Detection And Parameters Estimation Of LFM Siganl Based On Stochastic Resonance

The LFM (linear frequency modulation) signal is a kind of classic Nonstationary signal in the information system which includes wireless communication, radar, sonar and ultrasound system. The detection of LFM signal at low SNR (Signal to Noise Ratio) has become significant isusses. The main way to analysis LFM signal is the traditional time-frequency representation which can make full use of the time-frequency characteristics of LFM signal, but these algorithms have large amount of computation and storage because they require two dimensional search. They also have another drawbacak of the detection failure in the ultra-low SNR.With the development of nonlinear system, Stochastic Resonance (SR) has become a novel signal detection method in the areas of weak signal detection of method. SR has changed the tradition that suppressing or eliminating the noise, in constrast it uses noise energy to achieve the detection of weak signals. This paper applies SR to weak signal detecting, and presents a new method for frequency and chirp rate detection of single-component LFM signal based on SR. The main work and contributions of the dissertation are in the several aspects as follows:1. The basic principles of signal detection using SR are introduced, and the numerical simulations of detection algorithm are achived. When using SR to detect weak signal, it encounters three issues:high frequency signal detection, non-periodic signal detection and adaptive SR establishing. The three points are detailed descripted and the corresponding processing algorithms are given.2. The effects of system parameters and noise to signal resonance are analysised and two adaptive parameters selection algorithms are introduced. Ganetic Algorithm (GA) is applied to periodic signal detection to achieve the adaptive selection of the parameters of SR. Simulation results show the effectiveness of the algorithm.3. Time domain, frequency domain characteristic of LFM signal and the typical time-frequency analysis method—Wigner-Ville distribution (WVD) are studied. The basic principles of signal detection using Fractional Fourier transform (FRFT) and dechirp are introduced. Detection of LFM signal based on SR is presented, and the simulations of this algorithm using two LFM signals are achieved. The influences of the times of average output spectrum and the numbers of signal segments to the algorithm are dicussed. Finally, the computational analysis of the algorithm is given.4. For bistable systems conneted in series, its basic principles are analysised. Two bistable systems conneted in series are used to detect the LFM signal and estimate its parameter. Numerical simulations show two bistable systems conneted in series can affect the basic detection algorithm. Two bistable systems conneted in series futther improve the detection and estimated parameter of LFM signal based on SR.