Envelope spectrum analysis of underwater target radiated noise based on harmonic wavelet

Traditional methods can hardly extract the modulated information of underwater target radiated noise accurately. In this paper, a new analysis method of broadband noise envelope based on harmonic wavelet is proposed because it has advantages of band-pass filtering and Hilbert transform in the frequency domain. By using harmonic wavelet transformation, the target radiated noise will be decomposed with multi-resolution and high-precision, so that the frequency band which contains the strongest modulation will be obtained, and the modulated information can be demodulated easily. The experiment results of the oceanic signal show that, compared with the traditional envelope analysis methods, the proposed method can extract the modulated information of target radiated noise faster and more accurately, which will be helpful to improve the underwater target radiated noise signal detection and feature extraction.