DocumentCode :
3243171
Title :
Parametric analysis methods of time-variant waveform and its pattern
Author :
Takizawa, Yumi ; Fukasawa, Atsushi
Author_Institution :
Commun. Syst. Lab., OKI Electr., Toyko, Japan
Volume :
5
fYear :
1992
fDate :
23-26 Mar 1992
Firstpage :
149
Abstract :
Time-variant signal analysis is described from a viewpoint of waveform. A parametric waveform analysis method is first proposed based on a model which effectively reflects the characteristics of time-variant signals. A parametric pattern analysis method is also proposed based on mutual relationships among the vibration components of a waveform. The proposed method has been confirmed through evaluations using artificially composed modulation signals to give improved tracking capability and reduced waveform approximation error compared to the Prony method. The derivation of pattern parameter has been found to give useful information on features of a waveform
Keywords :
modulation; signal processing; waveform analysis; modulation signals; parametric pattern analysis; parametric waveform analysis method; signal analysis; time-variant signals; time-variant waveform; tracking; vibration components; waveform approximation error; Approximation error; Bandwidth; Damping; Frequency; Information analysis; Pattern analysis; Performance analysis; Research and development; Signal analysis; Wavelet transforms;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1992. ICASSP-92., 1992 IEEE International Conference on
Conference_Location :
San Francisco, CA
ISSN :
1520-6149
Print_ISBN :
0-7803-0532-9
Type :
conf
DOI :
10.1109/ICASSP.1992.226636
Filename :
226636
Link To Document :
بازگشت