The threshold analysis of SVD-based algorithms

Author

Tufts, D.W. ; Kot, A.C. ; Vaccaro, R.J.

Author_Institution

Dept. of Electr. Eng., Rhode Island Univ., Kingston, RI, USA

fYear

1988

fDate

11-14 Apr 1988

Firstpage

2416

Abstract

The problem of analyzing the threshold effect of signal processing algorithms which use the singular-value decomposition (SVD) is addressed. The probability of obtaining an outlier is calculated and used to determine the threshold SNR at which the variance of parameter estimation errors depart from Cramer-Rao bound behavior. Simulation results using low rank approximation and linear prediction for frequency estimation verify the analysis. The same method of analysis can be applied to a broad class of parameter-estimation methods in which the principal-component technique or low rank approximations to matrices are used

Keywords

filtering and prediction theory; matrix algebra; parameter estimation; probability; signal processing; Cramer-Rao bound behavior; SNR; frequency estimation; linear prediction; low rank approximation; matrix approximation; outlier probability; parameter estimation errors; principal-component technique; signal processing algorithms; singular-value decomposition; threshold analysis; Algorithm design and analysis; Analytical models; Least squares approximation; Matrix decomposition; Parameter estimation; Predictive models; Probability; Signal analysis; Signal processing algorithms; Signal to noise ratio;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 1988. ICASSP-88., 1988 International Conference on

Conference_Location

New York, NY

ISSN

1520-6149

Type

conf

DOI

10.1109/ICASSP.1988.197129

Filename

197129