User Selection for MIMO Downlink With Digital and Hybrid Maximum Ratio Transmission

Abstract

In this paper, we propose a user selection algorithm suitable for digital and hybrid maximum ratio transmission (MRT) precoding to alleviate inter-user interference. For digital MRT, the minimum inter-user interference problem is reduced to the notorious NP-hard called max k -cut. We first relax the max k -cut problem into a semi-definite programming (SDP), which can be solved by the interior point method. Then, we propose a randomized rounding-based greedy user selection algorithm (RRGUSA) to get the approximate solution of the max k -cut from the SDP solution. For hybrid MRT, the minimum inter-user interference problem is more complicated. We first force the interference minimization problem of hybrid MRT to be a quasi-max k -cut problem according to the law of large numbers. Then, RRGUSA is still used to obtain an approximate solution. Simulation results show that, for digital MRT, the proposed algorithm obtains a near-optimal solution, and for hybrid MRT, the proposed algorithm brings better performance by deploying more antennas.