Title :
Forward covariance least-squares algorithm: a new method in AR spectral estimation
Author :
Scott, Peter D. ; Nikias, C.L.
Author_Institution :
State University of New York, Department of Electrical Engineering, Buffalo, USA
Abstract :
A new method for estimating autoregressive (AR) process coefficients for spectral estimation is introduced by minimising the average power of the forward covariance prediction error. AR spectra generated by this method are shown to have improved performance over AR spectra generated by the best alternative forward prediction error method, the Yule-Walker estimator.
Keywords :
least squares approximations; spectral analysis; Yule-Walker estimator; autoregressive spectral estimation; forward covariance least squares algorithm;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:19810079
