基于稀疏角域的6G信道估计智能反射面算法  

RIS algorithm for 6G channel estimation based on sparse angle domain

作　　者：朱文杰 伏胜洋 姚雨晨 ZHU Wenjie;FU Shengyang;YAO Yuchen(School of Computer and Information Engineering,Hubei Normal University,Huangshi 435002,China)

机构地区：[1]湖北师范大学计算机与信息工程学院,湖北黄石435000

出　　处：《湖北师范大学学报（自然科学版）》2024年第2期72-78,共7页Journal of Hubei Normal University：Natural Science

基　　金：国家自然科学基金(62172144);湖北省教育厅中青年人才科技项目(Q20192501);湖北师范大学人才引进项目(1000002)资助课题;2023年研究生科研创新项目(2023Z101)。

摘　　要：智能反射面(Reconfigurable Intelligent Surface,RIS)有着操纵性强、能耗低、方便部署等优势,已成为6G(第六代移动通信)的关键技术。研究了关于智能反射面的信道估计算法,对于传统算法在传播时路径会因角度原因发生偏移的情况,采用了Tucker分解的稀疏角度域高阶奇异值分解(High Order Singular Value Decomposition,HOSVD)信道估计算法来解决路径偏移的问题。为了验证所提出算法的鲁棒性,对比了传统的交替最小二乘法信道估计算法,可以得到不管是在用户数量、传播的路径偏移上都能取得比传统的信道估计算法更好的归一化最小均方误差(Normalized Mean Squared Error,NMSE)效果。Reconfigurable Intelligent Surface(RIS)has become a key technology of 6G(sixth generation mobile communication),due to its strong maneuverability,low energy consumption and easy deployment.Based on the study of the channel estimation algorithm for intelligent reflective surfaces,this paper uses and the High Order Singular Value Decomposition(HOSVD)channel estimation algorithm based on Tucker decomposition to solve the problem of path offset,considering that the path will be shifted due to angle during propagation of traditional algorithms.In order to verify the robustness of the proposed algorithm,it compares HOSVD with the traditional alternating least squares channel estimation algorithm.The result shows that the normalized mean squared error(NMSE)effect is better than the traditional channel estimation algorithm in terms of the number of users and the path offset of propagation.

关 键 词：智能反射面 稀疏角域 Tucker分解 归一化最小均方误差 信道估计 

分 类 号：TN295[电子电信—物理电子学]