基于DDPG算法的MISO通信系统优化  

Optimization of MISO Communication System Based on DDPG Algorithm

作　　者：赵冬鹤 徐龙艳[1] 翟亚红[1] ZHAO Donghe;XU Longyan;ZHAI Yahong(School of Electrical and Information Engineering,Hubei University of Automotive Technology,Shiyan Hubei 442002,China)

机构地区：[1]湖北汽车工业学院电气与信息工程学院,湖北十堰442002

出　　处：《重庆科技学院学报（自然科学版）》2024年第2期82-87,共6页Journal of Chongqing University of Science and Technology：Natural Sciences Edition

基　　金：湖北省教育厅科研计划重点项目“智能工厂的5G∕WiFi网络多域共存关键技术研究”(D20211802);湖北省科技厅重点研发计划项目“应用于商用车制造及检测的3D机器视觉及机器人集成技术开发”(2022BEC008)。

摘　　要：无线通信系统中部署大量天线时会增加硬件成本,且能耗和信号处理的复杂性加大,针对此问题提出了一种基于深度确定性策略梯度(DDPG)算法的多输入单输出(MISO)通信系统优化方案。研究智能反射面(RIS)辅助多用户的MISO系统,在连续状态和动作背景下,利用DDPG算法联合设计基站端波束成形矩阵和RIS处相移,以最大化系统加权和率。仿真实验结果表明,该方案与FP算法的加权和率性能相当且收敛性较好,相比未部署RIS的方案具有明显优势。A multi-input-single-output(MISO)communication system optimization scheme based on deep deter-ministic policy gradient(DDPG)algorithm is proposed to address the issues,such as,increased hardware cost,increased energy consumption,signal processing complexity,etc.caused by deploying a large number of antennas in wireless communication systems.The reconfigurable Intelligent Surface(RIS)assisted multi-user MISO systems is studied and the beamforming matrix at the base station end and the phase shift at the RIS are jointly designed u-sing the DDPG algorithm to maximize the weighted sum rate of the system in the context of continuous states and ac-tions.Simulation experimental results show that the scheme has comparable and better convergence of sum-rate performance with the FP algorithm,and has a clear advantage over the scheme without RIS deployment.

关 键 词：DDPG算法 智能反射面 波束成形 多输入单输出系统 

分 类 号：TN929.5[电子电信—通信与信息系统]