Linear Precoding-Based Geometric Mean Decomposition (LP-GMD) for Multi-User MIMO Systems

In this work, a novel linear precoding algorithm, called linear precoding-based geometric mean decomposition (LP-GMD), is proposed for the downlink of multi-user (MU) multiple-input multiple-output (MIMO) systems. The proposed scheme suppresses the multi-user interference (MUI) at the BS. Then it uses the GMD to decompose the equivalent single-user MIMO channel into multiple parallel sub-channels that have the same effective signal to interference plus noise ratio (SINR), which allows the use of the same modulation and coding scheme (MCS) for different spatial streams of each user. A system-level simulator is then set up based on the IEEE 802.11ac standard. The simulation results indicate that the proposed algorithm provides a significant gain as compared to conventional linear precoding algorithms such as block diagonalization (BD) and regularized block diagonalization (RBD).