Resource Allocation Algorithm of Unmanned Aerial Vehicles Assisted Mobile Edge Computing Based on Deep Reinforcement Learning

In this paper, we propose an unmanned aerial vehicle (UAV) assisted multi-user mobile edge computing system, where UAVs and base station (BS) transmit data jointly. We aim to minimize user delay and energy consumption by optimizing the UAV trajectory optimization and user power allocation. To solve this problem, a deep reinforcement learning resource algorithm is adopted. Firstly, we use multi-agent deep reinforcement learning to optimize the trajectory of UAVs. Secondly, we dynamically allocate user transmission power to improve network transmission speed. The simulation results show that this algorithm enables UAVs to track user movements, significantly reducing system latency and energy consumption, and providing users with better communication services.