Exploring Real-Time Malicious Behaviour Detection in VANETs

Vehicular ad hoc networks (VANETs) are a fundamental component of intelligent transportation systems, using real-time data from inter-connected vehicles to provide us faster and safer trips. However, these benefits raise a number of security challenges that need to be solved. This paper is proposing a real time malicious behaviour detection system that uses statistical and machine learning methods to detect malicious vehicles involved in Distributed Denial-of-Service (DDoS) attacks. The proposed system builds on prior work and improves on the state of the art by conducting evaluations using data from simulated VANET networks that are realistic, cover scenarios with different street topologies and can work in real-time. The proposed system first detects that an attack is ongoing and then determines the malicious vehicles responsible for the attack. Evaluations demonstrate that the proposed system is able to detect the onset of an attack within one second of its commencement with an F1-score of 100% and detects the malicious vehicles involved, on the average, with an F1-score of approximately 90%.