DocumentCode :
3096424
Title :
Modeling arbitrary polynomial bispectra using systems with multiplicity in one- and two-dimensions
Author :
Erdem, A. ; Tekalp, A.
Author_Institution :
Dept. of Electr. Eng., Rochester Univ., NY, USA
fYear :
1990
fDate :
10-12 Oct. 1990
Firstpage :
197
Lastpage :
201
Abstract :
Introduces a new class of systems, called systems with multiplicity (SWM), to model arbitrary polynomial bispectra. It is shown that an arbitrary polynomial bispectrum of a one-dimensional (1-D) process can always be modeled using an SWM with FIR components. Experimental results for the identification of SWM based on their output bispectra using a proposed method are then presented. In two-dimensions (2-D), it is shown that an arbitrary polynomial bispectrum cannot always be modeled using an SWM with 2-D FIR components, although a given SWM with 2-D FIR components can be identified using the bispectrum of its output.<>
Keywords :
filtering and prediction theory; polynomials; spectral analysis; 1-D process; 2-D process; FIR components; arbitrary polynomial bispectra; modelling; systems with multiplicity; Autocorrelation; Finite impulse response filter; Impedance matching; Large scale integration; Polynomials; Power system modeling; Signal representations; Sufficient conditions; Two dimensional displays; White noise;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Spectrum Estimation and Modeling, 1990., Fifth ASSP Workshop on
Conference_Location :
Rochester, NY, USA
Type :
conf
DOI :
10.1109/SPECT.1990.205574
Filename :
205574
Link To Document :
بازگشت