“An Algebraic Framework for Query Optimization in Sensor Networks"

Friday, February 25, 2005
11:00 a.m. – 12:00 noon
Room 404, IS Building

You are invited to meet the speaker before the talk
from 10:30 AM – 11:00 AM in the Large Commons Room of the IS Building

Abstract: Recent advances in wireless communications and micro-electronics enabled wide deployment of smart sensor networks. Sensor networks support a broad range of applications that involves system monitoring and information tracking (e.g., airport security infrastructure, monitoring of children in metropolitan areas, product transition in warehouse networks, fine-grained weather/environmental measurements, etc.). Such applications require efficient mechanisms for querying the sensor data and delivering the query result in a timely manner. Meanwhile, resource limitations of sensor nodes imply that excessive transmissions in response to sensor queries can lead to premature network death.

In this talk we introduce an integrated approach to sensor query optimization that utilizes performance and functional trade-offs between query processing schemes and lower layer transmission protocols. Specifically, we develop novel cross-layer optimization techniques that utilize information about how the medium access control layer operates while processing queries in large scale sensor environments. The optimizer uses a Data Transmission Algebra that uniformly captures the structure of data transmissions, and their constraints and requirements. Our framework enables both qualitative analysis and quantitative cost-based optimization of sensor queries. In particular, we demonstrate efficiency of our approach in vibration-based Structural Health Monitoring, where sensor network continuously examines the structure under critical integrity conditions.

Bio: Vladimir Zadorozhny is an Assistant Professor in Department of Information Science and Telecommunications, School of Information Sciences, University of Pittsburgh. He received his Ph.D. in 1993 from the Institute for Problems of Informatics, Russian Academy of Sciences in Moscow. Before coming to US he was a Principal Research Fellow in the Institute of System Programming, Russian Academy of Sciences. Since 1998 he worked as a Research Scientist in the University of Maryland Institute for Advanced Computer Studies at College Park. He joined University of Pittsburgh in September 2001. Vladimir's current research interests include scalable architectures for wide-area environments with heterogeneous information servers, networked information systems, query optimization in resource-constrained distributed databases, wireless data management.