Multi-UAV Collaborative Computing and Task Offloading Strategy Based on Non-orthogonal Multiple Access
In crowded application scenarios,there is a large demand for computing resources and unequal distribution of un-manned aerial vehicle computing tasks.For this kind of scenario,this paper presents a multi-UAV-assisted mobile edge computing task uninstallation scheme based on non-orthogonal multiple access.First,the non-orthogonal multiple access and serial interfer-ence deletion technologies are used to increase the transmission rate of users,and the cooperation of multiple UAVs is used to pre-vent uneven task distribution.Under the constraints of energy consumption and computing resources of user equipment,a non-con-vex optimization problem,the minimization of system energy consumption,is established through joint optimization of user uninstal-lation decision and unmanned deployment location.Subsequently,the optimization issue is divided into two subsections,the DDQN network is used to determine the user's uninstallation decision,the differential evolution algorithm is used to determine the un-manned aerial vehicle deployment under the uninstallation decision,and the two methods are iterated alternately to obtain the opti-mal solution of the problem.The simulation results show that non-orthogonal multiple access effectively reduces the transmission de-lay of user tasks compared with time-division multiple access technology.Compared with DQN network and greedy algorithm,the offload decision algorithm proposed in this paper reduces the system's total energy consumption.
mobile edge computingunmanned aerial vehicledeep reinforcement learningnon-orthogonal multiple ac-cessdifferential evolution
国家科技期刊平台
NSTL
万方数据
