Date of Award
5-16-2024
Publication Type
Thesis
Degree Name
M.Sc.
Department
Computer Science
Keywords
ITS (Intelligent Transportation System);Machine Learning;Sybil Attack;VANET;VeReMi
Supervisor
Arunita Jaekel
Abstract
Integrating Vehicular Ad-Hoc Networks (VANETs) into modern Intelligent Transportation Systems (ITS) introduces critical security concerns. This research addresses the emerging threat of Traffic Congestion Sybil Attacks, where malicious entities inject spurious data to fabricate artificial traffic congestion. The methodology involves a detailed examination of the VeReMi dataset, a benchmark for VANET research, coupled with state-of-the-art classification machine learning algorithms. The analysis includes training and evaluating these models to identify patterns indicative of Grid Sybil attacks. Preliminary results obtained through meticulous testing demonstrate a substantial enhancement in various classification metrics, showcasing promising improvements, especially in enhancing the recall value for accurately identifying maliciously induced traffic congestions. These initial findings underscore the potential for further refinements and heightened classification metrics in subsequent phases of the research. This thesis emphasizes the urgency of securing VANETs against Traffic Congestion Sybil Attacks, presenting an innovative solution through the fusion of machine learning techniques and the VeReMi dataset. The outcomes contribute to theoretical understanding and hold practical implications for enhancing the security of vehicular communication networks.
Recommended Citation
Khanduja, Sarthak, "Traffic Congestion Sybil Attack Detection in VANETs using Machine Learning Techniques" (2024). Electronic Theses and Dissertations. 9470.
https://scholar.uwindsor.ca/etd/9470