Reliability Sensitivity Analyses

Published: 16 December 2021| Version 1 | DOI: 10.17632/jnsgkc23xs.1
Contributor:
Inés Álvarez

Description

We want to show how Time-Sensitive Networking (TSN)-based networks can benefit from using time redundancy to tolerate temporary faults in the links. Specifically, we want to prove that we can increase the reliability of said networks using our Proactive Transmission of Replicated Frames (PTRF) mechanism to tolerate temporary faults in the links of the network (thoroughly described in https://www.mdpi.com/1424-8220/21/3/756). To that, we model the behaviour of TSN and of two different approaches to implement PTRF, namely approach A and B, using the PRISM probabilistic model checker. We have used our models to quantify the reliability and to study the impact that different network parameters have on it. In this dataset we show the results obtained from the parametric sensitivity analysis of the reliability of TSN, PTRF approach A and PTRF approach B in the presence of temporary faults. The parameters considered in our models are the following: - Bit error rate (BER): 1E-4 to 1E-12 - Frame size: 64, 782 and 1500 bytes - Number of frame replicas: 1 to 4 - Number of bridges in the communication path: 2 to 10 We executed an experiment for each combination of parameters for TSN, PTRF approach A and PTRF approach B, which resulted in 2,146 experiments. In each one of these experiments we modeled the operation of the network during 10 hours, which resulted in the modeling of the transmission of 1,800,000 frames. Our results allowed us to draw the following conclusions: -The BER is determinant in the reliability achievable by TSN, PTRF approach A and PTRF approach B. PTRF approach A is a more adequate solution than approach B to provide high reliability in networks with low BERs. On the contrary, approach B can reach significantly higher levels of reliability than approach A in networks with high BERs. - The frame size has an important impact on reliability, which increases significantly as the BER increases. - We have seen that the reliability decreases as the network size increases. We have also seen that this negative impact increases with the BER. - We can state that these analyses prove that we can increase the reliability of networks based on TSN by using proactive time redundancy to tolerate temporary faults in the links. The dataset contains the results of all the experiments. Specifically, they show the UNRELIABILITY (probability of the communication failing within 10 hours) for TSN, PTRF approach A and PTRF approach B, each in a different sheet. The BER is indicated in the top left of each table and below we find the frame size in bytes. k1 to k4 represent the number of replicas while b1 to b10 represent the number of bridges in the path.

Files

Institutions

Universitat de les Illes Balears

Categories

Applied Sciences, Informatics

Licence