Estimation of parameters of a harmonic chirp model

Abstract

Abstract Here, we address the problem of estimation of the parameters of a harmonic chirp model, often encountered in speech and music applications. This model was introduced recently by Christensen and Jensen [1] as an extension of a standard harmonic model. We propose two methods of estimation: the least squares estimation method and the approximate least squares estimation method. We establish the asymptotic properties of the least squares estimators as well as the approximate least squares estimators of the parameters of this model under the assumption of stationary errors. These asymptotic properties are proved analytically as well as corroborated through simulation experiments. We present two speech signal data sets and their analysis using both the estimation methods. The results show that the proposed methods perform reasonably well for estimating the unknown model parameters.

Keywords

• chirp modulation
• least squares approximations
• music
• parameter estimation
• speech processing