Abstract: In reconfigurable intelligent surface (RIS) assisted communication systems, existing wideband channel estimation schemes only consider the frequency selectivity fading of the channel, and ignore the beam squint problem which degrades the system performance severely. Therefore, a distributed Dice subspace pursuit (DDSP) scheme based on a grouping strategy is proposed in this paper. First, the cascaded channel estimation problem is transformed into a sparse signal recovery problem by exploiting the sparsity of Terahertz (THz) channels. Then, a wideband dictionary is designed and a grouping strategy is proposed to solve the effect of beam squint. Finally, the Dice coefficient criterion and backtracking idea are introduced to optimize the selection of support set. The simulation results show that the proposed algorithm can improve the accuracy of channel estimation and outperforms the traditional wideband channel estimation schemes.

Key words: reconfigurable intelligent surface (RIS), terahertz (THz), wideband channel estimation, beam squint