Research on UAV-Assisted Computation Offloading Based on PER-SAC

Unmanned Aerial Vehicles (UAVs) are becoming more popular in Mobile Edge Computing (MEC). However, UAVs face problems such as high task processing delay, high energy consumption, and the lack of fairness in service during the service process. This paper investigates a UAV-assisted mobile edge computing system which uses the reinforcement learning algorithm to minimize the task processing delay and the energy consumption of the UAV while maximizing the fairness of the user service. Firstly, a Markov decision process is formulated for this system. Then, the SAC algorithm is improved by employing the prioritized experience replay method based on instantaneous rewards and TD errors, and the PER-SAC algorithm is employed to tackle this problem. The simulation demonstrate that the proposed algorithm converges rapidly and outperforms other algorithms.