Research On Methodology Of Weak Signal Detection Based On Chaos Theory

Weak signal detection has been widely used in numerous fields such as communication and radar.While the weak signal detection under the background of strong noise has been an important research hotspot in modern information theory and has urged people to continue exploring and developing the new theories and new methods.Though there are some limitations of the traditional time domain signal detecting methods,especially the method is often limited by the SNR threshold.In recent years,with the study of chaos theory in nonlinear science,it has provided a novel viewpoint to resolve problem.The detection method based on the chaos theory has overcome the shortcomings of traditional methods.Compared with conventional schemes,chaos-based detection methods can detect lower signal to noise ratio of the signal so it provide a new research theory and methods.Based on the analysis of chaotic dynamical system,taking Duffing-Holmes mapping as a research subject.LCE(Lyapunov Characteristic Exponent)is used to judge the state of chaotic system as a quantitative criterion.The QR decomposition method is used to find the threshold value of dynamic system from chaotic state to periodic state.The basic principle and detection method of weak signal detection by using chaotic Duffing oscillator are further analyzed,and the feasibility of detecting the detection signal based on phase trajectory to recognize whether the target signal is contained is verified.To detect unknown frequency of the signal,improving the Duffing-Holmes system by using the sliding mode variable structure in control theory.The simulation result show that the improved chaotic Duffing system can effectively suppress the noise and measure the frequency of the weak signal through the power spectrum of the system.Making use of the support vector machine theory,genetic algorithm and particle swarm optimization algorithm,a one-step prediction model is established to predict the short-term chaotic signal.The phase space reconstruction parameter and the support vector machine model parameters are combined to optimize together.A prediction model is built with the optimal parameters,and the chaos time series is used to verify the accuracy of the model.At the same time,the chaotic and weak signals are combined to compare with the traditional parameters calculating methods.The results show that the prediction performance of the new method is improved obviously.