Title :
Spectral estimation using dynamical system theory
Author :
Klumpp, Eric ; Mailhes, Corinne
Author_Institution :
ENSEEIHT/TeSA, Nat. Polytech. Inst. of Toulouse, Toulouse, France
Abstract :
This paper addresses the problem of spectral estimation by using dynamical system theory. A new parametric model motivated by the Takens theorem is defined. This model referred to as dynamical AR model is showed to outperform the usual AR model to estimate the frequencies of sinusoidal signals.
Keywords :
Kalman filters; frequency estimation; spectral analysis; Kalman filter; Takens theorem; dynamical AR model; dynamical system theory; sinusoidal signal frequency estimation; spectral estimation problem; Abstracts; Delays; Estimation;
Conference_Titel :
Signal Processing Conference, 2002 11th European
Conference_Location :
Toulouse
