The predictable trajectory of underwater gliders can be used in geographic routing protocols. Factors such as drifting and localization errors cause uncertainty when estimating a glider's trajectory. Existing geographic routing protocols in underwater networks generally assume the positions of the nodes are accurately determined by neglecting position uncertainty. In this paper, a paradigm-changing geographic routing protocol that relies on a statistical approach to model position uncertainty is proposed. Our routing protocol is combined with practical cross-layer optimization to minimize energy consumption. Our solution's performance is tested and compared with existing solutions using a real-time testbed emulation that uses underwater acoustic modems.
Baozhi Chen, Patrick C. Hickey, Dario Pompili