Blockchain-Enhanced Federated Learning for Internet of Vehicles
Muhammad Fudholi Hamidul As’ad, Safa Otoum · 2024
The integration of the Internet of Vehicles (IoV) into urban environments introduces significant data security and network efficiency challenges. As vehicular networks expand, their susceptibility to cyber-attacks increases, demanding effective security solutions. This paper proposes a novel framework by integrating blockchain technology with federated learning (FL) to secure and enhance the efficiency of IoV systems. In this context, we propose a Blockchain-Enhanced FL for Internet of Vehicles (BEFL-IoV) framework designed to optimize communication efficiency and protect data privacy within IoV networks. Our approach distributes data processing tasks across vehicles, reducing latency and network congestion while ensuring high data privacy and integrity. Blockchain provides a decentralized, tamper-resistant layer for secure transactions and data exchanges between vehicles and infrastructure. Simulation results confirm the framework’s efficacy in maintaining high data accuracy under various cyber threats. Specifically, the model achieves 0.94% accuracy against property inference attacks and 0.92% accuracy against membership inference attacks. This integration of blockchain and FL significantly improves the scalability and reliability of IoV applications, marking a major advancement in smart transportation technologies.