Secure Localization and Tracking in Sensor Networks
Abstract
Localization and tracking of objects of interest are two canonical issues in sensor networks research. When the object of interest is static, we use localization algorithms to identify its location. When the object of interest is moving, we use tracking algorithms to estimate its path over time. Since sensor networks are often deployed in remote or hostile terrains, however, security becomes another critical issue. Hence the localization or tracking accuracy would go down as a result of the presence of malicious nodes. The objective of this dissertation is to correctly identify the malicious nodes during the localization and tracking processes. A novel algorithm based on relaxation labeling is presented to achieve this objective. Our approach provides a different perspective from the existing literature on secure localization and tracking. Current literature uses statistical measures to perform localization and tracking as accurately as possible given the in presence of malicious nodes. Instead, those malicious nodes are isolated first, and use only data from benign nodes to perform localization and tracking. Both simulations and field experiments are used to demonstrate the performance of our algorithm.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2008
- Accession Number
- ADA504623
Entities
People
- Chih-chieh G. Chang
Organizations
- North Carolina State University