1 / 5

Overview: Active Sensor networks

Overview: Active Sensor networks. Philip Levis, David Gay, and David Culler, Active Sensor Networks . In Proceedings of the Second USENIX/ACM Symposium on Networked Systems Design and Implementation (NSDI 2005) Enabling dynamic programming on the sensor system

bockd
Download Presentation

Overview: Active Sensor networks

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Overview: Active Sensor networks • Philip Levis, David Gay, and David Culler, Active Sensor Networks. In Proceedings of the Second USENIX/ACM Symposium on Networked Systems Design and Implementation (NSDI 2005) • Enabling dynamic programming on the sensor system • Recaliberate calibration after deployment • Post processing to reduce network traffic

  2. Problem • Sensor networks are deployed in areas where we don’t know apriori what to expect • Dynamically reprogram sensors: change what they sense • Question: How do we reprogram • send the new native code - could be large • send a script - new code depends on the script language • Code for a single VM (Mate) - not flexible • Limited functionality, not extensible for specific application requirements • JavaVM - too expensive for some applications • Send code as application specific virtual machine • Small size code, nearly as fast as native code • Extensible type support, concurrency control, code propagation

  3. Design • Each VM is made of a template • Operations: primitives (language specific), functions (language independent)

  4. Data model • Stack architecture • No program data beyond stack • Scheduler: FIFO thread scheduler • Concurrency manager: manage multiple handler threads to avoid race conditions • Threads only allowed to run when all resources are available • Reboots when new code arrives to reset all variable • Capsule store: propagation • Selective execution, everyone has code, only some nodes execute it • Code is trickled to other nodes using version vectors • Version packets, capsule status packets, capsure fragments

  5. Evaluation • Flexibility: Languages: Tinyscript, Motlle, Applications: RegionsVM, QueryVM • Performance: With 6% of hand coded and 20% energy saving • Low overhead • Sensor CPU are mostly idle - not necessary to be as efficient as the native code

More Related