User Cluster-based Unmanned Aerial Vehicle Swarm Mission Planning Strategy
In order to solve the communication command problem caused by"power outage,network outage,and circuit outage"caused by natural disasters,a unmanned aerial vehicle assisted network system is proposed,which loads popular access content of ground users through airborne edge server and issues emergency notification messages.Based on ground service requirements,the un-manned aerial vehicle introduces a Q-Learning reinforcement learning algorithm for trajectory planning and divides the entire system into a unmanned aerial vehicle exploration subsystem and a unmanned aerial vehicle service subsystem.The unmanned aerial vehicle explo-ration subsystem aims at the problem that signal coverage area of a single unmanned aerial vehicle is limited and it is impossible to col-lect the positions of all mobile devices in the area.A boundary exploration method is created to determine the optimal number of un-manned aerial vehicle groups.The unmanned aerial vehicle service subsystem determines the cluster center by performing K-means,K-medoids,and AGNES clustering on the mobile device respectively,and uses the cluster center as the guide to carry out trajectory plan-ning so as to maximize the service life of mobile device.Simulation results show that the proposed unmanned aerial vehicle auxiliary sys-tem is feasible to design,and the minimum configuration quantity for full coverage of unmanned aerial vehicle swarm signals is deter-mined,and the applicable scenarios of unmanned aerial vehicle swarms with different clustering algorithms are obtained.Research results can be used to evaluate the support service experience of ground users and to provide a basis for the architecture design and con-trol optimization of unmanned aerial vehicle assisted communication networks.
国家科技期刊平台
NSTL
万方数据
