Joint Spectrum Allocation and Power Control in Vehicular Networks Based on Reinforcement Learning

In this paper, we investigate the joint channel al-location and power control problem in vehicular networks. Considering the different quality-of-service (QoS) requirements for vehicle-to-infrastructure (V2I) and vehicle-to-vehicle (V2V) links, we transform the optimization problem using reinforcement learning (RL) and then propose a distributed resource allocation scheme based on the deep Q network (DQN) and deep determin-istic policy gradient (DDPG), which enables joint optimization of continuous power control and discrete channel allocation. Additionally, we consider the reward fluctuation caused by the strong dynamics of vehicular networks, and propose the advantage reward to alleviate this instability. Simulation results demonstrate that the proposed DQN-DDPG based resource allocation algorithm improves both the total capacity of V2I links and the payload delivery rate of V2V links, achieving higher QoS satisfaction compared to other baselines.