The binary algorithm for the analysis of the spectrum amplitude and recover of harmonic components signals in the presence of uncorrelated background noise

The problem of detecting and distinguishing between the informative components of multicomponent signals in the presence of additive uncorrelated noises. The solution of this problem is based on harmonic analysis Fourier series using the sign-function analog-stochastic quantization of the investigated signal. This kind of quantization makes it possible to carry out extremely coarse two-level quantization without a systematic error, regardless of the statistical properties of the analyzed signals. A developed algorithm for harmonic analysis provides a high computational efficiency. The practical implementation of these algorithms basically reduces to performing the simplest arithmetic operations of summation and subtraction of the discrete values of the filtering functions calculated at the instants of time when the result of sign-function analog-stochastic quantization changes its sign. For the decision about the detection of harmonic components, we use a criterion for exceeding of a threshold value by estimating the amplitudes of the harmonic components. The results of experimental studies show that the use of sign-function analog-stochastic quantization has allowed to effectively eliminate the effect of the additive uncorrelated noise and to obtain high quality recover of continuous multicomponent signals useful.