Deep Reinforcement Learning for Resource Allocation in V2V Communications

In this article, we develop a decentralized resource allocation mechanism for vehicle-to- vehicle (V2V) communications based on deep reinforcement learning. Each V2V link is supported by an autonomous "agent", which makes its decisions to find the optimal sub-band and power level for transmission without requiring or having to wait for global information. Hence, the proposed method is decentralized, with minimum transmission overhead. From the simulation results, each agent can effectively learn how to satisfy the stringent latency constraints on V2V links while minimizing the interference to vehicle-to-infrastructure (V2I) communications.