A low-complexity 3D MU-MIMO downlink scheme based on two-step precoding

Abstract

The emergence of Internet of Things (IoT), which connects billions of smart devices together wirelessly, has brought great pressure on capacity of current wireless networks. Fortunately, a new technology, three-dimension multiple-input multiple-output (3D-MIMO), with a large number of active antennas equipped in a 2D planar array, has been shown significant potential in improving system capacity. A problem is that with the number of antennas increasing, the complexity of the conventional multi-user (MU) precoding algorithms increase rapidly. To cope with this problem, a two-step 3D MU-MIMO precoding scheme is proposed to suppress the interference not only in horizontal but also in vertical domain in this paper. Meanwhile, considering the strong correlations of channel matrices in vertical dimension for different users, we adopt chordal distance-based user selection algorithm to find the users with correlations of the vertical antennas' channels as low as possible for the first step precoding. Simulation results show that the proposed transmission scheme combining two-step precoding with user selection in vertical dimension can achieve better cell average spectrum efficiency.