A Modified Off-grid SBL Channel Estimation and Transmission Strategy for RIS-Assisted Wireless Communication Systems
Mengnan Jian, Yajun Zhao
Abstract
In this paper, we design a compressed sensing based Uplink/Downlink (UL/DL) channel estimation (CE) scheme for RIS-aided terahertz MIMO systems and develop an integral transmission strategy. Specifically, we adopt the off-grid sparse Bayesian learning (SBL) approach to work directly on the continuous AOA/AOD domain and avoid the severe grid mismatch. Compared with most state-of-the-art algorithms, i.e., LS, MMSE and on-grid compressive sensing (CS) approaches, the proposed CE method achieves better channel estimation accuracy. The angle-domain reciprocity is exploited to obtain a much simplified overall transmission scheme with significantly reduced training and feedback overhead. A novel frame structure is developed to successfully handle the RIS-aided transmission problem. Moreover, a multi-user downlink strategy is developed where the limited scattering nature of terahertz channel is exploited and a region-separation aided user grouping algorithm is explained. Simulation results are provided to demonstrate the superior performance of the proposed method over existing ones.