Sensor networks are notoriously difficult to program, given that they encompass the complexities of both distributed and embedded systems. To address this problem, we present the design and implementation of a declarative sensor network platform, DSN: a declarative language, compiler and runtime suitable for programming a broad range of sensornet applications.
We demonstrate that our approach is a natural fit for sensor networks by specifying several very different classes of traditional sensor network protocols, services and applications entirely declaratively – these include tree and geographic routing, link estimation, data collection, event tracking, version coherency, and localization. To our knowledge, this is the first time these disparate sensornet tasks have been addressed by a single highlevel programming environment.
We address a number of systems challenges that arise when building a generic compiler and runtime environment for the sensornet context; these include not only issues of limited resources, but also the management of asynchrony and requirements of predictable execution. Our results suggest that the declarative approach is well-suited to sensor networks, and that it can significantly improve software productivity and quality while still producing efficient, resourceconstrained code.
Download Latest Release
The platforms supported are TelosB, Tmote Sky, micaZ and mica2 (although it would not be difficult to add support for other platforms already supported by TinyOS).
You can access the latest tutorial here.
David Chiyuan Chu, Lucian Popa, Arsalan Tavakoli, Joseph M. Hellerstein, Philip Levis, Scott Shenker and Ion Stoica. The Design and Implementation of A Declarative Sensor Network System. The 5th ACM Conference on Embedded networked Sensor Systems (SenSys 2007), Sydney, Australia, 6-9 Nov, 2007. Also see: Technical Report No. EECS-2006-132, EECS Department, University of California, Berkeley, 2006. [pdf]
This is a good place to start for understanding the rational behind DSN, the declarative language Snlog along with many example applications, as well as the architectural and implementation challenges of the DSN system.
Tavakoli, David Chu, Joseph Hellerstein, Philip Levis, and
Scott Shenker. A Declarative Sensornet Architecture.
Workshop on Wireless Sensor Network Architecture (WWSNA 2007),
Cambridge, Massachusetts, April 25-27, 2007. [pdf]
This work explores the architectural benefits and concerns raised by a declaratively constructed sensor network from both a practical and philosophical angles.
David Chu Arsalan Tavakoli Lucian Popa Joseph Hellerstein. Entirely Declarative Sensor Network Systems. In Proceedings of the 32nd International Conference on Very Large Data Bases. Seoul, Korea. September 12 - 15. Demonstration. [pdf]
We demoed our system at VLDB 2006 running a multi-hop tracking application akin to real sensor network tracking applicaiton deployments.
David Chu and Joseph M. Hellerstein. Automating Rendezvous and Proxy Selection in Sensor Networks. In Proc. Eigth International Conference on Information Processing in Sensor Networks (IPSN 2009). San Francisco, California, April 13-16, 2009. [pdf]
This research demonstrates how high-level declarative protocols can be automatically optimized based on network state to make efficient decisions about rendezvous points and proxy selection.
David Chu. Building and Optimizing Declarative Networked Systems. Ph.D. Dissertation. University of California, Berkeley, May 2009. [pdf]
In addition to covering the above topics in more detail, Chapter 4 of the dissertation develops various cross-layer optimizations achievable in declarative sensor networks.
David Chu. A Declarative Sensor Network Architecture. 2nd iCAST/CMU/TRUST Joint Conference on Security and Privacy Technologies. Taipei, Taiwan, June 4-5, 2007. [ppt]
David Chu. Declarative Sensor Networks with Applications in Landslide Detection. iCAST/CMU/TRUST Joint Conference, 2007. Crossing Boundaries: International Themes for Computer Security in Academia, Industry, and Services. Taipei, Taiwan, January 8-10, 2007. [ppt]
David Chu. Declarative Sensor Networks. UC Berkeley EECS Database Seminar. Berkeley, CA, March 2, 2007. (also gave similar talk at Syslunch Seminar, November 28, 2006.) [ppt]
Students: David Chu, Lucian Popa, Arsalan Tavakoli
Professors: Joseph M. Hellerstein, Philip Levis, Scott Shenker, Ion Stoica
We always welcome talking to UC Berkeley undergraduates looking for research or development opportunities. Please email davidchu @t cs if you are interested.
P2 - a system for Internet-scale declarative networking
TinyOS - a popular operating system for sensor networks