Channel Estimation for IRS-Assisted mmWave Massive MIMO Systems in Mixed-ADC Architecture
Rui Zhang, Weiqiang Tan, Shidang Li, Maobin Tang
Abstract
The accuracy of channel state information (CSI) acquisition is nontrivial for intelligent reflection surface (IRS)-assisted massive multiple-input–multiple-output (MIMO) systems due to the IRS comprises of a large number of low-cost electromagnetic reflection units that can smartly reflect the impinging signal. In this article, we investigate the downlink channel estimation problem for IRS-assisted millimeter-wave (mmWave) massive MIMO systems, where the IRS provides effective reflected paths to enhance the coverage and spectral efficiency. To reduce the hardware costs and power consumption, the massive MIMO system is adopted the mixed-analog-to-digital converter (ADC) architecture, in which a fraction of antennas is equipped with high-resolution ADCs and a large number of antennas are deployed with low-resolution ADCs. Considered the finite-dimensional millimeter-wave channel model, we utilize the special row–column sparse characteristics of the channel matrix and propose a row-structured sparsity based on the orthogonal matching pursuit (RS-OMP) algorithm. The RS-OMP algorithm sequentially calculates the row–column support sets of the angular cascaded channel matrix and then uses the least square (LS) algorithm to reconstruct the cascaded channel matrix, which aim to achieve a lower computational complexity. Simulation results demonstrate that under the same system configuration, the proposed RS-OMP algorithm not only improves the accuracy of channel estimation but also reduces more than 75% of the pilot overhead compared to the traditional orthogonal matching pursuit algorithm.