QR decomposition and Gram Schmidt Orthogonalization based low-complexity multi-user MIMO precoding

Block diagonalization (BD) is a well-known linear precoding scheme for multi-user multiple-input multipleoutput (MU-MIMO) downlink transmission. BD precoding achieves good performance but results in high computational complexity overheads due to two singular value decomposition (SVD) operations for each user. By replacing the first SVD operation with a less complex solution, the QR decomposition based block diagonalization (QRBD) precoding scheme reduces the computational complexity. In this paper, we present a QR decomposition and Gram Schmidt Orthogonalization (QR-GSO) based MUMIMO precoding scheme as a new approach for MUMIMO downlink transmission. The proposed scheme introduces an improved QR decomposition and Gram Schmidt Orthogonalization method to significantly reduce the computational complexity. Our results show that the proposed QR-GSO precoding scheme achieves the same performance as BD and QR-BD precoding schemes, while requiring much lower computational complexity.