Analysis of the possibilities of application of a nonparametric criterion for investigation of noisy pulse signals

The paper presents the results of applying the algorithm based on the nonparametric Wilcoxon criterion for evaluating individual parameters of noisy pulse signals. The signals were sawtooth pulses with different values of the amplitude and duty cycle, as well as rectangular pulses with different amplitude values and their repetition period. Gaussian noise was chosen as the additive noise. As a result of applying the algorithm, the original noisy signal without preliminary filtering was transformed into a random rank function. The behavior of a rank function is characterized by the presence of extrema in it at the points of analysis. In this case, the differences in the rank functions make it possible to identify characteristic points by which it is possible to estimate individual parameters of the pulse signal, and the estimation accuracy depends on the signal-to-noise ratio.