Maximum SNR decision-feedback equalization with FIR filters: filter optimization and a signal processing application

An alternative approach to filter optimization for decision-feedback equalization with FIR filters is presented. The optimum feedforward and feedback filters under the design criterion of maximum SNR instead of minimum mean-squared error (MMSE) are derived. After a short review of the MMSE solution, it is shown, that both solutions are equivalent, if a bias in the reference tap, inherent in the MMSE solution, is removed. Finally, the derived solution is used to solve a problem in signal processing, namely factorization of a nonnegative function. This procedure may be superior to the traditional factorization techniques in certain applications