Secure and Efficient Task‐Offloading for Decentralized Vehicular Communication Networks

Abstract

Recently, the combination of Internet of Vehicles (IoV) and blockchain has emerged as a promising solution for enhancing the security and efficiency in vehicular communication networks. However, the deployment of blockchain technique in IoV inevitably derives additional computation and communication overheads, which significantly hinders the development of IoV. In addition, efficient task offloading in IoV is essential to support computation‐intensive and delay‐sensitive vehicular services under dynamic network conditions. To address the above challenge, this paper proposes a deep reinforcement learning‐based joint task‐offloading framework for blockchain‐empowered IoV communication networks. Specifically, it formulates the blockchain‐based task‐offloading problem in IoV as a continuous control Markov decision process, aiming at improving long‐term system performances by jointly optimizing latency, computational cost, throughput and security. Then, a twin delayed deep deterministic policy gradient‐based algorithm is customized to learn the optimal offloading policy efficiently in high‐dimensional continuous action space. Furthermore, a trust‐aware mechanism is incorporated into the state representation and reward design to mitigate the impact of malicious vehicles. Finally, simulation results demonstrate that the proposed method outperforms conventional baseline methods with respect to communication latency, computational cost, throughput and security.