Improved precoding algorithm design for downlink large scale MU-MIMO system

Abstract

Large Scale Multiuser multiple input multiple output (LS-MU-MIMO) communication system is considered as an encouraging solution to attain greater spectral efficiency for the next-generation wireless communication system. In this paper, we propose an improved design of a precoding algorithm for LS-MU-MIMO communication system. The proposed precoding algorithm is based on Block diagonalization (BD) precoding which is a combination of QR decomposition and Zero forcing channel inversion (ZF-CI) operation named as QR-ZF-BD precoding algorithm. To enhance the performance of LS-MU-MIMO system, the proposed precoding algorithm utilizes ZF method and QR decomposition operation instead of more complex Singular value decomposition (SVD) operation as compared to the conventional Block diagonalization (BD) precoding technique. The proposed precoding algorithm is designed in two stages. In the first stage, the proposed QR-ZF-BD precoding algorithm first utilizes QR decomposition to diminish multiuser interference (MUI), the second stage utilizes ZF based channel inversion (ZFCI) and QR decomposition operation again to further enhance the spectral efficiency of a single cell downlink LSMU-MIMO system. The enhanced spectral efficiencies of values 334 bit/s/Hz and 388 bit/s/Hz are obtained with our proposed precoding algorithm design when LS-MU-MIMO system having 100 and 200 base station antennas respectively. The numerical results indicate that the proposed QR-ZF-BD precoding algorithm achieves the best spectral efficiency as compared to current LTE-A standard and other recently used conventional linear precoding algorithms.