Evaluating Indexing And Routing Schemes For Spatial Queries In Sensor Networks

Abstract

Recent advances in low-power sensing devices coupled with widespread availability of wireless ad-hoc networks, has fueled the development of sensor networks. These senor networks have various applications such as to monitor conditions at different locations (temperature, pressure, rainfall, vibrations etc.,), tracking of objects and so on. Each device is equipped with an energy source (usually with a battery), memory, CPU and communication bandwidth, which is severely constrained. Hence each sensor network is comprised of hardware for sensing, software for communication and computational algorithms. Spatial queries are commonly applied to sensor network, for example: "Find the highest temperature sensed in a particular region". Spatial query processing is considered extensively in the context of centralized databases. In this thesis, we examine spatial queries in distributed sensor networks. We use a binary tree, which is constructed by a MULT routing protocol and the meta-data that has been collected in the base station to construct an index structure, thus resolving the spatial queries, we evaluate and compare the behavior of our approach with other kind of indexing structure and show that our mechanism is efficient in terms of increased in-network computation and lowered communication.