Blind Channel Estimation for Reconfigurable Intelligent Surface Assisted OFDM Systems

This study presents a novel subspace-based blind channel estimation method for reconfigurable intelligent surface (RIS) assisted orthogonal frequency division multiplexing systems. In existing research on RIS control, reflection characteristics of RIS is assumed to be fixed during data transmission to maximize data rates. However, this fixed reflection pattern undermines the uniqueness of the estimation result by the subspace-based estimation methods. To address this issue, a new data transmission scheme is proposed, comprising an implicit channel estimation phase followed by a rate maximization phase. In the initial phase, the reflection pattern dynamically varies to realize blind channel estimation without any training symbols transmission. Subsequently, in the rate maximization phase, the optimal reflection pattern is determined based on the estimated channel obtained during the first phase. Computational simulations demonstrate the effectiveness of the proposed method in accurately estimating the channel without the need for training signals.