Localization of small surface vessels through acoustic data fusion of two tetrahedral arrays of hydrophones

Abstract

Detection and tracking of vessels is important in confined areas such as marine parks or harbors. Nowadays, the presence of ships can be accurately monitored either by radar or via AIS system, while small vessels, which have weak radar signature, may be easily missed. The paper presents the detection and localization algorithms optimized for small- and mid-sized boats and based on data either from a single underwater sensor station of four hydrophones, or from data fusion between two hydrophone volumetric arrays. Each platform hosts a sparse tetrahedral array of broadband hydrophones and pan, tilt, compass and depth sensors. Both acoustic and non-acoustic data from the two stations are transferred to shore, where they are stored and processed on a PC. The basis of localization algorithm is the crosscorrelation between pairs of hydrophones along time (crosscorrelogram). The wavevector estimation of a vessel from each tetrahedron is achieved through Least Mean Square method. Adequate data association algorithms allow the fusion of estimates obtained from each array in order to provide precise and robust tracking of each vessel. At-sea results demonstrate the system capability for detecting and localizing small vessels in a shallow-water harbor environment. [Work partially funded by EU within ARGOMARINE Project].