QUAD: Quadrant-based relative location estimates for representative topologies in wireless sensor networks

• Authors: May Wong;Demet Aksoy
• Venue: Computer Networks: The International Journal of Computer and Telecommunications Networking
• Year: 2009

Sensor networks consist of devices that make various observations in the environment and communicate these observations to a central processing unit from where users can access collected data. In this regard, users' interpretation of collected data highly depends on the reported location of the sensor making an observation. GPS is an established technology to enable precise location information when deployed in open field. Yet, resource constraints and size issues prohibit its use in small sensor nodes that are designed to be cost-efficient. Instead, locations are estimated using a number of approaches. To date, however, the focus of such estimations was based on individual accuracy of sensor locations in isolation to the complete network. In this paper, we discuss problems with such approaches in terms of data management and analysis. We propose a novel location estimation algorithm called QUAD, quadrant-based localization, to enable representative topology information. In particular, QUAD makes use of relative distances from landmark points to determine the quadrant a node resides in and refines estimations according to neighbour provided information. QUAD makes use of uncertainty levels in estimates to further assist data analysis. Our experiment results suggest significant improvements in individual accuracy prior to optional refinements. Drastic improvements are achieved in the overall topology using refinements.