| Increasingly dense of electromagnetic signals in modern battlefield, makes the radarsignal recognition technology based on the five conventional parameters unable to meet thereal needs. Therefor it’s an urgent need to explore new recognition methods to improve thetechnological level of existing radar against equipments in the electronic warfare.This paper was mainly interested in the theory and method of intentional modulation andunintentional modulation of radar signal recognitions.Considering two typical intentional modulation of radar signal recognition, Phase ShiftKeying (PSK) signal and Frequency Modulation (FM) signal, a classification and recognitionmothed based from rough to the detailed was proposed in this paper. Considering thesignificantly different characteristics of the3dB bandwidth of frequency spectrum betweenPSK signal and FM signal, firstly apply rough classify and recognize. On this basis, for thedetailed characteristics of PSK signal and FM signal, a mothed based on wavelet ridgefrequency characteristics was proposed in this paper, PSK signal was in detail classified andrecognized as the Binary-Phase Shift Keying (BPSK) signal and Quarternary-Phase ShiftKeying (QPSK), FM as Linear FM (LFM) signal and Nonlinear FM (NLFM), thuscompleting two typical intentional modulation of radar signal recognition at certain signalnoise ratio (SNR).Considering unintentional modulation of the radar pulse signals’ conventional pulserepetition interval (PRI) characteristics, this paper applies the time of arrival (TOA) differencemethod for recognizing radar emitter individuals. Considering unintentional modulation of theradar pulse signals’ carrier frequency offset characteristics, this paper applies fast fouriertransform (FFT) method for recognizing radar emitter individuals. Considering unintentionalmodulation of the radar pulse signals’ waveform characteristics, based on the analysis of theactual measured pulse, this paper respectively extract from the time domain characteristicsand frequency domain characteristics to apply the Hausdorff distance method to recognize.Above-mentioned, the result was given in the corresponding probability of correctrecognition. Considering unintentional modulation of radar signal characteristics caused by the phasenoise, the bispectrum analysis mothed based on transform domain was proposed this paper,and the signal bispectrum calculation method and its physical significance were given.Secondly, surrounding-line integral method was applied to optimize and extract characteristicsfrom the results of bispectrum estimate, also quantify and accumulate characteristic extraction.Considering cumulative characteristic quantity, barycenter distance method and FuzzyC-means (FCM) clustering method were respectively applied to establish radar emitterindividual recognition template. Finally, computer simulation experiment results verify thecorrectly recognizing probability of based on the template to recognize radar emitterindividual at the certain SNR, and the characteristic extraction method of this paper is verifyby using data from the actual measured experiment. The performance of bispectrumestimation of radar signal is analysed at the end of this paper. |
PDF Full Text Request