基于统计假设检验的虚拟传感网容错策略-JournalofNortheastern.PDFVIP

第33 卷第3期 东 北 大 学 学 报 ( 自 然 科 学 版 ) Vol133, No. 3 2 0 12 年 3 月 Journal of Northeastern U niversit ( Natural Science) Mar. 2 0 1 2 宋 欣, 王翠荣 ( , 110819) : , , , , # , , , , , # # : ; ; ; ; : TP 393 : A : 1005-3026( 2012) 03-036 1-04 Fault-Tolerant Strategy for VSN Based on Statistical Hypothesis Testing S ON G X i n, WA N G Cu i-rong ( School of Information Science Engineering, Northeastern Universit , Shen ang 110819, China. Corresponding author : SONG Xin, E-mail: bravesong @ 163. com) Abstract : In order to minimize energ dissipation and obtain higher fault sensor nodes recognition rate for deplo ing mult-i target monitoring wireless sensor networks ( WSN ) , a fault-tolerant strateg for virtual sensor netw orks ( VSN) based on statistical h pothesis testing w as proposed. On the cluster tree based VSN architecture, the fault occurring was judged b checking the matching degree betw een local data reading sequence and event statistical characteristics. The simulation results show ed that before an node energ exhaustion, the mult-i event monitoring produced less unicast messages and more multicast messages than single-event monitoring. In addition, the proposed strateg can maintain the fault identification rate and event region monitoring probabilit at the satisfactor level w ith increase of the sensor nodes fault probabilit . Key words: w ireless sensor netw orks ( WSN ) ; statistical h pothesis testing; fault-tolerant strateg ; virtual sensor netw orks( VSN) ; event monitoring ( virtual sensor networks,