Tomlinson-Harashima Precoding Optimization over Correlated MIMO Channels

Fast temporal variations in wireless Multiple-Input Multiple-Output (MIMO) channels cause a significant degradation in traditional Tomlinson-Harashima precoding (THP) performance. The use of channel feedback from receiver to transmitter is a standard approach in order to adapt the THP with channel variations. While knowledge of the channel at the transmitter would produce good benefits for wireless communications, the generation of reliable channel feedback is complicated in fast time-varying channel for mobile applications. In temporal correlated MIMO channels, the use of the correlation property of the channel can be led to significantly less channel feedback than other traditional THP's design. This paper proposes a novel robust THP design for a time-varying, frequency-nonselective MIMO broadcast channel in which the Channel State Information (CSI) is updated at the beginning of each block (where the perfect CSI is assumed to be known). As the channel varies over time, THP blindly pre-equalizes the channel during the next symbol time. The algorithm leads to a cost-efficient CSI updating procedure for the THP optimization. Simulation results are provided to confirm the new THP design performance against traditional methods.