Improvement of the Detection of Envelope Modulation on Noise (DEMON) and its application to small boats

The Detection of Envelope Modulation on Noise (DEMON) algorithm is popular for the identification of vessels from passive acoustic data. We suggest several improvements to the DEMON algorithm: First, an extension of the DEMON algorithm for determining time delay of arrival of an acoustic signal based on the phase delay of DEMON spectral components from two hydrophones. Then we show the DEMON spectra of a variety of small vessels measured in controlled conditions. Their various characteristics are discussed in the context of vessel classification from DEMON spectra. Finally, the dependence of the DEMON spectra on the passband of propeller noise were investigated using the Cyclic Modulation Spectrum, and optimum frequencies for demodulation were found.