On computing and implementing the running bispectra

Author

Perry, Richard J. ; Amin, Moeness G.

Author_Institution

Dept. of Electr. & Comput. Eng., Villanova Univ., PA, USA

Volume

43

Issue

4

fYear

1995

fDate

4/1/1995 12:00:00 AM

Firstpage

1017

Lastpage

1021

Abstract

The paper extends recursion to bispectrum estimation problems and presents the systolic array implementation of the recursive higher order spectrum in which the bispectrum estimate is updated every data sample. Forward and reverse sequence running Fourier transforms are first systolically realized. The results are then used to drive a second systolic array, whose outputs represent the FT of the data third-order moment. The proposed systolic arrays have no global communications with a number of processing elements independent of the size of the employed 2D lag window

Keywords

Fourier transforms; VLSI; digital signal processing chips; pipeline processing; recursive estimation; sequences; spectral analysis; systolic arrays; 2D lag window; bispectrum estimation problems; data third-order moment; forward sequence running Fourier transforms; recursion; recursive higher order spectrum; reverse sequence running Fourier transforms; running bispectra; systolic array implementation; Adaptive signal processing; Fourier transforms; Frequency estimation; Least squares approximation; Least squares methods; Recursive estimation; Signal processing; Signal processing algorithms; Speech processing; Systolic arrays;

fLanguage

English

Journal_Title

Signal Processing, IEEE Transactions on

Publisher

ieee

ISSN

1053-587X

Type

jour

DOI

10.1109/78.376859

Filename

376859