Estimation of broadband multiuser millimeter-wave massive MIMO-OFDM channels by exploiting their sparse structure

Abstract

In millimeter wave (mmWave) massive multiple-input multiple-output (MIMO) systems, acquiring accurate channel state information is essential for efficient beamforming (BF) and multiuser interference cancellation, which is a challenging task since a low signal-to-noise ratio is encountered before BF in large antenna arrays. The mmWave channel exhibits a 3-D clustered structure in the virtual angle of arrival (AOA), angle of departure (AOD) and delay domain that is imposed by the effect of power leakage, angular spread and cluster duration. We extend the approximate message passing (AMP) with nearest neighbor pattern learning algorithm for improving the attainable channel estimation performance, which adaptively learns and exploits the clustered structure in the 3-D virtual AOA-AOD-delay domain. The proposed method is capable of approaching the performance bound described by the state evolution based on vector AMP framework, and our simulation results verify its superiority in mmWave systems associated with a broad bandwidth.

More information

Identifiers

ORCID for Lajos Hanzo: orcid.org/0000-0002-2636-5214

Catalogue record

Export record

Altmetrics