Throughput optimization for orthogonal random beamforming in multi-user MIMO-TDMA downlink

We address the problem of throughput optimization in MIMO-TDMA downlink using orthogonal random beamforming (ORBF), which has been intensively studied recently because of its ability to exploit multi-user diversity and spatial multiplexing gains with partial channel state information (CSI). We formulate an optimization problem which aims at maximizing the total throughput under the constraint of individual long term throughput guarantee for each user. Although this problem is generally non-convex, we find that the optimal solution can be found with high probability by combine convex optimization algorithms with a well designed initial algorithm. By comparing the solution of the proposed method with the optimal value of the dual problem, we showed that the optimal throughput is usually achieved by the proposed method.