An improved MMSE Tomlinson-Harashima precoding for the joint transceiver of MIMO systems

The minimum mean square error (MMSE) Tomlinson-Harashima precoding (THP) algorithm can provide a solution for the joint transceiver of MIMO systems by pre-equalizing the interlayer interference. However, the MMSE THP algorithm can´t achieve full diversity and its bit error rate (BER) performance is unsatisfactory. In this paper, an improved MMSE THP algorithm is proposed. In this algorithm, the channel gain matrix is decomposed with geometric mean decomposition (GMD) to obtain identical sub-channels and translated into a new matrix with smaller condition number using lattice reduction. The simulation results show that the proposed algorithm can improve the BER performance, achieve full diversity and its complexity is applicable.