The Detection And Parameter Estimation Of The Linear Frequency Modulated Signals And Nolinear Frequency Modulated Signal In Low SNR

It is very noticeable to the Linear Frequency-Modulated signal(LFM) in the Time-varied signals. As a spreading frequency signal with Generalized Time-Bandwidth Product, it has been applied widely to many fields of information system,such as communications, radar, sonar, earthquake recognition etc. The Nolinear Frequency-Modulated signal has been also applied widely to these fields.By analyzing its parameters, we can identify its kind and model. the high polynomial phase of sigal received by the radar receiver reflects the movement character of the target relatived to the radar. In communications, AM/FM signal and other Phase-Modulated signals can be expanded to the high polynomial phase signal(PPS) and be approximated by finite order PPS.This paper first interprets simply the basic theory involved in it ,and introduces mainly Wigner-Hough Transform and Fractional Fourier Transform. Then it goes deep into research detailedly the Detection and Parameter Estimation of the Multi-Linear Frequency Modulated signals and Nolinear Frequency Modulated Signal in Low SNR. 1) Research the detection of the long and frequency-varied tardily LFM (LFV-LFM) signal in Low SNR. Because the movement rules of the peak of LFV-LFM on the multiple WHT plane assimilates with that of the peak of point target on the multiple infrared image,this paper puts forward a new algorithm.It can detect the long LFM in Low SNR which makes use of the detection way of image processing about exiguous infrare point target combined with WHT. 2) This paper proposes a new algorithm of the parameter estimation and the signal separation for Multi-component LFM signal with the Wigner-Hough transform (WHT) ,which can picks up many peak value from the WHT plane at a time and estimate the parameters of the each LFM signal and separate them combined with the Dechirp and Narrowband filter. The illusive pick can be eliminated by the ways about self-adapting filter and energy integration in the local field of the peak and autocorrelation function. At the same time, this paper discusses detailedly amplitude estimation when it is time-varing amplitude. 3) Researchs the Detection and Parameter estimation of the Multi-source Linear Frequency Modulated signals by Fractional Fourier Transform(FRFT). First, this paper proposes respectively the concentration Detection and the Kurtosis Detection algorithm for LFM based on FRFT. At same time, it analyses detection probability and false alarm probability of the algorithm. Finally,it researchs Parameter estimation of the Multi-source Linear Frequency Modulated signals by FRFT. 4) Using Fractional Fourier Transform, the window of Short-time fourier Transform is optimized. This paper dissertates the optimization theory and Generalized Time-Bandwidth Product and estimates the instantaneous frequency(IF) of Nolinear Frequency Modulated Signal. Finally, academic analysis and experimentation emulate for the estimate error of IF is realized in detail. 5) Using Fractional Fourier Transform and the minimum Generalized T-F domain support,three parameters of gause chiplet small wave such as scale, time-slope, frequency-slope are optimized,which makes the parameter optimization of gause chiplet small wave transforms into the parameter optimization of the gause-Modulated window fuction. At same time, based on the character gause-Modulated window optimized by FRFT, Multi-windows is introduced,which can estimate well the instantaneous frequency(IF) of Multi-source Nolinear Frequency Modulated Signals ,and make the T-F concentration of them more fine.