OUS: Optimal user selection in MU-MIMO WLAN

In multiuser MIMO (MU-MIMO) networks, different users can transmit packets to the AP concurrently, thereby network capacity could be significantly improved. Previous works, using the conventional user selection schemes, have shown the promise of MU-MIMO in practical WLANs. However, we find that, to harness the full potential of MU-MIMO, an optimal user selection scheme is required. In this paper, we formulate the user selection problem in MU-MIMO as a discrete constrained optimization problem, and by decomposing and solving this problem, we design an optimal selection scheme called OUS. OUS efficiently allocates network resources so as to improve both the overall throughput of a network and fairness among users. We conduct extensive simulations to evaluate performance of OUS, and the simulation results show that compared with previous works, OUS improves the overall throughput by 40% and improves the fairness by 30% in most cases.