Efficient iterative water-filling power allocation method in MU-MIMO broadcast channels

In this paper, we investigate the sum rate of multiuser multiple-input multiple-output broadcast (MU-MIMO) communication systems, the channel state information (CSI) is supposed perfectly known at the base station. Since power allocation is not a convex optimization problem in the presence of multiuser interference (MUI) in general, the optimal solutions proposed in the literature have exponential computational complexity, which is hard to implement for practice. We propose an iterative water-filling algorithm that takes advantage of the classical simple water-filling principle. The computational complexity is significantly reduced in the proposed algorithm compared with the methods in the literature only with a negligible performance degradation. In addition, the generalized eigenvalue technique for beamforming design is utilized in this paper for minimizing MUI, the number of users and the number of antennas of each user can be arbitrary. Simulations show that the sum rate of the proposed method is close to the sum capacity of the MU-MIMO broadcast channel, especially in low signal-to-noise ratio (SNR) region.