Improved Linear Precoding over Block Diagonalization in Multi-cell Cooperative Networks

Abstract: In downlink multiuser multiple-input multiple-output (MIMO) systems, block diagonalization (BD) is a practical linear precoding scheme which achieves the same degrees of freedom (DoF) as the optimal linear/nonlinear precoding schemes. However, its sum-rate performance is rather poor in the practical SNR regime due to the transmit power boost problem. In this paper, we propose an improved linear precoding scheme over BD with a so-called "effective-SNR-enhancement" technique. The transmit covariance matrices are obtained by firstly solving a power minimization problem subject to the minimum rate constraint achieved by BD, and then properly scaling the solution to satisfy the power constraints. It is proved that such approach equivalently enhances the system SNR, and hence compensates the transmit power boost problem associated with BD. The power minimization problem is in general non-convex. We therefore propose an efficient algorithm that solves the problem heuristically. Simulation results show significant sum rate gains over the optimal BD and the existing minimum mean square error (MMSE) based precoding schemes.