Research On Signal Saperating And Filtering Based On Franctional FOURIER Transformation

Fourier transformation did great help to communication and signal processing, but since Fourier analysis,which is a transform in full-time domain,can't describe the signal property in time-frequency domain, such as Linear frequency-modulated (LFM,chirp) signal that is often applied in radar, sonar and communications system is nonstationary signal,in which the frequency varieswith time.Since Fourier analysis,which is a transform in full-time domain,can't describe the signal property in time-frequency domain,Fractional Fourier transformbecomes one of the new time-frequency analysis theories now. Fractinal Fouriertransform is Fourier transform in broad sense,which is the representation of the signalin Fractional Fourier domain.As optical equipment carry out Fractional Fourier transform easily,the application of Fractional Fourier transform in optical signal processing come true first.With the presentation of Digital computation of Fractional Fourier transform,it has begun to apply to electric signal processing.On the basis of study of Fractional Fourier transform and its digital computation, an franctional polarization filtering technique is proposed in this thesis. The proposed method can suppress the interference perfectly,while does not introduce distortions in amplitude and phase of the target when the po-larized states are estimated exactly. Furthermore, it is also a valid tool when the target and the interference hold the same polarized angle. Basic principles of the proposed method are demonstrated. We also show the franctional polarization filtering is an extension to conventional ones and the NPSPF. Simulation and numerical results are shown to illustrate the performance of the proposed method.