Virtual AoA and AoD estimation for sparse millimeter wave MIMO channels

Abstract

In this paper, we investigate an angle of arrival (AoA) and angle of departure (AoD) estimation algorithm for sparse millimeter wave multiple-input multiple-output (MIMO) channels. The analytical channel model whose use we advocate here is the beam space (or virtual) MIMO channel representation. By leveraging the beam space MIMO concept, we characterize probabilistic channel priors under an analog precoding and combining constraints. This investigation motivates Bayesian inference approaches to virtual AoA and AoD estimation. We divide the estimation task into downlink sounding for AoA estimation and uplink sounding for AoD estimation. A belief propagation (BP)-type algorithm is adopted, leading to computationally efficient approximate message passing (AMP) and approximate log-likelihood ratio testing (ALLRT) algorithms. Numerical results demonstrate that the proposed algorithm outperforms the conventional AMP in terms of the AoA and AoD estimation accuracy for the sparse millimeter wave MIMO channel.