Acoustic beacon aided underwater vehicle navigation: Observability analysis on vehicle trajectory

Abstract

In the GPS-denied underwater environment, unmanned underwater vehicle navigation commonly depends on integrating the outputs of an inertial measurement unit (IMU) and Doppler Velocity Sonar (DVS) for recursive self-localization. However, dead reckoning navigation based on only an IMU and DVS suffers from the accumulation of errors over time. Analogous to using a GPS signal from a satellite, an acoustic signal from a source from a known location can be used to measure time-of-flight (distance) and motion-induced Doppler (relative velocity) to correct for these errors. However, due to the non-linear relationship between the pose and acoustic measurements, analyzing the observability of the positioning system in a traditional sense is nontrivial. For this work, we aim to analyze and design the trajectory of a vehicle that operates in order to ensure a form of observability. This approach provides a sufficient condition for the existence of a one-to-one relationship between the state of the system and the observed series of measurements. By analyzing this form of observability over possible trajectories, our goal is to find a vehicle path plan that can yield a unique pose based on measurements from the onboard sensors and received acoustic signals.