An EAADE: Effective Authentication Approach for Data Exchange in Vehicular Social Network for IoV

ABSTRACT In recent years, in order to protect the privacy of vehicular users, federated learning (FL) is mainly used for secure and reliable data transmission in the vehicular social network. However, there is a vulnerability to data attacks by adversaries when exchanging locally trained model data about the vehicle's trajectory with a federated server. Therefore, in this paper, we have proposed a federated learning‐based mutual authentication protocol to provide security in data transmission. First, we classified the vehicular social community by fixed, moving, and hybrid modes. Second, to develop a novel approach named EAADE (Effective Authentication Approach for Data Exchange). It provides efficient data dissemination in a vehicle social network and security of the main trade‐off between vehicles. Finally, experience it with in‐depth simulation and compare it with other modern protocols. During the analysis of the results, ensure its security against Sybil, replay and falsification attacks, and estimating processing costs and verification times. The outcomes of simulation on the SUMO and OMNeT++ platforms demonstrate that the packet loss rate is lowered to 4.78% and that the verification delay is the better, when compared to other protocols. This protocol has also better transmission cost and implementation costs.