高效的攻击检测与数据融合算法  

Efficient attack detection and data aggregation algorithm

作　　者：程宏兵[1,2,3] 容淳铭 黄晓 EGGEN Skjalg 曾庆凯[1] 

机构地区：[1]南京大学软件新技术国家重点实验室,江苏南京210093 [2]江苏城市职业学院信息工程系,江苏南京210017 [3]斯塔万格大学计算机科学与电子工程系 [4]江苏移动通信有限公司互联网中心,江苏南京210013

出　　处：《通信学报》2012年第9期85-94,共10页Journal on Communications

基　　金：国家自然科学基金资助项目(61170070;61021062;60873231);江苏省六大人才高峰计划基金资助项目(11-JY-009);江苏省高校自然科学基金资助项目(11KJB510003);中国博士后基金资助项目(2012M511252);江苏省博士后基金资助项目(1102014C)~~

摘　　要：给出了一种高效的无线多媒体传感器网络攻击检测和数据融合算法EIDSART。该算法从节点的多元属性方面对节点行为特征进行界定,通过选择合适的邻居节点集合,可以运用于任意规模的多媒体传感器网络;另外,在经过精确检测攻击行为的情况下,对传感数据进行了融合,降低了网络通信开销。仿真结果表明,EIDSART在攻击检测精度和误报率等方面具有优势,并能得到精确的数据融合结果。An efficient algorithm of attack detection and data aggregation for wireless multimedia sensor networks based on the previous work was proposed.The proposed algorithm concludes the action trait of sensor nodes from their attrib-ute vectors without any prior knowledge,at the same time;it was scalable and could be applied in large scale networks.The simulation results show that the proposed algorithm can detect the attacks action more accurate than other technolo-gies,and can make data aggregation efficiently.At the same time,the proposed algorithm can make the wireless multi-media sensor networks secure and reduce communication flow so that it will save a lot of resources in wireless multime-dia sensor networks.

关 键 词：物联网 无线多媒体传感器网络 攻击检测 数据融合 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]