Spectral and Energy Efficiency Analysis for SLNR Precoding in Massive MIMO Systems With Imperfect CSI

We derive tractable bound expressions on achievable spectral efficiency for a multiple-input multiple-output (MIMO) system with a signal-to-leakage-plus-noise ratio precoding scheme (SLNR-PS) under the condition of imperfect channel state information. These bounds are tight and approach exact values when the number of base station (BS) antennas is large. A problem of energy efficiency (EE) maximization is investigated by using a practical power consumption model. The effects of the system parameters and quality of channel estimation, including the number of BS antennas, transmit power, and training length on the performance metrics, are explicitly analyzed. Following that, an alternating optimization algorithm is employed to obtain the optimum EE. It has been shown that the proposed SLNR-PS performs better than the matched-filtering and zero-forcing schemes and a deployment of massive MIMO with the optimal transmit power and training length can achieve high EE.