Intelligent Framework for Multi-UAV-Enabled Joint Offloading Service and Trajectory Planning in MEC

Unmanned aerial vehicles (UAVs) have a pivotal role in mobile edge computing (MEC) applications. In a UAV-based MEC framework, data generated by user equipments (UEs) on the ground can be offloaded to servers carried by the UAVs for expedited processing. Nonetheless, the diversity of computational tasks leads to uneven data distribution across UEs. Additionally, the energy consumption of UAVs for hovering and movement must be carefully managed to ensure sustainable flight and effective data analysis and processing. In this work, we tackle the NP-hard problem associated with integer service association in MEC systems by developing a potential game-based model that efficiently resolves the service association between UAVs and UEs. Building on this strategy, we develop a Grey Wolf Optimization (GWO) based trajectory planning method to explore the flight paths of the UAVs. This intelligent framework ensures a balanced and efficient operation of the MEC system, addressing key challenges in service association during task offloading and UAV trajectory planning. Simulation results demonstrate that our proposed MUJST not only ensures efficient convergence but also minimizes the offloading time for UEs and reduces UAV energy consumption.