Deep Reinforcement Learning-Based Resource Allocation for Cellular V2X Communications

Vehicle-to-everything (V2X) communication is an essential technology for future vehicular applications. It is challenging to simultaneously achieve vehicle-to-infrastructure (V2I) and vehicle-to-vehicle (V2V) communications, given the shared spectrum. Deep reinforcement learning (DRL)-based algorithms have been proposed for resource allocation in V2I and V2V designs. Existing DRL designs focus on the objectives of high-capacity V2I and high-reliability V2V links. In this study, a multi-agent DRL algorithm is proposed to maximize the sum capacity of V2I links while ensuring capacity fairness among the V2V links. The simulation results demonstrate the balance between the V2I–V2V objectives achieved by the proposed algorithm.