基于移动用户群的多无人机覆盖在线学习算法  

Online Learning Algorithm of Dynamic Multi-UAV Communication Coverage for Mobile Ground User Groups

作　　者：张金辉[1] 芦方旭 米志超 王穆阳 ZHANG Jinhui;LU Fangxu;MI Zhichao;WANG Muyang(Logistic Support Center,PLA General Hospital,Beijing 100853,China;Unit 31121 of PLA,Nanjing Jiangsu 210042,China;College of Communications Engineering,Army Engineering University of PLA,Nanjing Jiangsu 210007,China)

机构地区：[1]解放军总医院服务保障中心,北京100853 [2]中国人民解放军31121部队,江苏南京210042 [3]陆军工程大学通信工程学院,江苏南京210007

出　　处：《通信技术》2022年第3期305-311,共7页Communications Technology

摘　　要：地面移动用户群使得传统的静态无人机覆盖算法无法取得较好的效果,因此在综合考虑无人机的覆盖性能与功耗性能的基础上,采用动态多无人机方法,提出了一种基于最大奖励函数值的在线学习覆盖算法。该算法在空间建模无人机的运动方向后,以无人机收到的最大奖励函数值为激励函数,在线学习选择下一步的运动方向,确定下一步的运动位置。仿真结果表明,该算法在实现移动无人机的覆盖和功耗性能指标上具有显著效果。Mobile user groups on the ground make the traditional static UAV coverage algorithm unable to achieve good results. Therefore, on the basis of comprehensively considering the coverage performance and power consumption performance of UAVs, through adopting a dynamic multi-UAV method, this paper proposes an online learning coverage algorithm based on the maximum reward function value. After modeling the movement direction of the UAV in space, the algorithm takes the maximum reward function value received by the UAV as the excitation function, and then learns online to select the next movement direction and determine the next movement position. Simulation results indicate that the algorithm has a significant effect in achieving the best coverage and power consumption performance of mobile UAVs.

关 键 词：移动用户群 多无人机 通信覆盖 在线学习 

分 类 号：TN929.52[电子电信—通信与信息系统]