Randomized Low- Rank Matrix Approximation-Based Channel Estimation for Massive MU-MIMO systems

It is a critical issue to estimate the large-dimensional channel matrix for massive multi-user multiple-input multiple-output (MU-MIMO) systems within the limited coherence time interval. In this paper, the low-rank property of the channel matrix under the finite scattering channel model is exploited to enhance the performance of least squares (LS) estimation. First, we propose a channel estimation method based on low-rank matrix approximation (LRMA) to substantially improve the performance of LS estimation. Then, in order to effectively eliminate the computational bottleneck brought by performing singular value decomposition (SVD) for the large-dimensional channel matrix and enhance the performance of LS estimation, a novel randomized LRMA-based channel estimation method is proposed. Finally, numerical results are provided to corroborate our theoretical results and to show the superiority of the proposed methods.