observe user behavior to improve the overall performance of the system seemingly contradicting privacy. Moreover, payment devices must be extremely cheap, mass produced, and tolerant of a wide range of passenger demands.
The expected results and activities include: (1) design of novel cryptographic algorithms achieving privacy at low cost while retaining the benefits of meaningful data collection for ITPS; (2) implementation of hardware and software to perform modern cryptographic operations on low cost devices; and (3) testing of human factors and performance under realistic conditions that must balance security and privacy with cost and usability. The broader impacts are extensive as transportation systems are a critical part of most aspects of society, from the economy, to defense, to public safety. Moreover, information gathered from ITPS applications can facilitate advanced traffic management, travel time estimation, emergency management, congestion pricing, carbon emissions control, and environmental justice assessments. A substantial outreach plan builds on the research team?s significant experience working with transportation professionals in industry and government. III: Medium: Collaborative Research: Linguistically Based ASL Sign Recognition as a Structured Multivariate Learning Problem The manifestation of language in space poses special challenges for computer based recognition. Prior approaches to sign recognition have not leveraged knowledge of linguistic structures and constraints, in part because of limitations in the computational models employed. In addition, they have focused on the recognition of limited classes of signs. No system exists that can recognize signs of all morphophonological types or that can even discriminate among these in continuous signing. Through integration of several computational approaches, informed by knowledge of linguistic properties of manual signs, and supported by a large existing linguistically annotated corpus, the team will develop a robust, comprehensive framework for sign recognition from video streams of natural, continuous signing. Fundamental differences in the linguistic structure of signs, distinguishing signed languages in 4D, with spatio temporal dependencies and multiple production channels from spoken languages, are critical to computer based recognition. This is because finger spelled items, lexical signs, and classifier constructions, e.g., require different recognition strategies. Linguistic properties will be leveraged here for (i) segmentation and categorization of significantly different types of signs, and then, although this subsequent enterprise will necessarily be limited in scope within the project period, (ii) recognition of the segmented sign sequences. Through the 3D hand pose estimation from a team developed tracker, w significant tracking accuracy, robustness, and computational efficiency will be attained. This 3D information is expected to greatly improve the recognition results, as compared with recognition schemes using only 2D information. The 3D estimated information from the tracking will be used in the proposed hierarchical Conditional Random Field (CRF) based recognition, to allow for tracking and recognition of signs that are distinct in their linguistic composition. Since other signed languages also rely on a very similar sign typology, this technology will be readily extensible to computer based recognition of other signed languages.

This linguistically based hierarchical framework for ASL sign recognition?based on techniques with direct applicability to other signed languages, as well?provides, for the first time, a way to model and analyze the discrete and continuous aspects of signing, also enabling appropriate recognition strategies to be applied to signs with linguistically different composition. This approach will also allow the future integration of the discrete and continuous aspects of facial gestures with manual signing, to further improve computer based modeling and analysis of ASL. The lack of such a framework has held back sign language recognition and generation. Advances in this area will, in turn, have far ranging benefits for Universal Access and improved communication with the Deaf. Further applications of this technology include automated recognition and analysis by computer of non verbal communication in general, security applications, human computer interfaces, and virtual and augmented reality. In fact, these techniques have potential utility for any human centered applications with continuous and discrete aspects. The proposed approach will offer ways to address similar problems in other domains characterized by multidimensional and complex spatio temporal data that require the incorporation of domain knowledge. The products of this research, including software, videos, and annotations, will be made publicly available for use in research and education. AF: Medium: New Directions in Computational Complexity Studies in computational complexity in three directions are proposed:
holographic algorithms, Darwinian evolution, and multicore algorithms.
In the first of these areas, holographic reductions have been shown to
be a fruitful source of new efficient algorithms for certain problems,
and evidence of intractability for othrs. In this research the aim is to
arrive at a better understanding of the possibilities and limitations of
holographic algorithms, by exploring ways in which specific currently
known limitations of this class of methods can be circumvented. For
evolution the goal is to understand better what classes of mechanisms
can evolve through the Darwinian processes of variation and selection
when only feasible resources in terms of population sizes and numbers of
generations are available. In the area of multi core algorithms, a
methodology will be developed for expressing and analyzing parallel
algorithms that are optimal for a wide range of hardware performance
parameters. Such algorithms would make possible portable software, that
is aware of the parameters of the machine on which it executes, and can
run efficiently on all such machines.

The work on multi core algorithms aims to have the practical goal of
increasing the effective exploitation of multi core computers as these
become more pervasive. The work on evolution will highlight the fact
that the question of how complex mechanisms could have evolved within
the resources available, is a question that is resolvable by the methods
of computational complexity, and aims to provide more precise
mathematical specifications of what the Darwinian process can achieve.
The work on holographic algorithms aims to make progress in our
understanding of what are widely regarded as the most fundamental
questions regarding the power of practical computation. CIF: Medium: Collaborative Research: Explicit Codes for Efficient Operation of Wireless Networks This project deals with models of wireless multi terminal networks incorporating practical constraints such as individual links that experience fading, applications that are delay sensitive, network communication that is subject to broadcast and interference constraints and nodes that are constrained to operate in half duplex mode. The network is assumed to be static for the duration of the message, but can change from one message to the next and channel state information is assumed to be present only at the receiver. In such settings, cooperative communication in which intermediate nodes facilitate communication between a particular source sink pair, is key to efficient operation of the network.

A key goal of any communication system, is one of achieving an optimal rate reliability tradeoff. The diversity multiplexing gain tradeoff (DMT) determines the tradeoff between relevant first order approximations to the rate and reliability of communication. The DMT of point to point communication links has been extensively studied and signal sets are available that are optimal under any statistical distribution of the fading channel. There now exist protocols and codes for two hop relay networks that come close to achieving the corresponding min cut upper bound on DMT. Goals of this project include: 1) determining the DMT of various classes of multiterminal networks ranging from broadcast, cooperative broadcast and multiple access channel networks to layered multi hop networks; 2) identifying the classes of networks for which the DMT of the network is given by the DMT of the min cut; 3) assessing the impact of asynchronous operation of the network, as well as of the presence of feedback along one or more links in the network; 4) the construction of codes with lesser decoding complexity. TC: Medium: Security and Privacy Preserving Data Mining and Management for Disctributed Domains A fundamental but challenging issue in information security is secure sharing and management of sensitive data and information among numerous organizations that form large scale e enterprises. Today, an increasing number of enterprises are using the Internet for managing and sharing users? and enterprise information through online databases. However, security and privacy of data is an overriding concern currently limiting the proliferation of information technology. Among others, the two key security issues are; (a) interoperability challenge of diverse security policies exercised by collaborating organizations for sharing sensitive information and; (b) using collaborative knowledge for detecting and responding to any emerging security or emergency threats. The objective of this project to develop a scalable multi domain information security framework which: (a) facilitates large scale integration, mining, and analysis of multimedia data residing in multiple domains for identifying sensitive information that require controlled dissemination (b) allows distributed real time content analysis of multimedia data streams across multiple domains for correlating detectable events, (c)l provides integration and evolution of inter domain access control policies for distributed multimedia applications, and (d) ensures protection of user?s information. The project uses several innovative approaches to meet these objectives. First, novel real time data mining and stream correlation methods is used for extracting useful content and detecting events embedded in data streams by using multiple sliding window operators. Second, an efficient privacy/security based multimedia data clustering mechanism allows a prevailing security policy to be adaptable to the changing context and the datasets. The uniqueness of this approach is the adaptive nature of security policies. In addition, a novel XML based integrated mechanism is used for encoding of multimedia data in conjunction with multi domain security policies for ensuring technical viability and portability of the system.
The proposed research is expected to have direct and long term impact on developing secure information infrastructures, such as e commerce, digital libraries, and healthcare systems, which are projected to have an important role at a global scale for the next several decades. The research results can help in alleviating the heavy financial risks associated with information and system security and in general, meeting the national information infrastructure security needs and threat management. The research will be pursued in collaboration with Cyber Center, the Center for Education and Research in Information Assurance and Security (CERIAS) and PURVAC facilities at Purdue University with which the PI?s have strong affiliation and on going collaboration. Results and tools developed will be made available over the Web to research community. In The team has a track record of providing their research results and tools to various research communities in the areas of multimedia databases and information security management.


For further information see the project web site at the
URL: http://cobweb.ecn.purdue.edu/~dmultlab/ SHF: Medium: MEDITA Multi Layer Enterprise Wide Dynamic Information Flow Tracking and Assurance Enterprise Information Systems (EIS) continually face attacks ranging from data leaks to the spread of malware; these attacks cost companies billions of dollars annually and can result in critical loss or leakage of data. Existing defenses typically either attempt to secure the hosts within the enterprise or add a security perimeter to the network. These conventional defenses are ineffective in the face of compromised hosts, mobile devices, and insider threats. Dynamic Information Flow Tracking (DIFT) techniques maintain data provenance information about objects within the system and control information flow by defining and implementing policies that dictate how that information should be allowed to flow. Although powerful, existing DIFT approaches are limited by the fact of targeting only a single layer on a single physical host, which limits their effectiveness and practical applicability.

This research will develop MEDITA, a multi layer DIFT mechanism that can precisely, securely, and efficiently track data flowing within a networked EIS and across layers, and control the flow of such data based on the data provenance and the security policy in place. Multi layer DIFT holds great promise for controlling information flow within an enterprise in many real world scenarios. Despite its appeal, however, realizing a system that could implement such DIFT policies in practice is extremely challenging because of the wide variety of attacks that can be mounted, ranging from copying and pasting the sensitive data to writing the document to removable storage or a mobile device. To address these and other challenges, this research will (1) refine existing techniques for performing DIFT within the individual layers of an EIS, (2) design and implement the integration and inter operation of DIFT techniques between layers, (3) define a language that can be used to express multi layer security policies for the EIS and mechanisms for translating those policies to tainting and enforcement mechanisms; and (4) Develop a prototype implementation of MEDITA and perform experiments by using the prototype to apply MEDITA to realistic information flow tracking control scenarios. AF: Medium: Algorithms Based on Algebraic and Combinatorial Methods Many advanced combinatorial problems have algebraic aspects. Even though the problem formulation can be entirely discrete, significant insight and efficient algorithms might be obtained by applying sophisticated algebraic methods. It is not uncommon that combinatorial problems have simple and elegant formulations, yet they are computationally hard, meaning that the obvious algorithms are useless for their solutions except for very small instances. It can also happen that even though traditional algorithmic approaches are successful, algebraic methods are still more efficient and provide additional insights into a combinatorial problem.

The graph isomorphism problem exemplifies a combinatorial problem where algebraic methods seem to be required for efficient solutions. Interesting for algebraic and combinatorial approaches is also the monomer dimer problem, the counting of matchings in grid graphs, which is of much importance in statistical physics. This proposal studies algorithms based on the scaling method. A particular goal is doing matrix scaling efficiently in parallel, as a tool for approximating the permanent.

This project will look at the computation of all coefficients of graph polynomials for trees and graphs of bounded tree width. The goal is to compute all coefficients together almost as fast as a single coefficient.

The other main focus of this project is the exploration of variations of the recent faster integer multiplication algorithm and the study of its application to polynomial multiplications and Fourier transforms. One goal is to develop a new algorithm, based on a more discrete method, improving the asymptotic complexity as well as leading to a more practical algorithm for computing products of very long integers.

Integer multiplication is such a fundamental arithmetic task that understanding and improving it is an obvious basic intellectual challenge. Such theoretical goals are foremost in this project. But there could be an impact on the search for Mersenne primes as well as on general purpose computations with high degree polynomials. Other aspects of this research involve topics with applications in Physics and Chemistry. SHF: Medium: How Do Static Analysis Tools Affect End User Quality The perceived quality of a software product depends strongly on field failures viz., defects experienced by users after the software is released to the field. Software managers work to constantly improve quality control processes, seeking to reduce the number of field failures. Static analysis is a powerful and elegant technique that finds defects without running code, by reasoning about what the program does when executed. It has been incubating in academia and is now emerging in industry. This research asks this question: How can the performance and practical use of static analysis tools be improved ? The goal of the research is to find ways to improve the performance of static analysis tools, as well as the quality control processes that use them. This will help commercial and open source organizations make more effective use static analysis tools, and substantially reduce field failures.

Using historical data from several open source and commercial exemplars, the research will retrospectively evaluate the association of field failures with static analysis warnings. The research will evaluate the impact of factors such as experience of the developer, the complexity of the code, and the type of static analysis warning on failure properties such criticality, and defect latency (time until a defect becomes a failure). A wide variety of projects will be studied, including both commercial and open source. The resulting data will be analyzed using statistical modeling to determine the factors that influence the success of static analysis tools in preventing field failures. Some field failures may have no associated static analysis warnings. This research will identify and characterize these failures, paving the way for new static analysis research. An integrated educational initiative in this proposal is the training of undergraduates by using bug fixes as pedagogical material; undergraduates will also help annotate the corpus of field failures with information relevant to our analysis. An important byproduct of this research, is a large, diverse, annotated corpus of field failures of use to other educators and researchers in empirical software engineering, testing, and static analysis.