Secure Data Sharing for Autonomous Vehicles in Mobile Blockchain Networks

As autonomous vehicle development progresses, data security has emerged as a critical concern due to threats emanating from multiple directions. Traditional centralized server-based data sharing approaches no longer suffice due to issues such as personal privacy leaks, location information tampering, and computing and storage loads. In this article, we first discuss key challenges in existing data sharing approaches for autonomous vehicles. To address these challenges, we propose a distributed data sharing approach that combines blockchain, edge computing, and federated learning. Specifically, blockchain is utilized to securely store data from autonomous vehicles across multiple nodes, while edge computing accelerates data processing and analysis, particularly reducing risks associated with data transmission. Federated learning is integrated to preserve privacy by enabling decentralized model training without sharing raw data, thus mitigating privacy concerns inherent in centralized approaches. Afterwards, we present a case study on road segmentation for autonomous vehicles to demonstrate the security benefits of the proposed distributed data sharing approach. We finally identify some future directions on data sharing of autonomous vehicles. Our research provides a feasible solution to the data security challenges facing the development and commercialization of autonomous vehicles.