A Low‐Pilot‐Overhead Two‐Stage Channel Estimation Scheme for XL‐RIS‐Assisted Multi‐User Terahertz Systems

ABSTRACT Accurate channel state information (CSI) is essential for exploiting terahertz communication and reconfigurable intelligent surfaces (RISs) in sixth‐generation (6G) wireless systems. However, CSI acquisition in RIS‐assisted terahertz systems requires prohibitive pilot overhead because the cascaded channel has high spatial degrees of freedom and the RIS contains a large number of passive reflecting elements. To address this challenge, we propose a low‐pilot‐overhead two‐stage channel estimation scheme for RIS‐assisted multi‐user terahertz systems. In the first stage, a reliability‐based rule selects a typical user, and a denoising network estimates its cascaded channel to extract the common BS‐RIS channel information shared by different users. In the second stage, a single‐link channel estimation algorithm estimates the channels of the remaining users by exploiting the correlation among users' cascaded channels. Specifically, only one RF link is activated to estimate the channel scaling factors between the typical user and the other users. A single‐link denoising network is further designed to suppress scaling‐factor errors and improve channel reconstruction accuracy. Simulation results show that the proposed scheme achieves higher channel estimation accuracy and spectral efficiency than benchmark methods while requiring fewer pilot symbols.