UC San Diego

There is increasing interest in deploying swarms of underwater vehicles for marine surveys. In such surveys, the cost, endurance, and utility of the vehicle swarm needs to be carefully optimized. One of the main challenges when designing these systems is coming up with an appropriate way to localize each vehicle in relation to the rest of the swarm. Current methods for underwater localization are restrictive in either cost, power consumption, or range.

This dissertation considers the self-localization of a deforming swarm of subsurface vehicles using ambient acoustic noise in the ocean. The experiments presented consider a group of independent underwater vehicles that passively record ambient sounds in the ocean with a single hydrophone while they float with subsurface currents. Three different self-localization approaches are considered. The first involves estimating the acoustic impulse response between moving vehicles using cross-correlations of ambient noise, a known first step towards a self- localization estimate. Accurate estimates of the acoustic impulse response are shown between moving vehicle pairs. However, motion between the receivers limits the amount of time averaging that can be done, making the estimation susceptible to anisotropies in the ambient noise field. To overcome these anisotropies, the next approach jointly estimates the vehicle geometry and the directionality of the ambient noise field, without prior knowledge of either estimate. This creates a viable method for estimating the vehicle geometry on short time scales using correlations of low-frequency noise in the ocean. Results are shown for a deployment carried out off of the coast of La Jolla. Last, a self-localization approach using impulsive noise from snapping shrimp is considered. Impulsive sound sources provide high intensity, broadband signals that facilitate accurate arrival time detections by each vehicle. However, the similarity between different signals presents a significant correspondence problem, which must be solved to provide accurate estimates of the changing geometry of the swarm. A geometric solution to this correspondence problem is shown and an estimation procedure is proposed to track the geometry of a swarm as it changes. The self-localization estimates are compared to estimates from an accurate acoustic localization system with good agreement.