Abstract

By an optimization algorithm, the signal reconstruction process is optimized using multiresolution analysis with the Daubechies mother wavelet and the Sure-Shrink thresholding algorithm. This work seeks to propose an optimization scheme and a threshold associated with the noise variance, which complements the signal reconstruction process and achieves better performances when comparatively analyzing the signal to noise ratio, the mean square error, and the norm, compared to the signal obtained from the unoptimized process.