Using OFDMA for MU-MIMO User Selection in 802.11ax-Based Wi-Fi Networks

Two innovative multi-user (MU) communication mechanisms, multi-user multi-input multi-output (MU-MIMO) and orthogonal frequency division multiple access (OFDMA), have recently attracted much attention as key technologies to boost the capabilities of 802.11 Wi-Fi systems. The method of MU transmission in Wi-Fi systems still suffers from severe problems with channel state information (CSI) feedback overhead, however, and this precludes obtaining much higher MU-MIMO gains through user selection. Although in 802.11ax-based Wi-Fi systems the methods used for uplink OFDMA transmission and downlink MU-MIMO transmission bear a great deal of resemblance to each other, in the sense that both are initiated by the AP, this similarity in protocol is not appropriately utilized for close collaboration between the two technologies. This paper proposes MUSE, a novel multi-user transmission scheme for 802.11ax networks. By exploiting uplink OFDMA transmissions for MU-MIMO user selection, MUSE takes advantage of both OFDMA and MU-MIMO functionalities. In particular, the AP estimates and gathers multiple CSI values from the uplink OFDMA frame, and this channel information is then used to find the optimal downlink MU-MIMO receiver group to maximize the system utility. To realize MUSE, a new OFDMA resource allocation algorithm and frame structure are developed. Extensive MATLAB simulation results show that MUSE significantly improves the network throughput, even in dense network scenarios, and also works effectively in co-existence with legacy nodes.