Intelligent Event Detection Mechanism For Wireless Sensor Networks

With the rapid development of micro-electricity system, computer technology and communications, the heterogeneous wireless sensor networks, as the most complicated, intelligent networks format of sensor networks, has been applied into all kinds of area including environment monitoring, health care, agriculture, and military. The major task of wireless sensor networks is to do environment surveillance and event detection. However, for heterogeneous wireless sensor networks, it is inadvisable to decide whether the events happen just according to single type node, since types of nodes in lower tier are different. Obviously, making decision of event detection based on comprehensive analysis and processing of different type of scalar data collected by same tier nodes becomes one of major significant research direction about wireless sensor networks.This thesis proposed a Scalar Data processing based Event Detection mechanism, which includes detailed discussing the related work about the event detection in WSN and establishing relatively complete mathematic model like data analysis model and event detection algorithm model and many details related to the proposed mechanism. This thesis also presented the simulation of the proposed mechanism, which is a comparison between proposed mechanism and traditional mechanism under the specific scenario. The results proved the reliability and efficiency of the proposed mechanism. The proposed mechanism efficiently enhances the performance including shortening the event detection period and decreasing the energy consumption. Finally, this Thesis also proposed a complete TinyOS based implementation with using Mica2 Mote and MTS310 Sensor Broad. Its systematic running is stable and available.