Weighted Subspace Fitting for Two-Dimension DOA Estimation in Massive MIMO Systems

Massive multiple-input multiple-output system can enable multi-stream transmission for high spectrum efficiency, which is a key technology in future 5G cellular networks. In this paper, we degenerate a direction of arrival (DOA) related 2-D weighted subspace fitting function into two independent parameterized 1-D versions. Based on this, we develop a novel 2-D DOA estimation algorithm, which can be utilized to assist in performing downlink precoding. Furthermore, we also make an analysis on the computational complexity and the theoretical Cramér-Rao lower bound. The direct merits are as follows: the proposed algorithm includes only once polynomial rooting and also does not require angle paring, hence it is of computational efficiency; in addition, compared with some existing algorithms, it can achieve higher 2-D angle estimating accuracy. A series of Monte Carlo simulations are subsequently carried out, which demonstrate the effectiveness of the proposed algorithm.