Quasi-LFM radar waveform recognition based on fractional Fourier transform and time-frequency analysis

Recent advances in electronics have increased the complexity of radar signal modulation. The quasi-linear frequency modulation (quasi-LFM) waveforms (LFM, Frank code, P1-P4 code) similar time-frequency distributions, and it is difficult to identify such signals using traditional analysis methods. To solve this problem, paper proposes an algorithm for automatic recognition quasi-LFM based on fractional Fourier transform analysis. First all, Wigner-Ville distribution (WVD) are used determine number main ridgelines tilt angle target component WVD. Next, standard deviation component's width signal's WVD calculated. Finally, assembled classifier neural network built recognize different by automatically combining three features. Simulation results show that overall rate proposed reaches 94.17% under 0 dB. When training data set test mixed with noise, 89.93%. best accuracy achieved when size taken as 400. can meet requirements real-time recognition.