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.
作者
赵冬鹤
徐龙艳
翟亚红
ZHAO Donghe;XU Longyan;ZHAI Yahong(School of Electrical and Information Engineering,Hubei University of Automotive Technology,Shiyan Hubei 442002,China)
出处
《重庆科技学院学报（自然科学版）》
CAS
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)。
