IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 26, NO. 9, DECEMBER 2008 1 Energy-Efficient Ranging for Post-Facto Self-Localization in Mobile Underwater Networks Diba Mirza, Student Member, IEEE, and Curt Schurgers, Member, IEEE Abstract—Many ocean processes, both biological and phys- ical, greatly depend on and interact with the intrinsic current dynamics of the underwater environment. A promising approach to understand small and large scale spatio-temporal correlations of these processes is to deploy a networked swarm of drifters that float freely with ocean currents. They form a coordinated distributed sampling system that observes ocean processes within their own moving frame of reference. As data interpretation is impossible without knowledge of sampling positions, a method is required to localize the drifters. Furthermore, the localization has to be repeated periodically as the network topology changes due to the inherent motion of the vehicles. This paper proposes a novel energy-aware, distributed solution based on inter-drifter range measurements. It leverages the realization that actual position es- timation can be performed after the mission is over. The proposed broadcast-based solution achieves sufficient localization accuracy with an extremely low overhead: around 0.5 transmissions per node per localization. Index Terms—Distance measurement, energy conservation, networks, underwater vehicle detection and tracking. I. I NTRODUCTION T HE PHYSICAL and biological processes occurring within the earth’s oceans greatly impact the climate and life on the planet in general. However, monitoring the oceans and the processes within has always been precarious, due to their vastness and difficulty to access. To help scientists in their exploration, we have begun working on a networked system of free-floating underwater explorers [1]. A rendering of what we envision this system to look like is shown in Fig. 1. Each individual device is a drifter that is subject to the motion of the currents. Compared to anchored instruments, such drifters are able to observe phenomena in their own moving frame of reference. Equipped with a multitude of sensors, each one essentially collects samples in its own local neighborhood. To obtain a large-scale view of the phenomena and their correla- tions, these drifters are not simply deployed individually, but as a swarm. In essence, they form a distributed and dynamic sampling system that is carried by the currents. However, to act as a sampling system, the drifters need to know their positions, so that spatial correlation of sampled data can be deduced. Because of the large area occupied by the swarm and their uncontrollable mobility, it is impractical to assume that nodes will always be in range of fixed localization Manuscript received March 6, 2008; revised . This work was supported in part by the NSF under award number ECCS-0622005. The authors are with the Electrical and Computer Engineering Department, University of California, San Diego, 9500 Gilman Drive, La Jolla, CA 92093- 0407 USA (e-mail: diba@ucsd.edu; curts@ece.ucsd.edu). Digital Object Identifier 10.1109/JSAC.2008.0812xx. Fig. 1. Underwater network of autonomous drifters. buoys. Those that are not will have to obtain their position through a process known as self-localization [2]. In self- localization, devices communicate with each other in order to estimate the distances between them, from which their positions can be deduced. For this purpose, we have networked our drifters using underwater acoustic modems. Transmitting data via such a modem consumes considerable power. Since our drifters are battery operated, energy efficiency is a crucial design constraint. This is the challenge we tackle in this paper: design of a self-localization scheme for underwater drifters that is extremely energy efficient. Our solution, which we have dubbed Sufficient Distance Map Estimation (SDME), achieves this goal by exploiting the unique properties of our networked swarm. The key ingredient is the realization that positions only need to be calculated after the mission is over and the devices have been retrieved. Indeed, the large amounts of sensor data will be interpreted by scientists only after the mission. So, it is sufficient to also calculate positions at this time. This means that SDME does not need to be a full “localization” scheme. While submerged, the drifters only need to collect distance estimates, a step referred to as ranging. The estimation of the actual drifter positions from ranging data can be done post-mission. However, SDME does not even need to be a “ranging” scheme in the traditional sense, as devices do not have to find distances while submerged. Instead, the only thing really needed, is that all the necessary information to perform the distance estimation is collected and stored somewhere in the network. The concrete challenge and design goal of SDME is to find which device needs to collect what data, while minimizing the negotiation between them, all when subjected to uncontrollable current motion. 0733-8716/08/$25.00 c  2008 IEEE