Symbolic Models of Non-Stationary Radar Signals

Abstract

Symbolic models of non-stationary radar signals and their possible application in radar systems are given. In comparison with traditional methods of regression analysis and various mathematic transforms (Fourier, Laplace, etc.) an approach based on computational intelligence is considered. The proposed approach to forming a symbolic model is a method of transforming a symbolic image of non-stationary radar signals based on the feature of scintillating radar images perception. The proposed transform of symbolic images of non-stationary radar signals of moving aerial objects with interperiod scintillating fluctuations from a certain correspondence of the asymptotic equality of perception of radar images changing in time and space, is reduced to the statement about the conditions of simple equality of perception of radar images that have different frequencies of fluctuations. The data obtained are considered as a mixture of true and various interfering signals, which are fuzzy samples and sets. Thus, it is possible to use an intelligent perception approach to determine the functional dependencies and relations of the mixture of radar images. The dependencies and relations obtained in the form of symbolic models of radar signals make it possible to determine distinctive features for detecting and recognizing the aerial objects in the presence of interferences. The paper shows how this approach can be used to improve the intelligent analysis of radar data by generating a pulse-frequency code of fluctuations of symbolic models of radar images and smoothing scintillating fluctuations.