Millimeter Wave Hybrid MIMO System Channel Estimation Using Variable Step Size Zero Attracting LMS

Abstract

Millimeter-wave multiple-input multiple-output (mmWave MIMO) has emerged as a viable technique for 5G and beyond 5G(B5G) wireless networks, promising higher spectral efficiency and increased data speeds. However, achieving high spectral efficiency and data rates requires precise channel estimation, which is difficult for mmWave MIMO due to scattering and blockages in general. Because of scattering and blockages, mmWave MIMO channels have intrinsic sparsity, which needs sparse-aware channel estimation algorithms. As a result, this work propose a variable step-size zero-attracting least mean squares (VSSZALMS) based channel-estimator. In VSSZALMS the step-size increases (or decreases) as the mean-square error (MSE) increases (or decreases) that’s result adaptive estimator based on VSSZALMS achieves better tracking and faster convergence rate. Convergence and steady-state behavior of estimator is analyzed. Simulations for a typical mmWave MIMO channels demonstrate the benefits of the proposed sparse channel-estimation approach and its convergence.