ARMA modelling based on root cepstral deconvolution

Abstract

Cepstral deconvolution has been successfully applied to many diverse fields, such as speech and seismic analysis. Thus far, all cepstral modelling performance has been empirical, relying on the judgement of the designer. A novel method for measuring root cepstral ARMA modelling performance is proposed, by introducing a cost function applied directly to the root cepstral domain. It is, therefore, possible to demonstrate the optimised modelling of an ARMA process and show that its performance is superior to that of a FIR Wiener filter and linear prediction.