Hybrid Beamforming with Random Analog Sampling for Wideband Channel Estimation in Millimeter Wave Massive MIMO Systems

Abstract

Hybrid analog and digital BeamForming (HBF) transceiver architectures realizing directive communication over large bandwidths in millimeter Wave (mmWave) massive Multiple Input Multiple Output (MIMO) systems require the availability of accurate channel estimation. This is, however, a challenging task mainly due to the short channel coherence time and the hardware limitations imposed by HBF architectures. In this paper, we capitalize on recent matrix completion tools and develop a novel wideband channel estimation technique for mmWave massive MIMO systems with HBF reception. The proposed iterative algorithm exploits jointly the low rank and beamspace sparsity properties to provide more accurate recovery, especially for short beam training intervals. We introduce a novel analog combining architecture that includes a random subsampling step before the input of the analog received signals to the digital component of the HBF receiver. This step supports the proposed estimation technique in providing the sampling set of measurements for recovering the unknown channel matrix. The impact of various system parameters on the effectiveness of the designed algorithm and the performance improvement of our technique over representative state-of-the-art ones is demonstrated through indicative simulation results.