SVD based information theoretic criteria for detection of the number of sinusoids

Author

Biradar, L. ; Reddy, V.

Author_Institution

Dept. of Electr. Commun. Eng., India Inst. of Sci., Bangalore, India

fYear

1990

fDate

3-6 Apr 1990

Firstpage

2519

Abstract

Following the information theoretic approach to model selection, criteria for detecting the number of sinusoids are developed. The criteria are well matched to singular-value-decomposition (SVD-) based methods of spectral estimation in that the additional computations required over and above those needed for the SVD calculation are marginal. The detection performance of these criteria is compared with that of the J.J Fuchs (1988) method. The results show that the minimum description length (MDL)-based criterion, developed in the paper, and the Fuchs method perform similarly. However, the latter method involves substantially more computations than the former

Keywords

information theory; parameter estimation; signal detection; SVD calculation; information theoretic approach; minimum description length; model selection; singular-value-decomposition; sinusoids detection; spectral estimation; Covariance matrix; Density functional theory; Difference equations; Eigenvalues and eigenfunctions; Frequency; Gaussian processes; Matrix decomposition; Random processes;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 1990. ICASSP-90., 1990 International Conference on

Conference_Location

Albuquerque, NM

ISSN

1520-6149

Type

conf

DOI

10.1109/ICASSP.1990.116112

Filename

116112