Networked GPS approach to tracking marine animal schools

There has been a great need to track marine animal schools in order to study their seasonal and regional behaviors, to study species diversity and gene flow, to protect endangered species, and to study pelagic environments and global climate change. However, existing means, such as GPS tags, do not suffice because of their low accuracy, low availability, and low integrity. This paper presents a networked GPS approach to tracking marine animal schools. The marine animals are tagged with GPS receivers. The GPS receivers in a marine animal school are networked via acoustic links and cooperatively calculate their positions. We formulate cooperative positioning as a high-dimension maximum likelihood estimation problem, and derive an iterative solver based on linearization. The simulation results validate our proposed positioning algorithm and further show that our proposed approach does not only achieve better positioning accuracy, but also enables underwater positioning (better availability) and intra-school relative positioning.