摘要
无人机作为空中基站,不仅可为地面用户提供移动通信服务,还可预先存储热点内容,有效提升了边缘通信能力。对无人机位置部署、无人机缓存配置与用户接入进行联合优化可大幅提升网络性能,然而传统优化方法计算量庞大且收敛速度慢。为解决上述问题,将该联合优化问题分解成无人机编队部署子问题和缓存接入子问题,设计了一种集中式和分布式相结合的快速资源分配优化方法。针对部署子问题,分阶段进行集中式一致博弈建模和分布式势能博弈建模,提出了一种基于较优响应的分阶段部署算法;针对缓存接入子问题,构建“用户-内容”虚拟体,将缓存与接入的联合问题建模为虚拟体与无人机缓存空间的匹配问题,提出了一种基于匹配的虚拟体接入算法。仿真结果表明,所提方法和传统分布式较优响应方法相比,具有更快的收敛速度和更高的全局性能。
Unmanned aerial vehicles(UAVs),functioning as aerial base stations,not only can deliver mobile communication services to ground users but also can proactively store high-demand content to effectively enhance edge communication capabilities.The optimization of UAV deployment,cache configuration,and user access selection in tandem can significantly enhance the network performance.However,conventional optimization methods suffer from considerable computational complexity and sluggish convergence rates.To tackle this challenge,the joint optimization problem is decomposed into two distinct subproblems:UAV formation deployment and cache configuration/user access.A hybrid centralized and distributed rapid resource allocation optimization method is designed.For the first sub-problem,a phased approach is employed.UAV deployment is modeled as both a centralized uniform game and a distributed potential game,leading to the development of a staged deployment algorithm based on superior responses.Regarding the second sub-problem,a"user-content"virtual entity is created,transforming the original problem into a matching issue between these virtual entities and UAV cache locations.Subsequently,a matching-based virtual entity access algorithm is proposed.The simulation results unequivocally demonstrate that the proposed method,in comparison with traditional distributed superior response methods,achieves quicker convergence and superior overall performance.
作者
钟天尧
吴杜成
陈润丰
王路广
陈学强
ZHONG Tianyao;WU Ducheng;CHEN Runfeng;WANG Luguang;CHEN Xueqiang(College of Communications Engineering,Army Engineering University of PLA,Nanjing 210007,China)
出处
《陆军工程大学学报》
2024年第1期19-27,共9页
Journal of Army Engineering University of PLA
