Robust Channel Estimation for IRS-Assisted Millimeter-Wave MIMO Systems

In this paper, the downlink channel estimation of the intelligent reflecting surface (IRS) assisted MIMO systems is considered in the millimeter-wave (mmWave) band. It is challenging to acquire the channel state information (CSI) of the IRS-assisted wireless communication systems, due to the large amount of passive reflecting elements at the IRS. On account of the low-rank property of the mmWave channels, a bivariate optimization problem, which balances the minimization of the fitting errors and the kernel norms, is formulated to address the estimation of the cascaded channel. By leveraging the alternating optimization method, the problem is decoupled into two subproblems, each of which is solved with the alternating direction method of multipliers (ADMM) together with the singular value thresholding (SVT) technique. In comparison with its counterparts, the proposed method provides higher estimation accuracy and stronger resistance to the uncertainty of the ranks of the channel state information matrices.