ON THE YULE WALKER EQUATIONS FOR THE ALL-POLE COEFFICIENTS

In this paper, a method for estimating the coefficients of an all-pole random process based on measurements corrupted by additive white Gaussian noise is described. The system is driven by a zero-mean independent and identically distributed (i.i.d) non-Gaussian sequence. The method proposes the estimation of the all-pole coefficients by extending the Yule Walker Equations to a triple correlation of the contaminated observations of the output sequence. Simulations verify the performance of the proposed method. The proposed method was compared with the Yule-Walker and Burg methods at different levels of signal-to-noise ratio (SNR) on the output.