Projection Approximation Subspace Tracking-Based Channel Estimation for IRS Control

This paper proposes a projection approximation subspace tracking (PAST)-based channel estimation for time-varying channels in intelligent reflecting surface (IRS)-assisted communication systems. Conventional pilot-based estimation requires an excessive number of pilot signals proportional to the number of IRS elements, making it difficult to track time-varying channels. To address this issue, we propose a novel pilot allocation, reflection coefficient transition, and channel subspace tracking via the PAST algorithm. Numerical analysis conducted in a 28 GHz band multi-path environment demonstrated that this approach successfully suppresses the degradation of the spectral efficiency while achieving an approximate 25% reduction in pilot overhead compared to conventional methods.