Research On Resource Scheduling Mechanism Of Mobile Edge Computing Based On UAV

Mobile edge computing(MEC)can solve the problem of insufficient computing resources for user equipment(UE).Nevertheless,in the special scenes such as mountain area and post disaster,the Base Station(BS)and MEC devices are difficult to be deployed.The UAV(unmanned aerial vehicle)carries communication circuit and MEC equipment to fly or hover over the area where UEs are located,which can assist UE to complete data collection and real-time calculation,and can also be used as relay node to help UE use MEC services provided by BS.However,UAV-assisted MEC systems still face many technical challenges and theoretical topics.For UAV flight path planning,flight parameter and data transmission power control of ground terminal,there are some problems to be solved,such as system delay and energy consumption optimization.This paper aims to minimize the task completion time,reduce the delay in the system and improve the user experience by studying the resource scheduling mechanisms in UAVassisted MEC systems,such as task scheduling,association between UE and UAV hovering positions,and UAV trajectory.The proposed model in this paper allows UAV to either provide lightweight MEC services for UE or as a relay node to forward the collected data from the UE to the BS for execution,and to consider priority order dependencies for tasks in the application.The contents of this paper are as follows:Study 1: in a single UE scenario,a mathematical model for minimizing task completion time is established by jointly optimizing task scheduling and UAV hovering position.In this paper,the heuristic algorithm based on particle swarm optimization(PSO)is used to search the optimal hover position of UAV,and then an effective iterative algorithm is proposed to find the optimal solution of the problem.Through a series of simulation and performance comparison,this paper reveals the influence of task size,number of task groups and UAV maximum flight speed on task completion time in the application,and provides insights on how to schedule resources in multi-UE scenario.Study 2: for the UAV assisted MEC system in multi UE scenarios,this paper optimizes the association between UE and UAV hover position,the service sequence of each UAV hover position to the associated UE,and how to divide tasks,so as to minimize the task completion time of all UE applications.Since the established UAV hover time optimization problem is difficult to find the optimal solution to the twostage flow-shop scheduling problem.An algorithm is proposed in this paper for solving UE association and task division,and uses the Johnson algorithm to solve the service sequence optimization problem in this two-stage flow-shop scheduling.A series of simulation experiments show that the proposed algorithm can always complete all tasks with the shortest time compared with other baseline scheduling schemes,further verifying the performance of the proposed algorithm.