Excitation conditions for signed regressor least mean squares adaptation

Author

Sethares, W.A. ; Mareels, Iven M Y ; Anderson, Brian D O ; Johnson, C. Richard, Jr. ; Bitmead, Robert R.

Author_Institution

Sch. of Electr. Eng., Cornell Univ., Ithaca, NY, USA

Volume

35

Issue

6

fYear

1988

fDate

6/1/1988 12:00:00 AM

Firstpage

613

Lastpage

624

Abstract

The stability of the signed regressor variant of least-mean-square (LMS) adaptation is found to be heavily dependent on the characteristics of the input sequence. Averaging theory is used to derive a persistence of excitation condition that guarantees exponential stability of the signed regressor algorithm. Failure to meet this condition (which is not equivalent to persistent excitation for LMS) can result in exponential instability, even with the use of leakage. This persistence of excitation condition is interpreted in both deterministic and stochastic settings

Keywords

filtering and prediction theory; least squares approximations; signal processing; stability; LMS; averaging theory; deterministic setting; excitation condition; exponential instability; exponential stability; filtering; high data rate applications; input sequence; leakage; least mean squares adaptation; persistence; signal processing; signed regressor algorithm; stability; stochastic settings; Adaptive filters; Echo cancellers; Finite impulse response filter; Least squares approximation; Quantization; Speech processing; Stability; Stochastic processes; Strontium; Systems engineering and theory;

fLanguage

English

Journal_Title

Circuits and Systems, IEEE Transactions on

Publisher

ieee

ISSN

0098-4094

Type

jour

DOI

10.1109/31.1799

Filename

1799