Multi-band decomposition of the linear prediction error applied to the least-mean-square method with fixed and variable step-sizes

Author

Rasende, F.G.V. ; Diniz, Paulo S R ; Tokuda, Keiichi ; Kaneko, Mineo ; Nishihara, Akinori

Author_Institution

Dept. of Phys. Electron., Tokyo Inst. of Technol., Japan

Volume

4

fYear

1997

fDate

9-12 Jun 1997

Firstpage

2176

Abstract

New least-mean-square adaptive algorithms based on multi-band decomposition of the linear prediction error are presented. In the first part of this paper the mathematical background for a stochastic approach to the multi-band decomposition-base scheme, where time and frequency resolutions are traded off along the frequency domain by using different step-sizes in the analysis of distinct frequencies, is presented and a least-mean-square algorithm with fixed step-sizes is derived. In the second part, an algorithm with independent variable step-sizes to allow time and frequency resolutions to be traded off in both time and frequency domains is introduced. Computer experiments compare the performance of multi-band and conventional least-mean-square methods when applied to autoregressive spectral estimation

Keywords

adaptive signal processing; least mean squares methods; prediction theory; spectral analysis; adaptive algorithms; autoregressive spectral estimation; decomposition-base scheme; fixed step-sizes; frequency resolutions; least-mean-square method; linear prediction error; multi-band decomposition; stochastic approach; time resolutions; variable step-sizes; Adaptive filters; Art; Costs; Frequency selective surfaces; Gas insulated transmission lines; Least mean squares methods; Least squares approximation; Petroleum; Reduced instruction set computing; Wiener filter;

fLanguage

English

Publisher

ieee

Conference_Titel

Circuits and Systems, 1997. ISCAS '97., Proceedings of 1997 IEEE International Symposium on

Print_ISBN

0-7803-3583-X

Type

conf

DOI

10.1109/ISCAS.1997.612751

Filename

612751