OMP with Grid-Less Refinement Steps for Compressive mmWave MIMO Channel Estimation

Abstract

Millimeter-wave (mmWave) communications will play an important role in future wireless communication systems. Due to the high omni-directional path loss present at mmWave bands, mmWave channels are dominated by a few reflection paths, only. The resulting sparsity of mmWave channels in the angular domain can be exploited in order to obtain channel estimates from compressed measurements. We propose mmWave multiple-input multiple-output (MIMO) channel estimation for compressive measurements using extensions of the conventional grid-based orthogonal matching pursuit (OMP) algorithm with grid-less refinement steps based on the maximum likelihood principle. Numerical examples show that the proposed estimators are able to significantly outperform, e.g., grld-based OMP.