Time Series Snapshot of Event Boundary Detection and Area Estimation in Wireless Sensor Networks

Abstract

To monitor the events like forest fire, flood, earth tremor etc., event boundary detection and area estimation are two challenging tasks for Wireless Sensor Networks (WSN). Researchers, so far, have proposed various innovative solutions assuming the event boundary to be static. To monitor spatiotemporal events by WSN, in this paper, we study some existing techniques to report snapshots of event boundary in successive intervals, and compare their performance in respect of accuracy, latency and message communication overhead. Testbed experiments have been done to evaluate the performance and the results show that the computation intensive techniques perform better, provided they satisfy the application specific real time constraint.