Quasi-Static and Time-Selective Channel Estimation for Block-Sparse Millimeter Wave Hybrid MIMO Systems: Sparse Bayesian Learning (SBL) Based Approaches

This paper develops schemes for block-sparse channel estimation in millimeter wave (mmWave) multiple-input multiple-output (MIMO) systems that exploit the spatial sparsity inherent in such channels. Initially, a novel sparse Bayesian learning (SBL) based block-sparse channel estimation technique is developed for a mmWave hybrid MIMO system with multiple measurement vectors, which overcomes the shortcomings of the existing orthogonal matching pursuit-based framework. This is subsequently extended to a temporally correlated block-sparse mmWave MIMO channel. Further, an online recursive hierarchical Bayesian Kalman Filter is developed for the estimation of a time-selective mmWave MIMO channel. Bayesian Cramer–Rao bounds are also derived for the proposed static and time-selective mmWave MIMO channel estimation schemes followed by precoder/combiner design employing the SBL-based imperfect channel estimates. Simulation results are presented to demonstrate the improved performance of the proposed SBL-based channel estimation techniques in comparison to the popular OMP-based scheme proposed recently.