Performance Analysis of Sparse Channel Estimators for Millimeter Wave Hybrid MIMO Systems With Non-Ideal Hardware

Millimeter wave (mmWave) multiple-input multiple-output (MIMO) is the state-of-the-art physical layer technique for the fifth and beyond fifth-generation (5G/B5G) wireless communication systems. However, existing works in mmWave hybrid (analog and digital) MIMO systems do not adequately address the impact of unavoidable residual transceiver hardware impairments (HIs). This paper, considers a mmWave hybrid MIMO system with residual HIs and estimates the channel of considered system in a downlink scenario. The residual transceiver HIs are modeled as additive distortion noise, that severely affects the received pilot and information signals, which makes channel estimation a challenging task. As distortion noise is non-stationary, hence, an online adaptive filtering-based zero-attracting least mean square (ZALMS) algorithm is proposed. To ensure a lower mean square error the range of step-size and regularization parameters are obtained. Further, to achieve a faster convergence rate a sparse-initiated ZALMS (SI-ZALMS) is proposed. Furthermore, the impact of HIs on the mean square deviation and spectral efficiency is also analyzed. The proposed method offers significantly lower computational complexity as compared with the existing sparse channel estimation methods like Bayesian compressive sensing and sparse Bayesian learning. Simulation and analytical results corroborate the superiority of the proposed method as compared with existing methods.