Precoding for MU-MIMO system in low scattering channel environments

Abstract

In Multi-user Multi-input Multi-output (MU-MIMO) systems, the most popular method to eliminate the multi-user interferences is projecting every user's precoding matrix into the null space of the rests' channels, so that data streams intended for different users would be transmitted in orthogonal spaces. However, while users having mutual correlated channels caused by the increased density of smart phones nowadays, this method would make correlated subspaces unused by data transmission for any user, and cause the reduction of overall throughput. To address the corresponding issue, we in this paper study the efficient precoding scheme that can eliminate the multi-user interferences as well as maximize the system capacity for MU-MIMO system in the spatially low scattering channel environments. Unlike the existing methods, our approach allows high-QoS-requirement users' precoding matrices to lie in the correlated subspaces between the channels of these primary users and the rests, so that more data streams can be transmitted to the primary ones simultaneously. In the mean time, in order to avoid the multi-user interferences, we require that every correlated channel can only be occupied by one user's precoding matrix. Following the aforementioned principles, we propose a Correlated Channels Utilized Precoding (CCUP) scheme for MU-MIMO system. Under this scheme, we also propose a different utilizing criterion of correlated subspaces to satisfy the networks with same-QoS-requirement services, in which every user's precoding space can be broadened by adopting partial correlated subspaces, termed Fair CCUP (F-CCUP) scheme. Simulation results show that while keeping no multi-user interferences, our proposed CCUP scheme can greatly increase the number of allowed independent data streams, and improve the overall capacity of MU-MIMO system in spatially correlated propagation environments compared to the existing approach.