Scheduling strategies for multiuser MIMO OFDM systems with limited feedback

In a cellular system with partial channel state information, i.e. with limited feedback, and multiple antennas, we consider scheduling of downlink transmissions at the base station using beamforming and orthogonal frequency division multiplexing. As the complexity of optimal scheduling for throughput maximization under quality of service constraints grows exponentially with the number of subcarriers, in this paper we propose two low complexity suboptimal techniques. The first is a greedy iterative strategy that at each step selects one user with the aim of maximizing the weighted sum rate (WSR), without the need of recomputing the beamformer. We next propose a pre-selection strategy that removes from the selection process users that do not provide an increase in the WSR, thus speeding the search. Both complexity and performance of the proposed techniques are evaluated and compared with existing solutions in a long-term evolution 3GPP scenario.