Tracking dynamics using sensor networks: some recurring themes

Abstract

Much of the data consumed today is dynamic, typically gathered from distributed sources including sensors, and used in real-time monitoring and decision making applications. Large scale sensor networks are being deployed for applications such as detecting leakage of hazardous material, tracking forest fires or environmental monitoring. Many of these “natural” phenomena require estimation of their future states, based on the observed dynamics. Strategically deployed sensors can operate unattended (minimizing risk to human life) and provide the ability to continuously monitor the phenomena and help respond to the changes in a timely manner. In this paper, we show that in-network aggregation, in-network prediction, and asynchronous information dissemination form sound building blocks for addressing the challenges in developing low overhead solutions to monitor changes without requiring prior knowledge about the (dynamics of) the phenomena being monitored.