Multiple Antenna Gains in Low Complexity Cross-Layer Scheduling with Link Adaptation and QoS Requirements

Traditional works rely on the use of complete Channel State Information (CSI) at the base station to take space-time scheduling decisions in the uplink of multi-user MIMO wireless systems. However, this might be a complex and costly solution. The simplest solutions are the random scheduler and the Round Robin (RR) scheduler that schedule terminals either randomly or sequentially without using CSI. In multiple antenna systems, thanks to the additional spatial dimension, many terminals could be simultaneously scheduled. The optimal number of terminals to be scheduled by a low complexity space-time scheduler in a multi-user SIMO system with link adaptation and QoS requirements is investigated in this paper. Furthermore, by taking a cross-layer point of view, we also study how the number of antennas, the QoS requirements and the number of terminals affect system gains in terms of throughput and delay. In particular, when a ZF receiver is used, we show that the benefits of using multiple antennas increase with the QoS requirements or, what is the same, decrease by increasing the average SNR. Besides, when the number of antennas is much higher than the number of terminals, the throughput gains increase linearly with the number of terminals and exponentially with the QoS requirements.