Optimization Problems In MIMO Downlink Systems

Nowadays Multiple-Input Multiple-Output system is a widely used wireless communication systems. Precoding technology is a pivotal technology in MIMO systems. It can adjust precoding strategies which adapts to the change of the real channel state information. Consequently, it can improve channel capacity, suppress channel fading and reduce Bit Error Ratio of MIMO systems. Therefore, precoding technology and its related problems has drawn a lot of researchers' interests in recent years.In this paper, an optimization problem for the sum rate maximization in multiuser multiple-input multiple-output (MIMO) downlink systems where a single base station equipped with multiple antennas communicates with multiple users each with a single antenna is investigated. With the regularized channel inversion (RCI) technique proposed by Peel at al. as the precoding strategy at the base station, we focus on finding a new and efficient power allocation strategy to maximize the sum rate. Although this programming problem is non-convex and highly nonlinear, based on simplifying its KKT conditions reasonably, we delete a term reasonably,which is proven to be a higher-order infinitesimal amount as the variance of noise tends to zero or infinity. Then a simple and efficient method referred as the water-filling (WF)-like algorithm is proposed for the power allocation of the MIMO downlinks. The proposed algorithm can get the approximate KKT points by solving the modified KKT conditions directly.Numerical experiments by MATLAB show the comparison between the proposed WF-like algorithm and other algorithms, including the classical regularized channel inversion, Zero-forcing beamforming and three typical algorithms. Furthermore, the experiments mainly show the comparison on the sum rate, computational complexity and computational time. The simulation results show that the proposed WF-like algorithm achieves relatively better performance compared with other algorithms in the three aspects.