DocumentCode
334802
Title
Covariant estimators of time-frequency descriptors for nonstationary random processes
Author
Scharf, Louis L. ; Friedlander, Benjamin ; Thomson, David J.
Author_Institution
Colorado Univ., Boulder, CO, USA
Volume
1
fYear
1998
fDate
1-4 Nov. 1998
Firstpage
808
Abstract
For a nonstationary random process, the time-time correlation function, frequency-frequency Loeve spectrum, time-frequency Wigner-Ville distribution, and frequency-time ambiguity function are each complete theoretical descriptions of second-order behavior. They are complete in the sense that they determine realizations of the random process according to the Cramer-Loeve spectral representation for a harmonizable process. In this paper we derive estimators for these descriptors by requiring them to have the same covariant dependence on time delay and complex modulation as do their theoretical counterparts.
Keywords
Wigner distribution; correlation methods; covariance analysis; delays; frequency modulation; parameter estimation; random processes; signal representation; spectral analysis; time-frequency analysis; Cramer-Loeve spectral representation; covariant dependence; covariant estimators; frequency modulation; frequency-frequency Loeve spectrum; frequency-time ambiguity function; harmonizable process; nonstationary random processes; second-order behavior; time delay; time-frequency Wigner-Ville distribution; time-frequency descriptors; time-time correlation function; Contracts; Delay effects; Delay estimation; Equations; Fourier transforms; Random processes; Time frequency analysis;
fLanguage
English
Publisher
ieee
Conference_Titel
Signals, Systems & Computers, 1998. Conference Record of the Thirty-Second Asilomar Conference on
Conference_Location
Pacific Grove, CA, USA
ISSN
1058-6393
Print_ISBN
0-7803-5148-7
Type
conf
DOI
10.1109/ACSSC.1998.750972
Filename
750972
Link To Document