Time Domain Channel Estimation for Time and Frequency Selective Millimeter Wave MIMO Hybrid Architectures: Sparse Bayesian Learning-Based Kalman Filter

Abstract

In this paper, a Sparse Bayesian Learning (SBL) based channel estimation technique for frequency selective millimeter wave (mmWave) channel in the time domain approach is developed. Further, SBL based Kalman filter (SBL-KF) for time and frequency selective mmWave multiple-input multiple-output (MIMO) hybrid architecture is presented. Hybrid precoders and combiners are designed to estimate the channel of mmWave MIMO systems. The hybrid precoding technique provides low power consumption and high achievable rate performance at mmWave frequencies. mmWave channels are sparse in nature, and the sparse recovery problem is estimated using the channel estimation technique. A simulation result of SBL-KF is improved by 4 dB and 10 dB of SNR compared to the conventional SBL and Orthogonal Matching Pursuit based scheme, respectively. The proposed SBL-KF scheme provides low estimation error at smaller training overheads M = 50 compared to the other existing work.