Nonlinear precoding for spatial multiplexing MIMO-OFDM downlink systems

Considering the error propagation effect and high complexity of vertical bell labs layered space time (VBLAST), a novel nonlinear module algebra precoding algorithm based on geometric mean decomposition(GMD) for VBLAST is proposed in multiple input multiple output orthogonal frequency division multiplexing (MIMO-OFDM) downlink systems, geometric mean decomposition is used for precoding matrix which have the same equivalent noise gain among each sub-channel, the nonlinear module algebra Tomlinson-Harashima precoding THP is used between sub-carrier channels of OFDM to eliminate interference from other signals at the transmitter, and can eliminate the error propagation effect of layered space-time codes effectively, At the receiver, zero forcing and minimum mean square error criterion are used, and the complexity of the downlink receiver can be reduced. Simulation results show that the proposed algorithm is better than the traditional algorithms.