Weak Signal Detection Based On Chaos Theory

Chaotic motion, which exits widely, can describe the very typical behavior of many nonlinear systems. Chaos has been paid wide attention because of its good intrinsic properties, and it has been widely and successfully applied to many fields. Chaotic oscillator detection for weak signal belongs to an important application of chaos theory in information science. Chaos system is sensitive to weak signal and immune to noise, which make chaos have good prospects and a great progress in signal detection technique.The idea of weak signal detection based on the chaotic oscillator is discussed in the paper, also a way of judging whether or not periodic signal exists in target signal based on the phase space portrait has been proposed. LCE (Lyapunov Characteristic Exponent) is employed into the field of chaos detection for weak signal, and it is used to get the critical value of the system and to judge the state of the weak signal detection system as a quantitative criterion. The methods for computing LCE of deterministic system and chaotic time series are researched. The theory of phase space reconstruction is introduced and the choice methods of embedded dimension and delay time are analyzed. Meanwhile the method of chaotic time series prediction is argued in detail. Based on chaos theory, the RBFNN (Radial Basis Function Neural Network) is used to build one-step prediction model. On the foundation of all the above, taking Logistic and Lorenz chaos signal as background, a series of emulation experiments are carried on, which validate the feasibility of the method and carry out the weak signal detection of chaos. At last, taking the reverberation, which has properties of chaos, as background, a series of emulation experiments of simulated echo signal detection are carried out.