A comparison of methods for adaptively designing minimum phase FIR models

Abstract

Given a set of filters with the same magnitude response, the minimum phase filter is the one with the minimum phase lag. This filter will also have the minimum group delay and the minimum energy delay. Obtaining the minimum phase FIR estimate of an unknown system is a constrained optimization problem because the zeros of the model must be constrained to be within the unit circle. The authors compare several structures for adaptively obtaining the optimum FIR minimum phase filter model for an unknown FIR system.<>