No abstract

The paper 1 presents experiments on the use of higher order statistics in the procedures of identification of radar emitters. Studies aimed at testing "sensitivity" of higher order statistics of selected characteristics of the individual sources of emissions. Attention is focused on methods based on analysis of parameters of single-pulse radar emissions. I. INTRODUCTIONMethods of analysis of signals based on higher order statistics (HOS) -took its beginning in the early 70s, from the economics and geophysics. The development of technologies supporting computational techniques made the range of applicability of higher order statistics is steadily expanded. The most common use of HOS are connected with issues of noise reduction, identification non minimalphase systems, nonlinear systems analysis, multidimensional signal processing, biomedical signal processing. HOS techniques are techniques with spreading possibilities with respect to statistics, the lower order [2][3][4]. Tasks requiring the application of techniques first or second order can be formulated in terms of higher order statistics. Significant advantage of higher order statistics respectively to the first and second ones, are included in the range information extracted from the signals using particular transformation.Many of the signals which we deal in practice with, cannot be transformed with sufficient accuracy using the traditional methods of second order (such as power spectrum). When determining the power spectrum of the signal only the information on phase shifts between components is lost. There are a lot of applications when this information has a key role. HOS are very efficient techniques enabling to solve that problem.The paper presents experiments use of higher order statistics in the procedures of identification of radar emitters. Studies aimed at testing "sensitivity" of higher order statistics of selected characteristics of the individual sources of emissions. Attention is focused on methods based on analysis of parameters of single-pulse radar emissions. The paper presents a brief introduction to themes of higher order statistics and some results of the ongoing transformation of the simulated and real signals. The presented results justify the thesis about the potential usefulness of higher order statistics in the procedures of identification of radar emitters

No abstract

ELINT/ESM type of electronic intelligence in the primary layer uses parameters measurements of intercepted radar signals. Nowadays modern radar uses more and more complex waveforms. Some waveforms are developed intentionally to make their intercept almost impossible. The main distinctive features of modern radar signal are hidden in its time-frequency structure. In the near past the problem of radar signal feature extraction was considered in time or frequency domain separately, because radar waveforms were relatively simple. Today, however, the signals should be observed simultaneously in both domains. Timefrequency distribution concept offers a new approach in radar signals classification/identification. The paper presents some results of compressive concept and Hough transform application to intra-pulse modulation analysis of radar signals. Linear frequency modulation within the pulse was considered.

No abstract