Privacy-Preserving Data Sharing in IoV: A Federated Learning and Blockchain-Based Approach
Zhuoqun Xia, Jiahao Sun, Jingjing Tan · 2024
With the development of in-vehicle sensors and communication infrastructures, vehicles facilitate intelligent transportation systems (ITS) and traffic flow management by enabling information exchange between vehicles and between vehicles and infrastructures through data sharing in Internet of Vehicles (IoV). Meanwhile, these data sharing schemes also pose security and privacy protection issues for the sharers. Traditional data encryption and access control algorithms are difficult to address data integrity issues and centralized characteristics. Additionally, those encryption and decryption algorithms require large amounts of computational and storage resources. To address these issues, we propose a privacy-preserving data sharing scheme based on blockchain and federated learning, where blockchain technology is used for authentication and data sharing, and asynchronous federated learning allows participants to train local models on local devices to reduce the risk of privacy leakage. Moreover, we screen out unrelated vehicles by calculating the similarity of vehicle trajectories to select participants. Experimental data and analysis validate the effectiveness and scalability of the proposed scheme.