双RIS辅助MIMO系统混合张量信道估计算法  

A Hybrid Tensor Channel Estimation Algorithm for Double-RIS Assisted MIMO Systems

作　　者：李双志[1] 邢益博 雷豪杰 LI Shuangzhi;XING Yibo;LEI Haojie(School of Electrical and Information Engineering,Zhengzhou University,Zhengzhou 450001,China)

机构地区：[1]郑州大学电气与信息工程学院,郑州450001

出　　处：《电讯技术》2025年第3期407-414,共8页Telecommunication Engineering

基　　金：国家自然科学基金青年科学基金项目(61901416);中国博士后科学基金第3批特别资助(站前(2021TQ0304));河南省青年人才托举工程(2024HYTP026)。

摘　　要：为解决在双可重构智能超表面(Reconfigurable Intelligent Surface,RIS)系统中获取高维信道状态信息(Channel State Information,CSI)的挑战,提出了一种基于混合张量分解的多链路联合信道估计算法。首先,通过设计导频传输机制,将单反射链路和双反射链路的接收信号分别建模为平行因子模型和平行因子塔克(Tucker)张量模型,将信道估计问题转化为混合张量因子矩阵的拟合问题。然后,考虑到多条链路之间共享的CSI,采用一种基于交替最小二乘迭代算法来分解混合张量,以有效估计出因子矩阵。最后,通过对该混合张量进行唯一性分析,与传统的Khatri-Rao分解方法相比,所提方法具备更为灵活的参数设计特点。仿真实验结果表明,该方法能够在训练块数小于RIS单元数的情况下有效估计反射链路CSI。To solve the challenge of obtaining high-dimensional channel state information(CSI)in reconfigurable intelligent surface(RIS)systems,a multi-link joint channel estimation algorithm based on hybrid tensor decomposition is proposed.Firstly,by designing a pilot transmission mechanism,the received signals of single reflection links and double reflection links are respectively modeled as parallel factor models and parallel factor and Tucker models,thereby transforming the channel estimation problem into a fitting problem of hybrid tensor factor matrices.Subsequently,considering the shared CSI among multiple links,an alternating least squares iterative algorithm is employed to decompose the hybrid tensor,effectively estimating the factor matrices.Finally,through a uniqueness analysis of the hybrid tensor,compared with the traditional Khatri-Rao decomposition method,this method possesses more flexible parameter design characteristics.Simulation results demonstrate that this method can effectively estimate the reflection link CSI when the number of training blocks is less than the number of RIS units.

关 键 词：多用户MIMO 信道估计 双可重构智能超表面 混合张量分解 

分 类 号：TN929.5[电子电信—通信与信息系统]