Real-time passive underwater localization using a compact acoustic sensor array

Abstract

This paper proposes an underwater passive localization system that can be installed on remote platforms and can locate in real-time sound sources using a compact 4-element array. The localization system relies on both the Time Difference of Arrival (TDoA) algorithm to localize sources that are impulsive, and a low-complexity implementation of the matched field processing (MFP) algorithm for broadband continuous waves, such as propeller noise. The array is intended to be mounted on small platforms and compact geometries are investigated to optimize the localization accuracy. Figures of merit are defined to assess the accuracy and reliability of the algorithm as a function of the distance between the elements. The proposed algorithm is implemented on a System-on-Chip (SoC) which executes the algorithm on an embedded system for real-time operation. Finally, the performance is assessed in realistic ocean conditions in different environments. It is shown that the MFP provides an excellent angle estimate for broadband continuous sources, but TDoA is more reliable, particularly when the deployment is sensitive to the environment conditions.