Dataset and Code for IoT-Based WISNE-SDN Detection and DDOS Attack mitigation using machine learning techniques
Description
There are using two datasets, one is UNSW_2018_IoT_Botnet_Final_10_best_Training.csv and dataset_sdn.csv, the dataset is, 1. UNSW_2018_IoT_Botnet_Final_10_best_Training.csv is a refined training dataset derived from the larger UNSW 2018 IoT Botnet dataset. It focuses on the ten most significant features selected through feature selection techniques, enabling efficient and effective model training. The dataset is designed for intrusion detection and includes both benign and malicious traffic, particularly emphasizing botnet activity in IoT environments. Each record consists of traffic flow attributes such as packet counts, byte sizes, durations, and statistical measures relevant to flow behavior. This selection aims to balance model accuracy and computational cost, making it suitable for building lightweight machine learning models for real-time IoT threat detection. The dataset reflects realistic IoT traffic scenarios by incorporating attacks like DDoS, information theft, and reconnaissance, captured using various tools and emulation techniques. 2. dataset_sdn.csv is a dataset structured for evaluating security mechanisms in Software Defined Networking (SDN) environments. It contains network flow records that represent both normal and attack traffic, enabling classification and detection tasks. The dataset is typically generated in a controlled SDN testbed and includes features like source and destination Session ID, ports, protocols, packet lengths, flow durations, and flags. Attack types might include DDoS, probing, and spoofing attacks, reflecting vulnerabilities specific to SDN-based networks. The dataset is often used to test anomaly-based intrusion detection systems in SDN controllers, where traffic patterns can be analyzed and malicious behavior flagged using machine learning techniques. It serves as a crucial resource for research in SDN security and intelligent traffic management. and also developing codes added in this datas.