基于K-means聚类的WSN异常数据检测算法  

Abnormal Data Detection Algorithm for WSN Based on K-means Clustering

作　　者：费欢 李光辉[1,2] 

机构地区：[1]浙江农林大学信息工程学院,浙江临安311300 [2]浙江农林大学浙江省林业智能检测与信息技术研究重点实验室,浙江临安311300

出　　处：《计算机工程》2015年第7期124-128,共5页Computer Engineering

基　　金：国家自然科学基金资助项目(61174023);浙江省自然科学基金资助项目(Y1110791)

摘　　要：为提高无线传感器网络应用系统的可靠性,对传感器节点采集的环境数据集进行检测,提出一种改进的异常数据检测算法。采用K-means算法思想,结合无线传感器网络数据的特点,以欧式距离作为指标,比较数据点的相似度并划分聚类,根据数据点与聚类中心之间的距离区分正常数据与异常数据。实验结果表明,当数据规模超过1 000时,与基于噪声的密度聚类算法相比,该算法对于异常数据的检测率较高,误报率较低。In order to improve the reliability of Wireless Sensor Network（WSN） application system,it detects abnormal data from sensor environmental data set. An algorithm of abnormal data detection based on clustering of data mining is proposed in the paper, which not only adopts K-means clustering but also takes the characteristics of WSN data into account. This algorithm uses Euclidean distance to compare similarity of data for cluster partitioning, and identifies the abnormal data according to the distance between data point and cluster center. Experimental results show that when data is more than 1000, compared with the algorithm based on Density-based Spatial Clustering of Applications with Noise （DBSCAN） ,the detection accuracy of this algorithm is higher and the false positive rate is lower under the same conditions.

关 键 词：K-MEANS算法 无线传感器网络 聚类 异常数据检测 密度聚类 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]