Queries for historic events in geosensor networks

Abstract

Mining for historic data is relevant to study the behaviour of complex systems. This article addresses the problem of querying a geosensor network (GSN) for historical data. GSNs - sensor networks of mobile, location-aware nodes-are typically deployed for real-time observations and local analysis. This article's interest is in reconstructing historic states or events. The problem is relevant in studying the emergence of patterns in complex systems, where previous states or events have to be recovered to explain a current situation. This article compares conceptually different network architectures and querying strategies with respect to their performance in reconstructing events or processes..