Global LMS estimate of wavelet coefficients for signal denoising

Wavelet transform is a powerful tool for signal denoising. Also, some wavelet domain filters were presented, which can be used to obtain an optimal estimate of the wavelet coefficients. In this paper, a global least mean-squared error estimate of the wavelet coefficients is developed, which can yield the least mean-squared error estimate of the original signal. In addition, an optimal threshold for the soft thresholding is introduced.