Spectral Efficient Precoding Algorithm for Large Scale MU-MIMO Communication System

We analyze the performance of a large scale multiuser-multiple-input multiple-output (MU-MIMO) communication system, where the base station (BS) has perfect channel state information (CSI). The performance is analyzed with various precoding techniques i.e. maximum ratio transmission (MRT), zero-forcing (ZF), Block diagonalization (BD), QR-BD, minimum mean square error (MMSE), and proposed QR-MMSE-BD precoding techniques under the Rayleigh fading channel. The BD precoding scheme eliminates multiuser interference (MUI) in the MU-MIMO system. Also, the BD and QR-BD precoding techniques are difficult to implement practically due to the huge computational complexity involved. In this work, we have designed an improved model for the performance enhancement of BD based precoding algorithm for a large scale multiuser-MIMO communication system. The complexity of the proposed precoding design is reduced due to the utilization of QR decomposition operation in place of singular value decomposition (SVD). The performance of a large scale multiuser-MIMO system is analyzed in terms of spectral efficiency with the number of antennas at BS and SNR values. It is found that proposed QR-MMSE-BD precoding outperforms as compared to conventional precoding methods for large scale multiuser-MIMO communication system.