Joint analysis on radiated noise of underwater target using wavelets and higher-order cumulants

The joint analysis of wavelets and higher-order cumulants is first used to analyze underwater target radiated noise in the paper. First, we use wavelet transform to extract the dynamic modulation of radiated noise, then a FFT transform of fourth-order cumulants' diagonal slices is used to analyze harmonic features of this process. The results show that the method can effectively separate the dynamic modulation from other parts of radiated noise, and extract harmonic features. Because fourth-order statistics can restrain the gaussian noise, the method has high stability in noise conditions. Moreover, comparing with bispectrum and trispectrum, 21/2-spectrum analysis evidently decreases calculational amount, costs little computer time and can be easily performed.