Passive Detection of Scooter's Underwater Radiated Noise

Acoustic detection of submerged vessels or submarines by small low-cost devices arranged in an internet of underwater things (IoUT) offers distributed sensing with easy deployment from moorings, vessels or air. In this paper, we propose a probabilistic approach to detect the underwater radiated noise (URN) of submerged scooters. Our method builds upon expected distortions in the radiated noise of the scooter that leads to a complex harmonic pattern. A stability test for harmonics then verifies detection. Results from two sea experiments with two different scooters show favorable tradeoff between the false alarm rate and the detection rate that improves as the vehicle's speed increases.