Performance of ZF and CB Precoding Techniques for Large-Scale MU-MIMO System

Abstract

Large Scale Multi User-MIMO (MU-MIMO) is a key technology with reference to 5G to achieve higher spectrum as well as energy efficiency. The new technology refers to the use of a large number of antennas at the base station serving many user terminals in the same time and frequency resource allowing the channel vectors nearly orthogonal as a result, there is a reduction in inter-user interference and users may be served with the significant data rate. The linear precoding techniques play a vital role in the reduction of interference among users and cells. In this paper, we have derived, analyzed and compared two important precoding techniques i.e. Zero-forcing (ZF) and Conjugate beamforming (CB) for large-scale multiuser-MIMO. We analyze these precoding techniques with respect to spectral efficiency and downlink power with imperfect channel state information (CSI) as well as with perfect CSI. It is shown that ZF performs better as compared to CB precoding for achieving higher spectral efficiency and requires lower downlink power. CB outperforms the ZF in terms of downlink transmit power when there is a requirement to achieve low spectral efficiency and also for cell- edge users, hence energy efficient in these cases. It is shown from simulation results that ZF precoding is the better choice for attaining higher spectral and energy efficiency for a large scale multiuser-MIMO communication system.