EE360 Winter 2010 Project: Power Efficiency with Correlated Data Compression in Sensor Networks

Abstract: Wireless sensor networks with long lifetime requirements have severe power constraints. This paper attempts to address these power constraints by taking advantage of data correlation between nodes to compress data and conserve energy in data transmission to the base station. A translationally and rotationally invariant correlation model is presented, and data compression ratios are calculated from the relationship between correlation and mutual information. Intuition about the energy trade-offs in data compression is drawn from simulations. Possible extensions of this paper, such as the use of different correlation models, are also discussed.

Project Proposal

Revised Project Proposal 2/1/2010

Progress Report 2/19/2010

Final Report 3/14/2010