CENS Technical Seminar Series

Collection Routing and Application-Informed Energy Management in Sensor Networks

Invited Speaker: Omprakash Gnawali
Date: January 23, 2009
Time: 1:00 PM - 2:00 PM
Venue: Boelter Hall 4760

Abstract

In the last few years, we have built a general-purpose and dynamically re-taskable sensor network system called Tenet and deployed it on a variety of application scenarios. In this talk, I will describe my recent work on Collection Tree Protocol (CTP) and Application-informed energy management (AEM) in the context of Tenet. CTP is a routing protocol that allows the nodes to form a routing tree along which the data packets are forwarded from the sensor nodes to the sinks in a network. CTP uses a link quality estimator called the 4-bit estimator that uses information from the physical, link, and network layers to accurately estimate wireless link quality. CTP uses the Trickle algorithm and result from data packet transmissions to make routing path discovery and repair quick while remaining efficient over the long term. The second work, AEM, performs static analysis of Tenet tasks to characterize the expected traffic profile. AEM then tunes the radio duty-cycling parameters to provide efficient radio duty-cycling while accommodating the application traffic profile. AEM is the first radio duty-cycling system that works under these constraints: dynamic multi-hop routing and link qualities, dynamic tasking, multiple and concurrent applications, and reliable end-to-end delivery of data.

Biography

Omprakash Gnawali is a Ph.D. candidate in computer science at the University of Southern California. He has worked on routing protocols, wireless link quality estimation, radio duty-cycling and network architectures of sensor networks. Previously, he has also worked on Peer-to-peer networks. He received his Bachelors and Masters degrees from the Massachusetts Institute of Technology.