A Proof-of-Multiple-State Consensus Mechanism for Mobile Nodes in Internet of Vehicles

Blockchain technology provides a reliable information access environment for the Internet of Vehicles, but the high latency and complex computing consensus mechanism in blockchain make it difficult to port to onboard devices. Recently, there are many methods to reduce the time cost of consensus by optimizing node grouping or reducing redundant calculations, but this would lower the security level of the blockchain. To address these issues and reduce the adverse effects of frequently changing channel quality on consensus results, a consensus mechanism based on vehicle comprehensive state factors for nodes selection (PoMS) is proposed. Firstly, the vehicle nodes utilize the machine learning model to predict local driving parameters and broadcast the predicted results to the other nodes. Secondly, each node uses interactive data to calculate the state values, and the leader comprehensively evaluates the nodes participating in the consensus and selects the nodes as relays. Finally, we also adopted a double-layer blockchain structure to accelerate the selection process of relay nodes. In order to verify the performance of the proposed consensus algorithm, we conducted tests on transmission time and communication quality. The experimental results show that compared to traditional consensus mechanisms, the algorithm proposed in this paper can reduce time overhead by an average of 12.7% and maintain a good transmission rates under a certain number of malicious nodes.