DocumentCode
3102805
Title
High-order spectra of mixture processes
Author
Dwyer, Roger F.
Author_Institution
US Naval Underwater Syst. Center, New London, CT, USA
fYear
1988
fDate
3-5 Aug. 1988
Firstpage
186
Abstract
Summary form only given. In certain processes higher-order spectra reveal more information than second-order spectra. For example, the bispectrum has been applied to determine deviations from Gaussianness, to determine phase information in nonGaussian linear processes, and in nonlinear interactions. The trispectrum has also been used to extract phase information from nonGaussian signals. For the present study, the application is to determine whether or not a nonGaussian process is present from its higher-order spectrum. Initially, a nonGaussian process is modeled as a parameterized mixture process. Then the theoretical spectra and higher-order spectra are derived based on the mixture process. Simulations are shown of the measured spectra and higher-order spectra from mixture processes. The results are pertinent to the modeling of underwater acoustic signals.<>
Keywords
acoustic signal processing; information theory; random processes; spectral analysis; underwater sound; Gaussianness; bispectrum; higher-order spectra; mixture processes; nonGaussian linear processes; nonlinear interactions; parameterized mixture process; phase information; trispectrum; underwater acoustic signals; Acoustic measurements; Acoustic scattering; Acoustic waves; Data mining; Gaussian processes; Markov processes; Multidimensional systems; Signal processing; Underwater acoustics;
fLanguage
English
Publisher
ieee
Conference_Titel
Spectrum Estimation and Modeling, 1988., Fourth Annual ASSP Workshop on
Conference_Location
Minneapolis, MN, USA
Type
conf
DOI
10.1109/SPECT.1988.206188
Filename
206188
Link To Document