Performance Analysis of Multi-User MIMO Downlink System in Time-Varying Channels

The channel state information (CSI) at the base station (BS) in multi-user multiple-input multiple-output (MU-MIMO) downlink system can provide a significant performance enhancement in terms of data transmission rate and interference reduction. However due to the time varying nature of the channel the available CSI at the BS can be outdated and this leads to the performance loss. This paper provides an analysis of the effect of outdated CSI on MU-MIMO downlink system when mobile stations (MSs) experience different speeds and how the channel prediction algorithms can be used to compensate for these delays in the time-varying channel. The simulation results show that even few samples delay can have a profound effect on bit error rate (BER) performance in MU-MIMO downlink system and the channel prediction algorithms can be used to compensate these delays to improvise the performance. For the first time, we propose to use normalized least mean square (NLMS) and recursive least square (RLS) adaptive filters for predicting the channels in MU-MIMO downlink system. Compared to the conventional linear extrapolation channel prediction method, both NLMS and RLS provided better performance, with RLS performing better with an increased computational complexity.