Title :
Application of maximum entropy spectral analysis to estimation of parameters of ARMA models by the ITIF method
Author :
Mart¿¿n, R. G¿¿mez ; Perez, M.C. Carri¿¿n ; Ibrahim, S. Al Khouri ; Olmedo, B. Garc¿¿a
Author_Institution :
Facultad de Ciencias, Dpto. de Electricidad, Granada, Spain
Abstract :
In the letter we show how it is possible to improve the results obtained by the ITIF method, modelling a deterministic, linear, time-invariant system through a rational function in the z-domain, when the input of the system is a white-noise signal. This is achieved by using the maximum entropy method (Burg algorithm) during the deconvolution stage instead of the covariance method.
Keywords :
filtering and prediction theory; parameter estimation; signal processing; spectral analysis; ARMA models; Burg algorithm; ITIF method; deconvolution stage; deterministic linear time-invariant system modelling; maximum entropy method; maximum entropy spectral analysis; parameter estimation; white-noise signal; z-domain;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:19850431