Fair and Efficient Resource Allocation via Vehicle-Edge Cooperation in 5G-V2X Networks

To cope with safety risks and operational efficiency problems, it is of paramount importance to ensure high data rates and meet the latency requirements in Connected and Autonomous Vehicles. The problem in such environments is two-fold: 1. Heavy load on the network due to increasing demands; 2) Resource imbalance, due to variations in the vehicular traffic density in certain regions. The consequences of these two phenomena may lead to service disruptions, as well as the fairness of resource allocation across vehicles. In this work, we propose a resource allocation method that distributes high workload among edge nodes and allocates network resources efficiently and fairly. Performance of the proposed method is evaluated under realistic scenarios, and compared to the state-of-the-art approaches in the literature. Behavior and performance of all methods in overload conditions in certain regions were analyzed. Simulation results exhibit a 38% improvement in the successful demand rate and a 48% improvement in capacity usage compared to state-of-the-art approaches.