Detection of broadband signals by their spectral features

Abstract

The paper is about a method for detecting wideband signals on a noise background, which uses the symmetry property of its spectrum. The authors of the article proposed the method. An analysis of literary sources has shown that the detection of WBS is relevant for electronic warfare and information security tasks, and the speed of detection is very important. Also, based on a literature review, it was concluded that the energy detection method is not effective enough, including due to the need for long-term accumulation of signal energy. As the SNR deteriorates, the required accumulation time increases. In paper proposes to use the fact that the energy spectra of most WBS, in particular phase-shifted one, are symmetrical with respect to their central frequency.

The proposed method involves applying a Fast Fourier Transform to the analyzed mixture of signal and noise. Next, a function is equal to the sum of the products of spectral samples taken symmetrically with respect to the expected central frequency calculated. It is analytically shown that such processing makes it possible to detect the PM WBS on a white Gaussian noise background of based on the characteristic maximum of this function. If this function is calculated only for noise (without a signal), then the maximum does not arise, since white noise has a uniform spectral density.

If the value of the average frequency of the signal is not known, then, according to the proposed algorithm, it is necessary to perform several cycles of calculations for different estimated average frequencies. Thus, both the presence of a signal in the mixture and the value of its central frequency are detected. Computer modeling, the results of which are also presented in the paper, shows that the use of the proposed method can reduce the accumulation time and increase resistance to noise exposure, compared to the energy detection method.