Minimum MSE Transmitter and Receiver FIR MIMO Filters Including Decision Feedback and Near-End Crosstalk

An FIR multiple-input multiple-output (MIMO) filter communication system is optimized by minimizing the block decision point mean square error (MSE) between the desired vector and the input vector to the decision unit. Decision feedback is included in the receiver such that the receiver contains both feedforward and feedback linear FIR MIMO filters. The FIR MIMO transfer functions that model the communication link between the transmitter and receiver and the undesired nearend crosstalk channel are assumed known both at the transmitter and receiver. The channel input vector time series has constrained average power. Under the assumption that the output of the hardlimiter in the receiver is equal to the desired signal, an iterative numerical solution is proposed that minimizes the block decision point MSE with respect to the coefficients in the transmitter and receiver filters. When compared to a method available in the literature, the proposed method yields better results since it is applicable to a more general scenario that may include near-end crosstalk and the transmitter filter is optimized. In the proposed method, all feedforward and feedback filter structures are treated in a unified way.