A Two Round Reporting Approach to Energy Efficient Interpolation of Sensor Fields

Abstract

In-network aggregation has been proposed as one of the main mechanisms for reducing messaging cost (and thus energy) in prior sensor network database research. However, aggregated values of a sensor field are of limited use in natural science domains because many phenomena, e.g., temperature and soil moisture, are actually continuous and thus best represented as a continuous surface over the sensor fields. Energy efficient collection of readings from all sensors became a focus in recent research literature. In this paper, we address the problem of interpolating maps from sensor fields.

We propose a spatial autocorrelation aware, energy efficient, and error bounded framework for interpolating maps from sensor fields. Our work is inspired by spatial autocorrelation based interpolation models commonly used in natural science domains, e.g., kriging, and brings together several innovations. We propose a two round reporting framework that utilizes spatial interpolation models to reduce communication costs and enforce error control. The framework employs a simple and low overhead in-network coordination among sensors for selecting reporting sensors so that the coordination overhead does not eclipse the communication savings. We conducted extensive experiments using data from a real-world sensor network deployment and a large Asian temperature dataset to show that the proposed framework significantly reduces messaging costs.