Research On Task Offloading And Trajectory Scheduling Of UAV-Assisted Wireless Energy Transfer Mobile Edge Computing Network

As a key technology in 5G research,Mobile Edge Computing(MEC)technology is processed by uploading complex computing tasks generated by Mobile Io T Device(MID)to the MEC server near the terminal device for processing.Compared with mobile cloud computing,MEC is a technology which can effectively utilize the computing power of network edge servers,reducing the service delay of MID.However,traditional MEC equipment is installed on fixed infrastructure,and there are certain limitations for some infrastructure-limited and emergency wireless network scenarios(emergency rescue,remote rural areas or military exercise areas).In recent years,Unmanned Aerial Vehicle(UAV)-assisted MEC networks have attracted widespread attention,UAVs can utilize their maneuverability to get closer to each device and reduce transmission delays.On the other hand,considering that MID energy is limited,Wireless Power Transfer(WPT)technology can provide continuous energy supply for MID nodes through radio frequency signals.Based on this,this paper aims to reduce system energy consumption and improve system energy efficiency,and conduct in-depth research on task offloading and trajectory scheduling in UAV-assisted MEC networks in WPT environment.The main contents are summarized as follows:Firstly,for deterministic single-task static task offloading,a joint optimization of MID’s Central Processing Unit(CPU)cycle frequency,offload calculation bits,UAV’s CPU cycle frequency and UAV is proposed.The static offloading and trajectory scheduling scheme of the flight trajectory to minimize the total UAV energy consumption(defined as the sum of UAV computing energy consumption,communication energy consumption with MID,and UAV flight energy consumption).Due to the coupling relationship between the optimization variables,the objective optimization problem is difficult to solve.Therefore,an offloading and resource allocation algorithm based on alternating optimization is proposed.First,the CPU cycle frequency and task offloading of the MID are obtained by using the Lagrangian dual method.The closed-form expressions for the CPU cycle frequency of the bit and UAV are calculated,and the non-convex trajectory problem is then solved using the Successive Convex Approximation(SCA)method.The simulation results show that the algorithm proposed in this paper can converge faster and can obtain lower energy consumption compared with other benchmark algorithms.Secondly,for the dynamic offloading of randomly arriving tasks,a dynamic task offloading and trajectory scheduling scheme that jointly optimizes MID’s CPU cycle frequency,transmission power,time allocation and UAV flight trajectory is proposed to minimize the long-term average energy efficiency of the system(definition is the ratio of long-term total energy consumption to long-term cumulative processing computing tasks).Since the optimization problem is a nonlinear fractional programming problem and the coupling relationship between variables,the objective problem is difficult to solve.Therefore,an online computing offloading and trajectory scheduling algorithm based on the Lyapunov optimization method is proposed.The problem is transformed into a linear fractional programming problem,and then the closed-form expressions for the CPU cycle frequency,transmission power,and time allocation of the MID are obtained using the Lagrangian dual method and the subgradient method,and then the SCA method is used to solve the non-convex trajectory problem.The simulation results show that the algorithm proposed in this paper can achieve a balance between system energy efficiency and service delay.Compared with other schemes,it can obtain better EE-delay performance and can converge faster.