Title :
Compensation for saturation effects in FIR-NLMS system identification
Author_Institution :
Dept. of Signals, Sensors & Syst., R. Inst. of Technol., Stockholm
fDate :
2/19/1998 12:00:00 AM
Abstract :
An adaptive scheme is derived to be used in combination with the normalised least mean squares algorithm for finite impulse response models (FIR-NLMS) to combat nonlinear imperfections of the plant to be identified. The scheme is shown to outperform FIR-NLMS when a test system that consists of a variable static nonlinearity in cascade with a linear filter is identified
Keywords :
adaptive estimation; identification; least mean squares methods; FIR-NLMS system identification; adaptive scheme; finite impulse response model; linear filter; nonlinear imperfections; normalised least mean squares algorithm; saturation effects; variable static nonlinearity;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:19980311
