Extracting Sinusoidal Frequency Modulation Signal Signature Using Fourier-Bessel Transform and Short-Time Fraction Fourier Transform

A method using Fourier-Bessel transform (FBT) and short-time fraction Fourier transform (ST-FRFT), for accurate and robust parameter estimate and separation of sinusoidally frequency modulated (SFM) nonstationary signal, is presented. The rotation micro-motion coupling with the rigid body's motion provides an approximately sinusoidal phase modulation, which is known as the micro-Doppler (m-D) signal. According to priori research, the SFM signal shows the nearly linear frequency modulation characteristic in the local part and sparsity features in fraction domain. Firstly, in order to extract the slope of nearly FM and the rotation target's angular frequency, here using FBT method can decompose the rotation target's echo into stationary and nonstationary component. Then, using ST-FRFT method separate the target's signal into the fraction domain where through peak search to catch the SFM signal's slope and rotation angular frequency. This function shows enormous potential in solving the problem that nonstationary component always can't be extracted because of weak echo amplitude. The rotation angle frequency is an important characteristic of the rotating target, which shows great significance to the classification of low-altitude and low-speed small targets. This proposed method has been demonstrated by the air target's simulation data which shows effective results of estimating these parameters.