Research And Implementation Of Radar Signal Intra-pulse Modulation Type Recognition

With the increase of the radar signal density and the modulation type in the space,modern electromagnetic environment is also becoming increasinglycomplex.In order to achieve high reliability with reliably identify to distinguish and intercept the radar signals,besides the traditional five parameters,the modulation characteristics of intra-pulse analysis and recognition of enemy radar signal modulation are also needed.This paper studies the recognition of radar signal intra-pulse modulation.Firstlythe recognition methods are put forward based on the time-frequency analysis.Then MATLAB simulations verify the correctness of the methods.Finally realize one of the methods through the "Wideband Receiving and Signal Sorting and Tracking Processing Board",which is applied in the engineeringproject.This paper thoroughly studies the recognition of radar signal intra-pulse modulationmainly by the following parts:Firstly,the theory and properties of short-time Fourier transform(STFT)and fractional Fourier transform(FRFT)are detailedlyexpounded.Secondly,the recognition method which is based on the STFT is studied.A coarse to fine recognition method is adopted.Firstly,recognize the radar signal as phase modulated(PM)signal and frequency modulated(FM)signal to realize the coarse recognition.For the fine recognition,the PM signal will be recognized as normal radar signal,BPSK signal and QPSKsignalby means of specific implementation.TheFM signal will be recognized as FSKsignal,LFMsignal and NLFMsignal.For the extraction of the instantaneous frequency of FM signal,this paperproposes an algorithm which is to add a median filter after instantaneous phase difference method.The simulation shows that this method improves the precision of frequency measurement.Thirdly,the recognition of modulation types of radar signals based on FRFT is studied,which is realized by using decision tree based on the different features.Byextracted normalization FRFT envelope curve feature to classify the radar signal into two major categories,one class is frequency modulated signal and the other one class is PSK signal and FSK signal.Then,by setting the kurtosis threshold will obtain the fine recognition result separatelyin two major categories.For LFM signal and NLFMsignal,there is a large overlapin kurtosis under low SNR.To solve the overlapperformance under low SNR,thelocal featureextraction of normalization FRFT envelope curve is proposed,which is called the difference of peak point.By setting the difference of peak point threshold to obtain the fine recognition result offrequency modulated signal under low SNR.The validity of the new method is certified by the recognition probabilitycontrastsimulation.Finally,the recognition method based on STFT is implemented by FPGA throughthe hardware platform "Wideband Receiving and Signal Sorting and Tracking Processing Board".Recognize the modulation type of the output signal by the wideband digital channel receiver.After software design and the coding of Verilog program,through the Modelsim simulation verifies its function.Then,implement on-line simulation through the SignalTap II logic analyzer in Qaurtus II,which verifies the availability and practicability of the method in the engineering project.