Digital Processing and Spectral Estimation of Ultrawideband Signals by Atomic Functions and Wavelets

In this article the new approach based on combination of atomic functions (AF) and classical spectral kernels is proposed and proved. Shown that obtained structures have some advantages in comparison with spectral wave analysis of ultrawideband (UWB) signals. The new approach based on application of spectral kernels to problems digital UWB signal processing is considered. In theory of communication signals are represented in the form of a linear combination of elementary functions. Application of spectral analysis of UWB signals in basis of digital functions in comparison with the spectral harmonious analysis gives certain advantages which are defined by the physical nature of representation of signals. Optimal processing in spectral area in comparison with the time possesses has advantages on application of numerical algorithms. It is known that spectral signal processing possesses a high noise stability to action of the noise which own basis differs from own functions of a signal. The physical characteristics and analysis of numerical experiment confirm efficiency of new spectral kernels and WA-systems in tasks of spectral estimation and digital UWB signal processing.