Low Complexity Channel Estimation for Massive MIMO Systems

In this paper, a low complexity channel parameter estimation method is proposed for two-dimensional (2D) uniform rectangular array (URA) massive multiple-input and multiple-output (MIMO) systems. Instead of assuming independent fading between different transmit-receive antenna pairs, a physical channel which models the realistic scattering environment via the angles and gains associated with different propagation paths is studied. A novel 2D Fourier transform (FT) based on elevation and azimuth steering factors is designed to derive the spatial spectrum distribution of received signals at base-station (BS). Accordingly, the received data matrices are used to estimate the channel parameters, which includes the direction of arrival (DOA) angles and the channel gains respective to each resolvable path. Since the channel coefficients are estimated from the DOA perspective and the proposed 2D FT can be realized by Fast-Fourier-Transform (FFT), the computational complexity is reduced significantly. Simulation results are provided to verify the accuracy and the complexity of the proposed scheme.