Joint Spatially Sparse Channel Estimation for Millimeter-wave Cellular Systems
Journal
IEEE Global Conference on Signal and Information Processing
Date Issued
2017
Author(s)
Abstract
We consider the channel estimation for millimeter-wave (mmWave) cellular systems in this paper. Because of the limitation of the hybrid RF/digital beamforming and the adoption of large-scale antennas, the acquisition of channel state information (CSI) is challenging. By exploiting the joint spatially sparse nature of practical mmWave channels, we develop a novel channel estimation algorithm based on compressive sensing (CS) with low training overhead and computational cost. A multiple measurement vectors (MMV) problem is formulated to exploit the joint sparse structure of the unknown channel, and then solved by the proposed rank-aware channel subspace matching pursuit (RA-CSMP) algorithm. Numerical results demonstrate the superiority of the proposed approach in terms of estimation quality in the more scattered channel environments. Moreover, it achieves about 35∼150 times faster runtime than other previous CS-based works in the experiment.
Type
conference paper