A UAV-assisted dynamic offloading based on maximum clique algorithm with weighted graphs in mobile edge computing

With the increase of user devices (UE), traditional mobile edge computing (MEC) has struggled to provide adequate resources in high-density areas. To provide a reliable and flexible offloading service, unmanned aerial vehicles (UAVs) equipped with edge servers have been introduced to assist MECs. However, existing studies tend to focus on a single metric and ignore the ability of UAVs to coordinate with each other. Therefore, a UAV-assisted MEC system unloading scheme is proposed in this paper to solve the deployment problem of multiple UAVs and multiple UEs. To solve the problem of limited resources and low energy efficiency of MEC system in UAV deployment, Maximum Clique with Weighted Graph Algorithm (MCWG) was proposed. The algorithm optimizes the HOTSPOT UAV deployment algorithm, calculates the UAV deployment position in the weighted graph, minimizes the average distance between ue and UAV, and maximizes the average CPU cycle of ue. The simulation results show that compared with other benchmark algorithms, the MCWG algorithm can effectively shorten the communication distance and calculation task length during the deployment of UAVS, and improve the resource utilization rate of UAVs by 34%.