Estimation of optimum number of poles for random signal by Yule-Walker method

Abstract

The Yule-Walker method is an effective method to estimate the system response or spectrum for random signal. Hence most of the noise and spurious signals are random in nature, so it is very convenient to estimate their spectrum by Yule-Walker method successfully. The Yule-Walker method is an autoregressive process to estimate the poles and errors also based on the number of poles for Wide Sense Stationary (WSS) process as well. Moreover the value of zero will be correspondingly calculated based on the poles in case of all poles model. The main concern of this paper is to analyze the Yule-Walker method and estimate the poles and zero along with the error based on the number of poles for a random signal. Moreover analyze the results to find out the optimum number of poles for least possible error.