摘要
当蜂窝连接无人机(Unmanned Aerial Vehicle,UAV)与地面用户(Ground User,GUE)共存时,彼此之间会产生严重的空地干扰.针对这一问题,在网络辅助全双工(Network Assisted Full Duplex,NAFD)无蜂窝大规模多输入多输出(Multiple Input Multiple Output,MIMO)系统中研究了一种资源优化方案,该方案具有协作能力且允许上下行传输同时进行,可以提升系统频谱效率(Spectral Efficiency,SE),但引入了交叉链路干扰.构建了一个同时最大化上行和SE(Sum SE,SSE)以及下行SSE的多目标优化问题(Multi-Objective Optimization Problem,MOOP),并提出了一种基于深度Q网络(Deep Q Network,DQN)的资源优化算法,通过联合优化接入点(Access Point,AP)的工作模式以及用户的传输功率,以较低复杂度提升系统全局SSE.仿真结果表明,所提算法能够在复杂约束条件下有效降低场景中性能较差用户的比例,相比于固定功率的传输方案有更高的系统SSE,间接性地缓解了 UAV和GUE间的空地干扰问题,提升了系统整体性能.
Abstract
When cellular-connected Unmanned Aerial Vehicle(UAV)coexists with Ground User(GUE),there is severe air-ground interference with each other.To solve this problem,a resource allocation optimization scheme in cell-free massive Multiple Input Multiple Output(MIMO)system with Network Assisted Full Duplex(NAFD)is investigated.This system has cooperation ability and enables uplink and downlink transmission simultaneously,which can improve system Spectral Efficiency(SE),while also introduces cross link interference.A Multi-Objective Optimization Problem(MOOP)is formulated to simultaneously maximize the uplink Sum SE(SSE)and the downlink SSE.A Deep Q Network(DQN)-based resource optimization algorithm is proposed,which jointly optimizes the operation mode of the Access Point(AP)and the transmission power of the users to improve the global SSE of the system with low complexity.The simulation results show that the proposed algorithm can effectively reduce the proportion of poor performance users in scenario under complex constraints.Compared with the fixed power transmission scheme,the proposed algorithm has higher system SSE,indirectly alleviates the air-ground interference between UAV and GUE,and improves the overall performance of the system.
国家科技期刊平台
NSTL
万方数据