Weak Signal Detection Method Based on Duffing Oscillator.

Junguo Wang,Jianzhong Zhou,Bing Peng

DOI: https://doi.org/10.1108/03684920910994015

IF: 2.352

2009-01-01

Kybernetes

Abstract:Purpose - The purpose of this paper is to detect the periodic signal tinder strong noise background, and estimate its amplitude/phaseDesign/methodology/approach - Melnikov method is adopted as calculating the threshold value when chaos occurs, and the detected signal is taken as a system parameter The system's output state is changed if the parameter has a slight change near the threshold Meantime, the phase of system's output is recognized to judge whether the output state changes, and the signal parameter is estimated according to the necessary condition.Findings - A small periodic signal in noise can be detected by Duffing oscillator via a transition from chaotic motion to periodic motionResearch limitations/implications - The paper shows how to calculate the amplitude/phase in low signal-to-noise ratiosPractical implications - The Duffing system is sensitive to the weak periodic signal and has definite immunity to noise, so it is easy to construct a system composed of many oscillators that could process complex signals, even though the environmental noise is intenseOriginality/value - This paper presents a nonlinear method for detecting and extracting the weak signal.