Misbehavior Detection in VANET Based on Federated Learning and Blockchain

As an irreversible trend, connected vehicles become increasingly more popular. They depend on the generation and sharing of data between vehicles to improve safety and efficiency of the transportation system. However, due to the open nature of the vehicle network, dishonest and misbehaving vehicles may exist in the vehicular network. Misbehavior detection has been studied using machine learning in recent years. Existing misbehavior detection approaches require network equipment with powerful computing capabilities to constantly train and update sophisticated network models, which reduces the efficiency of the misbehavior detection system due to limited resources and untimely model updates. In this paper, we propose a new federated learning scheme based on blockchain, which can reduce resource utilization while ensuring data security and privacy. Further, we also design a blockchain-based reward mechanism for participants by automatically executing smart contracts. Common data falsification attacks are studied in this paper, and the experimental results show that our proposed scheme is feasible and effective.