Design of nonlinear adaptive digital filter by wavelet shrinkage
Tomoya Tokairin, Yoshiharu Sato · Electronics and Communications in Japan (Part III Fundamental Electronic Science) · 2002
Abstract The wavelet shrinkage method is a nonlinear signal denoising procedure to remove noise by shrinking the empirical wavelet coefficients in the wavelet domain. In this paper, we apply the wavelet shrinkage method to nonlinear adaptive digital filters, and we propose the adaptive wavelet shrinkage filter which updates the threshold levels adaptively by the LMS algorithm according to the reference signal. The adaptive wavelet shrinkage filter is robust and has the property that one can control the filter easily by modifying the threshold levels. We describe the theorems concerning the convergence of the LMS algorithm for updating the threshold levels, and demonstrate several examples of the adaptive wavelet shrinkage filter. © 2002 Wiley Periodicals, Inc. Electron Comm Jpn Pt 3, 86(2): 59–71, 2003; Published online in Wiley InterScience ( www.interscience.wiley.com ). DOI 10.1002/ecjc.10032