南京大学学报(自然科学版) ›› 2016, Vol. 52 ›› Issue (2): 261.
胡 石1,2,李光辉1,2,3*,冯海林1,2
Hu Shi1,2,Li Guanghui1,2,3*,Feng Hailin1,2
摘要: 异常数据检测在基于无线传感器网络的环境监测系统中起着十分重要的作用,不仅有助于对传感器网络健康状况的监测,而且能够及时发现外部环境发生的突发事件(如森林火灾、环境污染等).通过对topk算法的改进,提出了一种基于topk(σ)的无线传感器网络异常数据检测算法.不同于topk算法,该算法根据传感器节点采集到的数据分布规律,构造合适的数据网格,将多维数据归一化处理后置入相应的网络单元.然后通过增设距离阈值σ来重构PC列表(populatedcells list).除了对每个单元格及其邻域内的数据点个数分别进行排序,还计算不同数据子集之间的欧氏距离,并与阈值σ的比较,确认数据子集与正常值集合的偏离程度,从而提高检测结果的准确性.通过MATLAB仿真实验发现,距离阈值σ的选取对算法效果具有较大的影响,当σ∈[2.5,3]时,topk(σ)算法在维持较高检测率的同时,最大程度地降低误报率.当取σ=3时,对于给定的5个数据集,topk(σ)算法的检测率平均达到了93.70%,比topk算法平均提高了4.94%,误报率则比topk算法平均降低了4.48%.
[1] 周贤伟,王 培,覃伯平等.一种无线传感器网络异常检测技术研究.传感技术学报,2007,20(8):1870-1874.(Zhou X W,Wang P,Qin B P,et al.An anomaly detection technology in wireless sensor networks.Chinese Journal of Sensors and Actuators,2007,20(8):1870-1874.) [2] Subramaniam S,Palpanas T,Papadopoulos D,et al.Online outlier detection in sensor data using nonparametric models.In:Proceedings of the 32nd International Conference on Very Large Data Bases.Almaden,Spanish:VLDB Endowment,2006,187-198. [3] Perrig A,Stankovic J,Wagner D.Security in wireless sensor networks.Communications of the ACM,2004,47(6):53-57. [4] 曹冬磊,曹建农,金蓓弘.一种无线传感器网络中事件区域检测的容错算法.计算机学报,2007,30(10):1770-1776.(Cao D L,Cao J N,Jin P H.A faulttolerant algorithm for event region detection in wireless sensor networks.Chinese Journal of Computers,2007,30(10):1770-1776.) [5] 曹 磊,韩 涛,张 婧等.无线多媒体传感器网络拥塞感知的流控制机制.南京大学学报(自然科学),2014,50(2):173-180.(Cao L,Han T,Zhang J,et al.Congestionware flow control strategy in wireless multimedia sensor networks.Journal of Nanjing University(Natural Sciences),2014,50(2):173-180.) [6] Martincic F,Schwiebert L.Distributed event detection in sensor networks.In:International Conference on Systems and Networks Communication.IEEE,2006,43. [7] 毕 冉,李建中,程思瑶.无线传感器网络(ε,δ)—近似Topk查询处理算法.通信学报,2011,32(8):45-54.(Bi R,Li J Z,Cheng S Y.(ε,δ)-approximate Topk query processing algorithm in wireless sensor networks.Journal of Communications,2011,32(8):45-54.) [8] 姜旭宝,李光耀,连 朔.基于变宽直方图的无线传感器网络异常数据检测算法.计算机应用,2011,31(3):694-697.(Jiang X B,Li G Y,Lian S.Outlier detection algorithm based on variablewidth histogram for wireless sensor network.Journal of Computer Applications,2011,31(3):694-697.) [9] 肖政宏,陈志刚,李庆华.WSN中基于分布式机器学习的异常检测仿真研究.系统仿真学报,2011,23(1):181-187.(Xiao Z H,Chen Z G,Li Q H.Simulation study of anomaly detection based on distributed machine learning for WSN.Journal of System Simulation,2011,23(1):181-187.) [10] 刘 群,顾 金,张足生.基于无线传感器网络的车辆分型算法.南京大学学报(自然科学),2013,49(5):655-663.(Liu Q,Gu J,Zhang Z S.Vehicle lassification algorithm based on wireless sensor networks.Journal of Nanjing University(Natural Sciences),2013,49(5):655-663.) [11] Aggarwal C C,Philip S Y.Outlier detection with uncertain data.Society for Industrial and Applied Mathematics,2008,483-493. [12] Sheng B,Li Q,Mao W,et al.Outlier detection in sensor networks.In:Proceedings of the 8th ACM International Symposium on Mobile ad hoc Networking and Computing.York,USA:ACM,2007,219-228. [13] Shaikh S A,Kitagawa H.Distancebased outlier detection on uncertain data of Gaussian distribution.Web Technologies and Applications.Berlin,Germany Springer Berlin Heidelberg,2012,109-121. [14] Shaikh S A,Kitagawa H.Fast topk distancebased outlier detection on uncertain data.WebAge Information Management.Berlin,Germany Springer Berlin Heidelberg,2013,301-313. [15] Knorr E M,Ng R T,Tucakov V.Distancebased outliers:algorithms and applications.The VLDB Journal—The International Journal on Very Large Data Bases,2000,8(3-4):237-253. [16] Shaikh S A,Kitagawa H.Efficient distancebased outlier detection on uncertain datasets of Gaussian distribution.World Wide Web,2014,17(4):511-538. [17] Shaikh S A,Kitagawa H.Topk outlier detection from uncertain data.International Journal of Automation and Computing,2014,11(2):128-142. [18] Hill D J,Minsker B S.Anomaly detection in streaming environmental sensor data:A datadriven modeling approach.Environmental Modelling & Software,2010,25(9):1014-1022. |
No related articles found! |
|