Continuously Weighted Majority Voting for Detection of Data Tampering Attacks in Redundant CACC Systems

Cybersecurity emerges as a critical concern in the realm of cooperative adaptive cruise control (CACC) over wireless networks. Various approaches, including model-and redundancy-based strategies, have been suggested to identify data tampering attacks within CACC systems employing redundant communication paths. Our previous study integrated these two approaches, presenting an attack detection method that utilizes discretely weighted majority voting (DWMV) based on estimation error. Despite its efficacy, DWMV falls short in detecting attacks when faced with simultaneous occurrences of modeling errors and multipath attacks. The current study introduces a novel attack detection method employing continuously weighted majority voting (CWMV) to address this limitation. The proposed CWMV establishes the weight through an exponential function involving the estimation error and decay ratio. Through simulations, we validate that CWMV excels in detecting data tampering attacks, even in scenarios where modeling errors and multipath attacks unfold concurrently in a leader-follower CACC system equipped with three redundant feedback paths.