A node trust evaluation method of vehicle-road-cloud collaborative system based on federated learning

As the vehicle-road-cloud collaboration system develops rapidly, it is accompanied by serious information se-curity problems while solving the data transmission issues. For constructing secure transmission of data, trust is recommended as a relevant way to accomplish network security; that is, developing a trust model that can be used by sensor nodes to determine the reliability of another node is crucial. However, the heterogeneity of the network has different functional requirements for trust evaluation, and the openness of the network makes the nodes more vulnerable to attacks. Therefore, the research of trust evaluation model in the vehicle-road-cloud collaborative system is facing greater challenges than the traditional network. In this paper, a trust evaluation scheme of the vehicle-road-cloud collaborative system based on Federated Learning (FLT) is proposed. A hier-archical trust evaluation model is designed, and the complex model is simplified to an orderly hierarchical structure by using hierarchical analysis. The trust indexes of different layers are evaluated, and the influencing factors among different nodes are comprehensively considered. Combined with federated learning, it solves the problem of finding the most reliable route and realizes personalization at the level of equipment, data, and model. For the purpose of alleviating the heterogeneity and obtaining a high-quality personalized model for each device, trust values can be adaptively updated as changes in the topology of the network occur in real-time. The simulation findings demonstrate that, as compared to previous schemes, the energy consumption is lowered by 35%, and the accuracy is raised by 45% while maintaining trust stability.