Research On Energy Efficient Offloading Strategy For UAV-Aided Edge Computing Systems

With the rapid development of mobile communication technology,user equipments need to face more complex application requirements than in the past.At the same time,the rapid growth of the number of user equipments makes it difficult for the traditional "cloud computing + fixed base station" network coverage model to carry huge-scale application requirements.UAVs have attracted much attention due to their strong maneuverability,good maneuverability,and lower deployment cost compared with ground base stations.They can quickly improve communication coverage.Based on the idea of mobile edge computing,the academic community has proposed the use of UAVs equipped with edge servers to assist user equipments in task calculations to meet user application requirements.At present,most of the relevant research work is focused on the maximum throughput problem in the UAV-assisted edge computing process or the energy consumption of the UAV,and the research on the energy consumption of the user equipment is still incomplete.This thesis focuses on an energy-saving offloading strategy for UAV-assisted user equipment edge computing and the multi-UAV collaborative computing offloading mechanism.The main contributions are as follows:(1)Aiming at the system model of UAV-assisted edge computing,the mobile edge computing technology and the air-ground collaborative networking system based on UAV are introduced respectively.Based on the characteristics of both,a system model of UAV-assisted edge computing is established.Secondly,it introduces and compares the common mathematical algorithms for solving system models.(2)Aiming at the energy consumption of user equipments under single-UAV-assisted edge computing,a system model of single-UAV-assisted multi-user edge computing is constructed.The model adopts a partial offloading strategy,which fully considers the user’s computing bits and delay,as well as the energy constraints and trajectory of the UAV to minimize the overall energy consumption of user equipments.In order to solve this model,a two-step iterative optimization algorithm based on block coordinate descent is proposed,which jointly optimizes the bit allocation for local computing and offload computing and UAV’s trajectory to minimize the energy consumption within the agreed time.Simulation results show that the proposed strategy is suitable for different channel conditions,and is superior to other benchmark schemes.(3)Aiming at the problems of user equipments energy consumption and multi-UAV cooperation mechanism under multi-UAV-assisted edge computing,a system model of multi-UAV-assisted multiuser edge computing is constructed.Aiming at the characteristics of multiple UAVs and multiple users,this model additionally considers the problem of offloading matching between multiple UAVs and multiple users,as well as the safety distance between UAVs.In order to solve this model,a three-step iterative optimization algorithm based on block coordinate descent is proposed.By decomposing nonconvex problems and solving by fixed variables in turn,it has achieved better performance than the single UAV-assisted edge computing system,and further reduced the energy consumption of the user equipments.