无线传感网络局部离群节点定位模型仿真  

Simulation on Localization Model of Part Outlier Detection in Wireless Sensor Network

作　　者：王岩[1] 李洪亮[1] 

机构地区：[1]长春工业大学人文信息学院,吉林长春130122

出　　处：《计算机仿真》2013年第12期253-256,共4页Computer Simulation

摘　　要：关于网络节点优化定位问题,在无线传感器网络中的离群节点是指那些明显偏离传感数据正常模式的节点。有效的离群节点定位可以保证网络安全。离群数据往往在属性上大幅偏离原始特征,节点在空间上不具备直观关联性。传统定位方法仅仅依靠空间位置的关联性检测节点,一旦离群节点数据属性偏离原始特征,造成空间失连,很难被准确的定位。提出一种不规则节点形状模型的无线传感器网络的局部离群点定位算法,采用一种带有约束条件的不规则模型描述传感网络中的节点分布情况,节点间通过参数运算区域相异度,再通过邻域分割方法除了网络节点分布的分散性,避免了传统算法由于离群数据属性偏离原始特征,产生的空间失连定位不准的问题。实验结果表明,改进算法具有定位精度高以及耗能量低的优势,是一种高效的传感网络离群节点定位方法。In this paper, the optimization problem of network node localization was researched. In the wireless sensor network, outliers are those nodes that stray significantly from normal mode of sensor data. Effective outlier localization can ensure network security. The feature of outliers often deviate significantly from the original attribute and the node in space does not have an intuitive relevance. Traditional positioning methods rely solely on the relevance of spatial position to detect node. Once the attribute of outliers deviate from the original feature, it may cause the loss of even space, and it is difficult to position precisely. In this paper, we proposed a new algorithm for local outliers posi- tioning in wireless sensor network based on irregular shape model. In this method, an irregular sensor model with constraint condition can describe the distribution of nodes in the network. The area dissimilarity was computed by parameter between nodes, and then the dispersion of the network nodes was excluded by the method of the neighborhood distributed to avoid the inaccurate positioning problem in traditional method. Experimental results show that the improved algorithm has the advantages of high positioning accuracy and low energy consumption. It is highly efficient outliers positioning method for sensor network node.

关 键 词：无线传感网络 离群点检测 不规则模型 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论]