Performance Analysis of Multi-User Massive MIMO Systems with Perfect and Imperfect CSI

Abstract

A collection of independent terminals instantaneously transmits data streams to a compact array of antennas. With imperfect channel state information (CSI), the antenna makes use of transmitted pilots to find the knowledge about CSI and the power dissipated by the terminals can be made reciprocally proportional to the square root of the number of antennas without a reduction in performances used at the base station. In contrast, when the CSI is known at the base station, the transmitted power is inversely proportional to the number of used antennas. Lower capacity bounds for zero-forcing (ZF) and maximum-ratio combining (MRC) detection are derived. A ZF receiver outperforms the MRC. A power scaling law is being taken into consideration in the analysis of the uplink sum rate with perfect and imperfect CSI. It is analyzed that the uplink sum rate for MRC and ZR receiver with perfect and imperfect CSI, considering the power scaling law. As the number base station antennas increases, the difference between the uplink sum rate for ZF and MRC decreases and if the number of base station antenna is very high, then both of these lines coincide with each other. Especially at a very large number of base station antennas, all bounds are very tight. In comparison to a single antenna system, the uplink sum rate (or spectral efficiency) can be improved in order of several magnitudes by using moderately large antenna arrays in massive multiple input multiple output (MIMO) system.