A Privacy-Preserving Misbehavior Detection System in Vehicular Communication Networks
Sohan Gyawali, Yi Qian, Rose Qingyang Hu
Abstract
5 G based vehicular communication networks support various traffic safety and infotainment use cases and rely on the periodic exchange of information. However, these messages are susceptible to several attacks which can be detected using misbehavior detection systems (MDS). MDS utilizes trust score, feedback score and other evaluation schemes to identify abnormal behavior of the vehicles. However, the trust and feedback scores used in MDS may violate the location, trajectory, or identity privacy of the vehicle. In this paper, we propose a privacy-preserving misbehavior detection system that can detect or identify misbehavior without violating the privacy of the vehicle. In the proposed method, encrypted weighted feedbacks sent from vehicles are combined using additive homomorphic properties without violating the privacy of the information. The decryption of the aggregate feedback is done securely at the trusted authority which updates the reputation score of the vehicle according to the decrypted aggregate feedback score. We have also performed comprehensive security analysis and have shown the correctness and resilience of the proposed schemes against various attacks. In addition, we have done extensive performance analysis and have shown that the computation cost of the proposed scheme is better compared to the existing schemes.