Semi-Blind Joint Channel Estimation and Symbol Detection for RIS-Empowered Multiuser mmWave Systems
Jianhe Du, Xin Luo, Xingwang Li, Mingfu Zhu, Khaled M. Rabie, Ferdi Kara
Abstract
In this letter, we propose a semi-blind joint channel estimation and symbol detection scheme for reconfigurable intelligent surface (RIS)-empowered multiuser millimeter wave (mmWave) systems. Combined with the coding scheme at user equipments (UEs) and RIS reflection coefficient design, we prove that the received signals at the base station (BS) follow a PARATUCK2 tensor model, and then a two-stage fitting algorithm is derived by exploiting the low-rank structure of mmWave channel. Without a dedicated training stage, the proposed scheme can jointly detect information symbols of all UEs and estimate the channels of the UEs-RIS and RIS-BS links. In comparison to the existing methods, the proposed system can increase spectrum efficiency and obtain better channel estimation and symbol detection performance. Numerical results are presented to verify the effectiveness of the proposed scheme.