|Stochastic Systems Group|
Tracking With Sleepy Sensors
University of Illinios at Urbana-Champaign
We examine the problem of tracking objects (targets) in an energy efficient manner using a sensor network. For energy efficiency, the sensors in the network are allowed to enter a sleep state wherein they can save energy but not observe the targets. Our goal is to find sleeping policies that yield the best tradeoff between tracking error and energy efficiency. While optimal solutions to this problem are in general intractable, we show that we can find good sleeping policies that approach a lower bound on optimal performance.
(This is joint work with Jason Fuemmeler)
Problems with this site should be emailed to firstname.lastname@example.org