Real-time Trust-Building Schemes for Mitigating Malicious Behaviors in Connected and Automated Vehicles

Abstract

Vehicle-to-everything (V2X) communication has raised security and safety concerns. A malicious attacker can not only have a compromised vehicle broadcast spoofed positioning information to nearby vehicles, but also send fake emergency reports to traffic management systems. Previous works on mitigating such malicious behaviors focus on either developing cryptography mechanisms for identity authentication or conducting plausibility checking by individual vehicle to evaluate the trust of messages sent by others. This paper proposes two types of architectures (centralized and distributed) for mitigating malicious behaviors based on the collaboration between trust authorities and vehicles. They differ from previous approaches in that the trust authority and vehicle nodes collaborate but take different responsibilities in plausibility checking and maintaining trust in order to build trust on each vehicle in real-time, rather than only relying on the judgment from each individual vehicle. Also, the trust value maintained by the trust authority in our approach is for revoking certificates of vehicle nodes based on pre-defined patterns of trust changes. The pros and cons of each architecture are evaluated through an experimental simulation. In addition, the proposed architectures for trust-building can be extended in future works to encourage benign behaviors in the long run.