John Heidemann

Self-Configuring Localization Systems: Design and Experimental Evaluation

TitleSelf-Configuring Localization Systems: Design and Experimental Evaluation
Publication TypeTechnical Report
Year of Publication2002
AuthorsN. Bulusu, J. Heidemann, D. Estrin, and T. Tran
Date Publishedsep
Institutionucla-cens
Abstract

Embedded networked sensors–-those that coordinate amongst themselves to achieve a sensing task–-promise to revolutionize the way we live, work and interact with the physical environment. Fundamental to such coordination is \emphlocalization, or the ability to establish spatial relationships among such devices. In very large, ad hoc deployed sensor networks, a localization system based on beacons (special nodes that are position-aware by virtue of being endowed with more sophisticated ranging hardware) can be used to localize smaller devices consistently even in a completely decentralized and scalable manner. However, in unattended sensor networks, these localization systems must \emph self-configure, i.e., autonomously adapt to the dynamics of their environmental setting and the availability of beacons, instead of relying on extensive pre-configuration or manual reconfiguration. In this paper, we present the motivation, design, implementation and experimental evaluation of a self-configuring localization system based on beacons. We identify \emphdensity as an important parameter in determining localization quality, and propose HEAP and STROBE, two algorithms to enable system self-configuration based on beacon density. Building on the observation that the quality of localization saturates at a transition beacon density, these algorithms (i) automate placement of new beacons at low densities to significantly improve localization quality or (ii) rotate functionality amongst redundant beacons at high beacon densities to significantly increase overall system lifetime. Our performance results include experimental results from implementation of an RF-proximity based localization system using nodes with sharply limited resources (8-bit microprocessor, 8-K ROM, 512 bytes RAM, limited battery life).

URLhttp://www.isi.edu/%7ejohnh/PAPERS/Bulusu02c.html
Groups: