Bilkent University
Department of Computer Engineering


Scalable Self-Healing for Wireless Sensor Network Applications


Murat Demirbas
Computer Science and Engineering
SUNY Buffalo

Self-healing is the ability of a system to recover from faults and restore itself to normal operation without human intervention. Since wireless sensor networks are inherently fault-prone, scalable self-healing is crucial for enabling the deployment of large-scale sensor network applications. My work on "local-healing" addresses the scalability of the cost-overhead of self-healing: By confining the contamination of the network due to faults, this approach achieves healing within work and time proportional to the perturbation size instead of the network size. My work on "specification-based healing" addresses the scalability of the design effort for adding self-healing: By using only the specification of an application and avoiding its implementation details for the addition of self-healing, this approach yields efficient design of self-healing even for large implementations. In my talk, using a tracking application as a running example, I will present samplers from my work in these two directions.

Bio: Murat Demirbas received his B.Tech. from Middle East Technical University, Turkey in 1997, and his Master's and Ph.D. degrees from The Ohio State University in 2000 and 2004, respectively. While at the Ohio State University Murat was involved in the development and deployment of a 100 node wireless sensor network, ``Line In The Sand'', for detection, classification, and tracking, which paved the way to the ``ExScal'' network with 1000 nodes. After a one-year post-doc with the Theory of Computing Group at MIT, Murat is currently an assistant professor at the Computer Science and Engineering Department at SUNY Buffalo. His research interests are in the area of distributed systems, sensor networks, and fault tolerance.


DATE: May 17, 2006, Wednesday@ 13:40