The article constitutes a set of research referring to the removal of the noise of Gaussian distribution of signal using wavelet analysis. The thesis has been launched of presentation of existing analysis algorithms, starting of discrete transform related to Mallat’s algorithm, and ending with the selected methods on inverse wavelet transform. The task was carried out by using the simulation scripts of package Wavelet Toolbox in MATLAB. The studies of using on-line filtering algorithms has been undergone. The delay value, which was implemented by the various filtration methods, was determined through simulations. The inclusion of the comparison of results allows to evaluate the quality of filtration.