Research Of Gradient Event Boundary Identification And Tracking Technology In WSNs

Wireless Sensor Networks (WSNs) is considered one of the21most influential andone of the10change the world technology in the21stcentury. It is widely used in military,healthcare, environmental monitoring, and daily life, because of it has some advantagelike small size, self-organization and convenient. The graded event tracking problem is oneof the important research topic in the wireless sensor networks. This research is primarilyused for area events tracking, just like, the spread fire, flood flows and toxic chemical leaketc. However, due to the limit enery and computing capabilities of wireless sensor networknode, when network run long time, in the tracking process, the accuracy of tracking thegraded event will be reduced and even worse lost which serious constraints the researchprocess of the graded event tracking in the wireless sensor network. So in this thesis, we arebased in reduce the energy consumption and improve the tracking accuracy, give thefollowing innovativework in the gradient event boundary identification and tracking on thebasis of the home and abroad research:Firstly, this paper presents an event boundary nodes selection algorithm based ongeometric calculations (NSBGC). First, to obtain the position coordinates of the sensor nodein accordance with the initialization of the wireless sensor network deployment, create theneighbor relations table by the communications between the wireless sensor node, andstatistics neighbor node information; sencond, find the node from the neighbor node tablewhich can create the triangle, caculate the triangle area which composition by the neughbornode and the node which wait to detect in heron’s formula and compare the raea size. Andalso judge the node which wait to detect whether is a boundary nodes. The simulationresults based in NS2show that, it can reduce the complexity of the algorithm when judgethe node, reduce the energy consume and improve the tracking accuracy.Secondly, this paper presents an gadient event boundary tracking method based on thenode calibration (EBTMNC). Firstly, sensor nodes state divided which combine with theNSBGC, let the active node to be sleep, reduce the node enery waste; Secondly, establishthe node sleep and wake-up mechanism based on the predicted, predict the position of thenext time, At last, compare the number of the NSBGC and the number of the predictmechanism, put the memssage feedback the master sensor nodes, adjustment the forecastparameters in real-time, make the prediction accuracy maintained at a high position. Thesimulation results based in NS2show that, this method reduce the energy consume from an overall, and improve the timeliness, effectiveness and accuracy during the tracking proess,get a greatly improved compared to the conventional method.