Using spatial correlation of ocean current for velocity estimate of underwater drifting nodes

摘要

The ability to navigate underwater is a key requirement for most underwater applications. Mobile devices, drifters, and also human divers, all require precise navigation capability to perform long term missions while being submerged. In the absence of GPS reception, performing underwater navigation (UN) requires an accurate state-space-model (SSM) to allow the tracked node (TN) to self-estimate its location. However, irregularities in nodes motion (mostly due to unpredictable changes of ocean current) makes it hard to reliably determine the SSM. In this paper, we rely on spatial correlation of ocean current, and propose to estimate the drift velocity of the TN as a combination of the drift velocities of anchors, and to directly use it as part of the SSM. This not only augments the amount of information available for the TN to track its own position, but also offers an unbiased velocity estimate which increases the reliability of the SSM. Since ocean current may not be always correlated (for example, in the case of turbulence), we offer two unbiased confidence indexes: one which is based on the range to the anchor, and a second which is based on the homogeneity of the current velocity field. To evaluate the potential of utilizing node spatial dependencies for UN, we collect trajectories of drifting nodes from both model-based simulations and a sea trial performed in Israel. Our results suggest that nodes drifting motion shows strong spatial correlation which could greatly enhance the performance of tracking algorithms.