.png){kind=icon}
.png){kind=icon}
Return
F-CP-RALS Algorithm for Channel Estimation in RIS-Aided mmWave MIMO Systems
.png){kind=icon}
PRE
AI
DOI：10.1109/tgcn.2026.3693090.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Accurate channel estimation remains a pivotal challenge in reconfigurable intelligent surface (RIS)-aided millimeter wave (mmWave) multiple-input multiple-output (MIMO) systems for next-generation wireless networks, particularly in dense urban scenarios where ultra-high data rates, massive connectivity, and seamless coverage are pursued. In this paper, a multiple Frobenius norms regularized alternating least squares (RALS) algorithm is proposed based on CANDECOMP/PARAFAC (CP) decomposition to jointly estimate all channel matrices, termed as F-CP-RALS. Firstly, the received signals from both cascaded and direct links are modeled as a third-order tensor based on a structured time-domain protocol, which inherently preserves the multi-dimensional channel structure and hidden space correlation. Subsequently, all links are jointly estimated by the proposed algorithm which can stabilize convergence and enhance noise suppression to improve estimation accuracy due to the additional regularization terms. Moreover, the algorithm utilizes an alternating iterative process to refine the channel matrices without turning off the RIS, thereby reducing the pilot overhead compared to methods that rely on separate estimations of direct and cascaded links. Additionally, a rigorous convergence analysis is conducted to guarantee algorithmic convergence, and the feasibility conditions are investigated by leveraging Kruskal’s uniqueness conditions, followed by a computational complexity analysis. Finally, experimental simulations confirm that the presented approach delivers more accurate estimation results than alternative methods.
Keywords:
Channel estimation
CP decomposition
RIS
MIMO
pilot overhead
Journal
I
IF:
6.7
Papers:
1.3K
Citations:
4.3K