User selection schemes for maximizing throughput of multiuser MIMO systems using Zero Forcing Beamforming

The performance of a multiuser MIMO broadcast system depends highly on how the users being served are selected from the pool of users requesting service. Though dirty paper coding can obtain the optimal capacity of multiuser MIMO Broadcast systems, the computational complexity is high. We examine this problem in the context of systems using Zero Forcing Beamforming (ZFBF) which have low complexity and reasonable throughput. The channels of the selected users may not be orthogonal to each other resulting in harmful interference and reduction in the performance of the system. This work identifies and thoroughly investigates the main factors affecting the performance and proposes some mitigating schemes to fully exploit multiuser diversity. In particular, we develop improved user selection schemes designed to maximize the sum rate that offer a range of complexity to performance tradeoff. Simulation results show that the selection schemes developed compare favorably with existing techniques.