A new fast algorithm for system identification

Author

Mattis, W.E.

Author_Institution

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

fYear

1991

fDate

13-16 Oct 1991

Firstpage

1381

Abstract

A recursive least mean square technique for system identification is outlined, and applied to a recursive system, to determine an algorithm for updating the coefficients, or weights, of such a system. The effects of the convergence coefficients on stability are detailed, in addition to the all important synthesis in hardware. Of particular interest are the synthesis of this system with a charge-coupled device (CCD), to produce the delayed version of the signals, and the effect of the charge retention in the cells of the CCD on convergence and coefficient or weight changes. The application of this system to a telephone network is discussed with regard to echo cancellation of long distance telephone calls

Keywords

adaptive filters; convergence of numerical methods; echo suppression; filtering and prediction theory; identification; iterative methods; signal processing; telephone lines; adaptive filters; charge-coupled device; convergence coefficients; echo cancellation; long distance telephone calls; recursive least mean square technique; signal processing; stability; system identification; telephone network; Charge coupled devices; Convergence; Delay effects; Echo cancellers; Hardware; Network synthesis; Signal synthesis; Stability; System identification; Telephony;

fLanguage

English

Publisher

ieee

Conference_Titel

Systems, Man, and Cybernetics, 1991. 'Decision Aiding for Complex Systems, Conference Proceedings., 1991 IEEE International Conference on

Conference_Location

Charlottesville, VA

Print_ISBN

0-7803-0233-8

Type

conf

DOI

10.1109/ICSMC.1991.169881

Filename

169881