Deep Reinforcement Learning-Based Resource Management for Flexible Mobile Edge Computing: Architectures, Applications, and Research Issues

In this article, we introduce autonomous vehicle-assisted mobile edge computing (AV-MEC), including unmanned ground vehicle (UGV)- and unmanned aerial vehicle (UAV)-assisted MEC, where the UAV/UGV can be deployed and carry the computing server to serve ground mobile devices (MDs). We first discuss applications and main research problems. Then, deep reinforcement learning (DRL)-based solutions are introduced, explored, and demonstrated. We also discuss challenges and future research directions for an AV-MEC system with DRL being applied to it.