Robust Scheduler Prioritizing UEs with Time-Variant Channels in Small-Delay Slots from Channel Estimation Timing for 5G Large-Scale MU-MIMO

In fifth-generation (5G) mobile communication systems, it is necessary to significantly increase system capacity compared with 4G in order to support rapidly increasing mobile data traffic. The multi-user (MU) MIMO, which spatially multiplexes multiple user equipments (UE) in the same time and frequency resource, is the most promising technology. We have been studying large-scale coordinated MIMO, which uses zero forcing (ZF) precoding at the transmitter side, for a distributed antenna system. However, channel mismatch between channel estimation timing and data transmission timing drastically degrades throughput performance. In this paper, we propose a novel scheduler that preferentially schedules high speed UEs in small-delay slots from channel estimation timing by exploiting the diversity in terms of UE moving speed. The proposed scheduling is realized by the signal to interference plus noise ratio (SINR) prediction from channel variation between two most recent channel estimates and by inputting the predicted SINR into the proportional fairness (PF) scheduler. Computer simulations show that the proposed method significantly improves the mean UE throughput for high speed UEs while keeping almost the same mean UE throughput for low speed UEs.