Favorable Propagation with Practical Angle Distributions for mmWave Massive MIMO Systems

Abstract

Favorable propagation (FP) defined as orthogonality of channel vectors associated with different users is a key assumption in massive MIMO systems. In this paper, theoretical analysis on asymptotically FP of a sparse millimeter-wave (mmWave) channel is provided under the assumption that the angle of arrival (AOA) follows wrapped Gaussian and truncated Laplacian distributions which are practical distributions used to model AOA in 3D MIMO channel model standards. Further, we theoretically prove that asymptotically FP is satisfied under practical angle distributions and the FP condition under the uniform distribution is a special case of our results. To investigate the proximity of real channels to FP, the eigenvalue spread is studied and simulation results imply that environments with rich scatterers are easier to satisfy FP condition than that with fewer scatterers and the assumption of the uniform distribution underestimates the gap between real channels and FP.