Abstract: Noise affects strongly the detection of signals and target recognition. A lot of literatures have been printed on the linear noise reduction for the underwater target radiated signal. For linear systems, linear filtering methods are compatible. However, these noise reduction methods are not fit for the signal from nonlinear systems or the signal affected severely by the nonlinear factor.
Non-linearity of signals radiated by the Underwater target have been proved in the latest literature. This presentation provided the noise reduction results with the nonlinear method. Irregular signals from nonlinear sources exhibit genuine broad band spectra and there is no justification to identify any continuous component in the spectrum as noise. Instead, structure in the reconstructed phase space will be exploited. This paper presents the difference between the reconstructed attractor of experimental time series and the noise reduced time series. Also, the difference of the correlation dimension and the signal-to-noise between experimental time series and the noise reduced time series is provided. According to the result, the nonlinear noise reduction method is effective for the underwater target radiated noise time series.