Recognition Of Unknown Signal In The Complex Electromagnetic Environment Based On FRFT

In the future information warfare, numerous electronic weapons and equipments willbe widely used in a relatively limited space in the battlefield. With the addition of exsitingelectromagnetic radiation, a special electromagnetic environment will be formed which iscalled complex electromagnetic environment. There are many kinds of intense signals inthe complex electromagnetic environment. Thus, in order to improve the anti-interferenceability of the electronic weapons and equipments, the jamming signals need to be detectedand identified precisely and quickly. However, the signals in the complex electromagneticenvironment are usually non-stationary and there are serious time frequency couplingbetween different signals. As a result, the traditional signal processing methods can not bequalified for detecting and recognizing signals in the complex electromagneticenvironment.As a newly emerged signal processing method, fractional Fourier transform (FRFT) issuitable for dealing with non-stationary signals. What’s more, FRFT is a lineartransformation and there is no cross-term interference. So fractional Fourier transform canachieve good performance when processing multi-component time-frequency signals. Thispaper applies FRFT to the recognition of unknown signal in the complex electromagneticenvironment. The fast detection methods and parameter estimation of typicaltime-frequency signals based on FRFT are studied in depth. The main contributions of thisthesis are summarized as follows.1. The fundamentals and properties of FRFT is studied and analized. On that basis, thesignal detection method based on fractional Fourier transform is researched, especially themethod based on the2-D peak searching and the one based on the fourth order originmoment of fractional spectrum (OMFrS). The performance of these methods is comparedthrough simulations. 2. On the basis of analyzing the practical signal form, the estimation error of theexisting parameter estimation method is derived and then a novel and universal parameterestimation method is proposed. Furthermore, the proposed algorithm is developed whichallows estimation of the practical observed Gaussian windowed chirp signal.3. The fast detection method of the optimal transform angle of FRFT is studied andtwo new algorithms are proposed. The first method is realized based on sub-Nyquistsampling. The relationship between the optimal transform angle of the undersampledsignal and that of the original signal is studied. It is proved that the chirp-rate of noiselessChirp signal can be estimated correctly even though the signal is undersampled. Thesecond method is based on the symmetric property of the fourth order OMFrS of chirpsignal. By deducing the difference between the theoretical optimal transform angle and thedetected result, a more accurate estimation can be obtained. Compared to the existingmethods, the proposed algorithm achieves better accuracy with high detection speed.4. The existing detection methods based on FRFT is usually only detects one Chirpcomponent that has the largest energy accumulation in each iteration. Thus, it will needseveral iterations to detect multicomponent Chirp signals. In order to improve the detectionefficiency, the shading relations between the fourth order OMFrS of different Chirpcomponent is studied. On this basis, a new threshold setting method for rapid detection ofmulticomponent Chirp signals based on the fourth order OMFrS is proposed. Byemploying the proposed method, several LFM components whose fourth order OMFrS attheir own optimal rotation angles are not shadowed by strong components can be detectedat the same time in one iteration.5. The connotative meaning of complex electromagnetic environment and itsconstruction principle is studied. On this basis, a simulation system of complexelectromagnetic environment is realized by adopting the computer simulation. Using thecomplex signals produced by the simulation system, the recognition method of unknownsignal in the complex electromagnetic environment using FRFT which is proposed basedon the works of this thesis is verified.