Performance analysis for downlink massive MIMO system with ZF precoding

This paper investigates the performance for the downlink massive multiple-input multiple-output system when the base station serves multiple user terminals (UTs) using zero-forcing precoding. The tight lower bound of average area spectrum efficiency (A(2)SE) is derived as a function of the number of transmission antennas, the number of UTs and the equivalent transmission signal-to-noise ratio. Regarding the lower bound as the approximate A(2)SE, the optimal number of UTs maximising the A(2)SE is attained for given the number of transmission antennas and the equivalent transmission signal-to-noise ratio. After formulating the realistic power consumption model in consideration of both the transmission power and circuit power, the trade-off between energy efficiency (EE) and A(2)SE is established, and the optimal EE with respect to A(2)SE is deduced. Moreover, both the outage probability and bit error ratio expressions are derived related to the degrees of freedom, defined as the difference between the number of transmission antennas and the number of UTs. Simulation results coincide to the analysis well, and they indicate that deploying more transmission antennas or multiplexing a rational number of UTs can improve the A(2)SE and EE, increasing the degrees of freedom will better both the outage probability and bit error ratio. Copyright (C) 2013 John Wiley & Sons, Ltd.