Energy-Efficient Resource Allocation in UAV-Assisted MEC Systems via Two-Stage Alternating Optimization

The rapid proliferation of ground user devices (GUDs) has posed significant challenges to the computational capacity of traditional Mobile Edge Computing (MEC) systems. As the number of user devices continues to grow, the demand for system computing resources increases accordingly. Although the integration of Unmanned Aerial Vehicles (UAVs) into MEC systems can partially alleviate this bottleneck, it also leads to a substantial rise in overall system energy consumption. To minimize the energy consumption of UAV-assisted MEC system, a two-stage alternating optimization algorithm is proposed. In the first stage, given fixed transmission power, the UAV's trajectory, CPU frequency allocation, and offloading time are optimized using successive convex approximation (SCA). In the second stage, offloading power allocation is optimized with the UAV trajectory fixed. This iterative process continues until convergence to a suboptimal solution. Simulation results show that the proposed method outperforms fixed power, fixed trajectory, and benchmark schemes, achieving significant energy savings for both UAV and GUDs while ensuring rapid convergence.