Downlink SLNR based precoding and power control in massive MIMO systems with time-shifted pilots

Abstract

In massive multiple-input multiple-output (MIMO) system with time-shifted pilots, the interference from other cell groups, where the pilots are staggered with the ones in the target cell group, can be eliminated when the number of antennas at base station (BS) is infinite. However, with only hundreds of antennas at the BS in practice, both intra-group interference and inter-group interference can not be ignored. Aiming at this problem, a Signal to Leakage and Noise Ratio (SLNR) based downlink precoding and power control scheme is proposed in this paper. In the design of precoder, not only the interference leaked to the users within the same cell is considered, but also the interference leaked to the BSs which are receiving the uplink pilots is included by introducing a conversion factor. Furthermore, the channel estimation error (CEE) is modelled and considered to enhance the robustness of the precoder. To obtain the tradeoff between uplink channel estimation and downlink data transmission, the preferable transmission power and conversion factor are found by maximizing the asymptotic sum-rate, which is dependent on the large-scale fading coefficients and some constants. In the calculation of asymptotic sum-rate, the truncated polynomial expansion (TPE) method is used to simplify the precoder and further to decouple the sum-rate computation from the small-scale fading coefficients. Simulation results show that the proposed downlink transmission scheme outperforms the maximum ratio transmission (MRT) based downlink scheme remarkably. Moreover, the performance obtained by asymptotic sum-rate optimization is very close to the complicated instantaneous sum-rate optimization.