Spatial Denoising for Sparse Channel Estimation in Coherent Massive MIMO

In this paper, we present a new sparse channel estimation (CE) for phase-synchronized sub-6G Massive Multiple-Input Multiple-Output (MIMO) receivers. The algorithm employs an iterative search for propagation channel taps, spatial filtration, and denoising. For that, we split each channel tap signal into tap-aligned and orthogonal parts and train their weighting coefficients in non-line-of-sight channel realizations generated in the QuaDRiGa 2.0 software. Simulation results are presented and compared with the beamspace CE implemented by the digital transformation of the antenna signal to a priori selected subspace directed towards channel taps.