一种传感器网络假冒攻击源的测定方法  

Masquerader Detection and Identification Approach in Sensor Networks

作　　者：谢磊[1,2] 王惠斌[1,3] 祝跃飞[1] 徐勇军[2] 

机构地区：[1]信息工程大学信息工程学院,郑州450002 [2]中国科学院计算技术研究所,北京100190 [3]河南司法警官职业学院,郑州450002

出　　处：《计算机科学》2009年第6期68-71,124,共5页Computer Science

基　　金：国家863计划(2006AA01Z223,2006AA01Z225);国家自然科学基金(60772070)资助

摘　　要：传感器网络中的假冒攻击是一种主动攻击形式,它极大地威胁传感器节点间的协同工作。提出了基于邻居协同测定假冒攻击源算法(CNAMDI)。在CNAMDI算法中,节点根据主动报警规则和从动报警规则发现假冒行为,基于义务测定集传递规则的邻居协同实现对假冒攻击源的测定。CNAMDI算法无需全网拓扑信息及路由协议支撑,测定过程不借助密码算法。通过分析得出,当局部网络密度较高时,CNAMDI算法具有漏报率低、成功测定率高的特点。仿真分析表明,对比朴素算法,CNAMDI算法使漏报率平均降低了25.8%,成功测定率平均提高了45.5%,平均发包数仅增加了1.19个。Masquerade attack is one of active attack forms in sensor networks,which will be a serious threat to the synergies among sensor nodes. A collaborative neighbor based algorithm for masquerader detection and identification （CNAMDI） was proposed. In CNAMDI algorithm, nodes can use initiative or slave alarm rules to detect a masquerade attack and can identify the masquerader based on collaborative neighbor with volunteer rule of suspect set forwarding. CNAMDI algorithm can work without the need of any underlying routing protocols and global topology information,and the identification process does not rely on any cryptographic algorithms. Theoretical analysis shows that, when higher density in local area networks, this algorithm has a property of low leak rate and high success identification rate. Simulation analysis shows that, compared with simplicity algorithm, CNAMDI algorithm reduces averagely 25. 8% leak rate and improves averagely 45. 5% success identification rate with only introducing additional 1. 19 sending packets per neighbor.

关 键 词：传感器网络 邻居协同 假冒攻击 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]