Vibrational and hydroacoustic signal processing in the frequency domain and its software-hardware implementation

The paper discusses an algorithm for spectral density estimation in the frequency domain using wavelet-based smoothing by wavelet thresholding techniques. The suggested algorithm can be applied to vibrational and hydroacoustic signal processing in order to estimate signal parameters in the frequency domain. The algorithm can be applied to the Fourier periodogram without using signal samples in the time domain. We also propose a technique aimed at working with signals of arbitrary length by applying the maximal overlap discrete wavelet transform. We also study software-hardware implementation of the developed algorithms.