Intentional Modulation Analysis And Fingerprint Feature Extraction For Radar Signal

Pulse modulation characteristic of radar signal is the most important performance for radarsignal subtle characteristics, and is more subtle and more profound form of mathematicalphysics described for signal, relative to the most basic feature of pulse compression radar signal.Pulse modulation characteristics include pulse intentional modulation and pulse unintentionalmodulation. Pulse intentional modulation is that compression radar purposefully modulatepulse signal by the methods of frequency, amplitude and phase. In the modern field ofelectronic countermeasures, this research has been more and more attention. Pulseunintentional modulation characteristic (also known as fingerprint feature) mainly depends onthe modulator, the radar system, the transmitter type, power and oscillating tube and it is aninherent feature of the radar transmitter. The radar emitter fingerprint feature extraction has alsobecome a key technology to study electronic countermeasures system in recent years.This paper studies the following contents: to research intentional modulationcharacteristics; to extract fingerprint feature of different radiation sources under the samepattern and model.For radar signal intentional modulation characteristics, this paper mainly studies:two-phase coding (BPSK) signal, the four-phase encoding (QPSK) signal, linear frequencymodulation (LFM) signal and non-linear FM signal (NLFM), and analyzes their intentionalmodulation form. The paper proposes a method that can identify PSK signal and the FM signalbased on the signal power spectrum3dB bandwidth. The modulation parameter of the BPSKsignal and QPSK signal are estimated by using the section of the circulation spectral, namelycarrier frequency and chip-rate. We can estimate carrier frequency by searching the maximumpeak and estimate chip-rate by calculating the difference between maximum peak and thesecondary large peak. And the NMSE of carrier frequency estimation and chip-rate estimationis simulated by the computer. We can use cyclic spectrum analysis method and set the thresholdto identify BPSK signal and QPSK signal. We can use short-time Fourier method to identifyLFM signal and NLFM signal.For the radar signal unintentional modulation characteristics, this paper studies three emitter signal model about CW signal, BPSK signal and LFM signal, and analysis of eachunintentional modulation method. Bispectrum graph and contour graph of CW signal withphase noise, BPSK signal with phase noise and LFM signal with phase noise are simulated. Wecan use bispectrum slicing method and contour integral bispectrum estimation method toextract specific emitter fingerprint feature, and these methods can be implemented by thecomputer.