Trajectory, Computation Offloading, and Transmission Optimization Strategy Based on Age of Information in UAV-Assisted MEC System

Mobile edge computing (MEC) technology has deployed edge servers (ES) with powerful computing power to make ES closer to terminal devices (TD), so as to effectively meet the requirements of computing power and delay. However, in addition to strict delay requirements, more and more emerging applications have also put forward higher standards for the freshness and timeliness of the collected information. In order to ensure the freshness and high value of information, we propose a trajectory, computation offloading, and transmission optimization strategy based on age of information (AoI). The strategy considers the average AoI of the system in the trajectory planning stage, the AoI in the transmission stage, and the execution stage, respectively, because it is constrained by the delay and residual energy. Then, a joint optimization model is established based on AoI. In order to solve the strong coupling problem among trajectory, computation offloading, and transmission decision, the original optimization problem is divided into three stages. The solutions of trajectory decision, computation offloading decision, and transmission decision are obtained, respectively, by using the proximal policy optimization (PPO) algorithm, the deep deterministic policy gradient (DDPG) algorithm, and the dueling double deep Q network (D3QN) algorithm. The proposed joint optimization strategy considers the influence of trajectory planning on offloading decision and offloading decision on transmission decision and can adapt to the dynamic change of channel connection between ES and TD caused by unmanned aerial vehicle (UAV) mobility. The experimental results show that, compared with other offloading and transmission strategies, the proposed method has higher overall system revenue and lower AoI.