Two-Dimension Coherent Multipath DOA Estimation without Spatial Smoothing in Massive MIMO Systems

摘要

The ever growth of traffic data in mobile communications indeed impels the continuous technological evolution in wireless industry. To achieve such exponential data growth, a successful marriage of massive multiple-input multiple-output (MIMO) and millimeter wave (mmWave) may be an enabling technology in future 5G cellular networks. In this paper, we propose a novel two-dimension direction-of-arrival (DOA) estimation algorithm for coherent multipath signals without requiring spatial smoothing, which can be utilized to assist in performing downlink precoding in massive MIMO system. The key idea behind is to degenerate a two-dimension weighted subspace fitting problem into two one-dimension versions. Compared with the tensor-MODE algorithm and spatial smoothing algorithm, the proposed one is not only a computational efficient choice but also manifests improved estimation performance.