Enhanced Iterative Hard Thresholding-Based Cascaded Channel Estimation in IRS-Enabled mmWave MIMO Systems

The integration of Intelligent Reflecting Surfaces (IRS) with millimeter-Wave (mmWave) Multiple-Input Multiple-Output (MIMO) architectures presents a compelling strategy for improving coverage and boosting spectral efficiency in next-generation wireless systems. A central difficulty in such networks lies in estimating the cascaded channel that incorporates the base station (BS)–IRS and IRS–User Equipment (UE) links. This task is particularly challenging because IRS elements operate passively and lack inherent baseband processing capability. In this work, we introduce an Enhanced Iterative Hard Thresholding (EIHT)–based framework for cascaded channel estimation that is computationally efficient and well-suited for large antenna arrays. By representing the IRS-assisted propagation path in the angular domain, the estimation problem is formulated as a sparse recovery task. The proposed algorithm refines the solution through residual-guided updates using pseudo-inverse projections, eliminating the need for explicit support detection or prior channel statistics. Numerical evaluations show that across a range of SNR levels, the proposed method achieves lower Normalized Mean Square Error (NMSE) and faster convergence than benchmark approaches such as Orthogonal Matching Pursuit (OMP) and classical Iterative Hard Thresholding (IHT). These results position the method as a practical and scalable option for real-time deployment in IRS-enabled mmWave MIMO networks.