无蜂窝大规模MIMO中基于深度强化学习的无人机辅助通信与资源调度

UAV Assisted Communication and Resource Scheduling in Cell-free Massive MIMO Based on Deep Reinforcement Learning Approach

无蜂窝大规模多入多出(MIMO)网络中分布式接入点(AP)同时服务多个用户,可以实现较大区域内虚拟MIMO的大容量传输;而无人机辅助通信能够为该目标区域热点或边缘用户提供覆盖增强。为了降低反馈链路负载,并有效提升无人机辅助通信的频谱利用率,该文研究了基于AP功率分配、无人机服务区选择和接入用户选择的联合调度;首先将AP功率分配和无人机服务区选择问题联合建模为双动作马尔可夫决策过程(DAMDP),提出了基于Q-learning和卷积神经网络(CNN)的深度强化学习(DRL)算法;然后将用户调度构造为一个0-1优化问题,并分解成子问题来求解。仿真结果表明,该文提出的基于DRL的资源调度方案与现有方案相比,可以有效提升无蜂窝大规模MIMO网络中频谱利用率。

Distributed Access Points(AP)in the cell-free massive Multiple Input Multiple Output(MIMO)networks serve multiple users at the same time,which can achieve large-capacity transmission of virtual MIMO in a larger area.Unmanned Aerial Vehicle(UAV)assisted communication can provide coverage enhancement for hotspots or edge users in this area.In order to improve the spectrum efficiency and reduce the feedback overhead,a joint resource scheduling scheme that includes AP power allocation,UAV service zone selection and user scheduling is proposed in this paper.Firstly,the AP power allocation and the UAV service zone selection problems are jointly modeled as a Double-Action Markov Decision Process(DAMDP).Then,a Deep Reinforcement Learning(DRL)algorithm based on Q-learning and Convolutional Neural Networks(CNN)is proposed.Furthermore,the user scheduling problem is formulated as a 0-1 optimization problem and solved by dividing into sub-problems.Simulation results demonstrate that the proposed DRL-based resource scheduling scheme exhibits a higher spectrum efficiency than existing schemes.