Passive Diagnosis For WSNs Using Sensingdata

Node resource resource-constrained network dynamics and unattended, the forest ecologicalmonitoring sensor network easily during operation from various problems, and how to ensure thenormal work of the nodes in the forest wilderness environment, collect data for real and effective, andimplement of accurate, real-time, low-power fault diagnosis is a key issue in sensor network operationand management of forest ecosystem. The common diagnosis judgment and assessment strategies ofnetwork no longer suitable for the large scale wireless sensor network，which is shortage of highnetwork connectivity and network traiffc requirements, low failure identification rate, as well as badlyflexibility and scalability.In response to these problems, this paper design and implement a fault diagnosis method basedon Sensing data called DSD (Diagnosis based on Sensing Data), a passive diagnosis method usingsensing date in wireless sensor network. GreenOrbs is a Long-term deployment WSN system in theforest, which released a large number of actual monitoring data on the internet(http://www.greenorbs.org). By analyzing the temporal and spatial characteristics of forestenvironment sensing data, such as data sequence, supply voltage to study the correlation betweenfailures and characterizations. Determine the only corresponding relationship between the failure andperception data characteristics due to the rough set theory，establish knowledge library of failure tohelp diagnosis. And during operation, the knowledge library can continue to add the system todetermine the new relationship，automatic expansion to cope with the changing environment of foresteco-sensor network system.In order to further the development of DSD function test the efifciency of diagnosis, as well as tomeet the demand for sensor networks for environmental monitoring of forest ecosystems, design andimplement a wireless sensor network visualization test system-TestbedVisualizer. WithTestbedVisualizer, users can monitor WSN and the surrounding environment changes dynamically. Itsupports both olffine playback mode and real-time monitoring mode. It also supports remotemanagement on-line nodes and reprogramming individual or batch nodes，custom format data import,et al.On the basis of these studies DSD diagnostic efficiency test. The experiment carried out inTestbed and realistic environment show that DSD can improve diagnosis eiffciency，decrease the burden of network traffic, and have the flexibility with the arrangement of large-scale wireless sensornetwork.