Weak signal detection based on chaotic oscillator

The weak signal detection based on chaotic oscillator is analyzed, and the weaknesses of the existing Melnikov method and the phase trajectory observation method are pointed out in this paper. A new weak signal detection method based on Lyapunov exponents (LE) is proposed. Firstly, the model based LE calculation method is used to determine the accurate initial threshold value. Secondly, an improved LE calculation algorithm is proposed based on phase space reconstruction of the observed data. The characteristics of this improved method are the high precision of the threshold value setting and the automatic recognition. Simulation results verify analysis and the effectiveness of this method.