Adaptive denoising of low SNR signals

Abstract

In 1992 David Donoho introduced the term denoising in connection with the adaptive nonlinear filtering in the wavelets transform domain. Despite its advantages this method has not be used yet in the communications field. The goal of this paper is to improve this method and to apply it in communications. The improvements are a new threshold's value searching method and the use of a new discrete wavelet transform. This new transform enhances the diversity in the wavelets transform domain. The results of a simulation prove the performances of the new denoising method.