Queue-Aware Computation Efficient Optimization for MEC-Assisted Aerial-Terrestrial Network

Unmanned aerial vehicle (UAV)-assisted mobile edge computing (MEC) system is a prominent strategy, where a UAV equipped with a MEC server is deployed to serve on ground terminal devices. This paper considers a multi-UAV-assisted network in which multiple UAVs are deployed to provide MEC services to terrestrial users. The objective is to maximize the queue-aware computation efficiency of an aerial-terrestrial network by jointly optimizing task splitting, task offloading and MEC server selection, UAV trajectory, and CPU frequency allocation. The work utilizes Dinkelbach’s method and Lyapunov optimization to reformulate the defined problem. Moreover, an alternating iterative approach based on the block descent method is proposed to solve this mixed-integer problem. Simulation results are presented to show that the proposed approach outperforms various benchmark schemes.