Report ID
2006-08
Report Authors
Selim Gurun and Chandra Krintz
Report Date
Abstract

We present a new energy estimation model for sensor network intermediate nodes (i.e. the Crossbow XScale Stargate). Such devices are battery powered and resource constrained and commonly employed as communication, processing, and gateway elements within sensor networks. Understanding and accurately characterizing the energy behavior of such devices is key to conserving the battery life of sensor systems.

In this paper, we present a macro-model for estimating the energy consumption of the device as a whole that couples estimation techniques for computation and communication. We construct our model using empirical data that we collect via hardware performance monitors. Our contributions are two-fold. 1) We demonstrate that hardware performance monitors are effective in modeling the computational energy consumption of sensor network gateways, 2) we present an analysis of linear dependencies between various hardware performance events, discuss their effects on model stability, and show how we can use principal component analysis to remove the undesirable side effects of such dependencies.

Document
2006-08.pdf173.98 KB