Efficiency Optimization for Blockchain-Enabled V2V Energy Trading With Dynamic Clustering Based on Deep Reinforcement Learning

Vehicle-to-Vehicle (V2V) energy trading is a feasible solution to alleviate charging anxiety and enhance the range of electric vehicles (EVs). Clustering EVs can improve the efficiency of both energy transfer and V2V communication. However, in the complex large-scale Internet of Electric Vehicles (IoEV), the security and reliability of trading and communication, as well as the impact of cluster number and strategy on trading efficiency, require further discussion. This paper proposes a V2V energy trading system that leverages blockchain sharding and dynamic clustering to securely record energy trades as transactions on the blockchain. The system forms clusters using a clustering algorithm based on the mobility information and power levels of EVs, ensuring the reachability of energy trading. These trading clusters correspond to blockchain shards, enhancing transaction throughput through sharding scalability. Deep reinforcement learning (DRL) is employed to optimize the number of clusters, clustering algorithms, and parameters of the sharded blockchain system under security and latency constraints. The results demonstrate that the proposed scheme ensures timely power replenishment for low-power vehicles, improves overall economic utility and throughput for participants, and is effectively applicable to large-scale V2V energy trading scenarios.