RAVEN: Resource Allocation Using Reinforcement Learning for Vehicular Edge Computing Networks

Vehicular Edge Computing (VEC) enables vehicles to offload tasks to the road side units (RSUs) to improve the task performance and user experience. However, blindly offloading the vehicle's tasks might not be an efficient solution. Such a scheme may overload the resources available at the RSU, increase the number of requests rejected, and decrease the system utility by engaging more servers than required. This letter proposes a Markov Decision Process based Reinforcement Learning (RL) method to allocate resources at the RSU. The RL algorithm aims to train the RSU in optimizing its resource allocation by varying the resource allocation scheme according to the total task demands generated by the traffic. The results demonstrate the effectiveness of the proposed method.