Detection of malicious data in vehicular ad-hoc networks for traffic signal control applications
For traffic signal control applications relying on vehicular ad-hoc networks, this method addresses the security threat of Sybil attacks, but the results are simulation-based and incremental.
The paper tackles the problem of malicious data injection in vehicular ad-hoc networks for traffic signal control. It proposes a method combining driver behavior modeling with position verification to detect Sybil attacks, and simulation experiments show it mitigates negative impacts on traffic signal control performance.
Effective applications of vehicular ad hoc networks in traffic signal control require new methods for detection of malicious data. Injection of malicious data can result in significantly decreased performance of such applications, increased vehicle delays, fuel consumption, congestion, or even safety threats. This paper introduces a method, which combines a model of expected driver behaviour with position verification in order to detect the malicious data injected by vehicle nodes that perform Sybil attacks. Effectiveness of this approach was demonstrated in simulation experiments for a decentralized self-organizing system that controls the traffic signals at multiple intersections in an urban road network. Experimental results show that the proposed method is useful for mitigating the negative impact of malicious data on the performance of traffic signal control.