基于人工免疫的无线AD HOC网络MAC层自私攻击检测  

Detection of Selfish Attack on MANET MAC Layer Based on Artificial Immune

作　　者：杜秀娟[1,2] 金志刚[3] 

机构地区：[1]天津大学计算机科学与技术学院,天津300072 [2]青海师范大学计算机学院,西宁810008 [3]天津大学电子信息工程学院,天津300072

出　　处：《小型微型计算机系统》2011年第7期1360-1364,共5页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(90604013)资助;天津市应用基础及前沿技术研究计划(08JCYBJC14200)资助;青海省科技厅项目(2010-J-728)资助

摘　　要：通过对IEEE 802.11 MAC层攻击的分析与分类,提出将节点传帧率和退避时间序列作为行为特征,建立了MAC层攻击检测的人工免疫模型,给出饱和状态下的节点检测算法-基于滑动窗口残差阈值法,并对自私节点的退避时间序列进行编码和基因处理,给出非饱和状态下基于相对位置的子串相关函数匹配的基因检测算法,解决了传统检测算法不能有效检测智能攻击和与现有协议不兼容的局限性.The paper uses frame rate together with back-off time as nodes＇ behavior characteristic by analyzing and classifying the MAC layer attack of IEEE 802.11 protocols, sets up artificial immune modeling of MAC layer attack detection. The nodes in saturation status employ residual error threshold algorithm based on slide window to detect selfish attack node, further more encodes the selfish nodes＇ back-off time sequence and transfer it into gene detector which is employed by the node in non-saturation status to detect attack behavior based on relative position sub-string correlation function matching algorithm. So the paper resolves the localization of traditional detection algorithm which can＇t detect availably smart attack or incompatible with current 802.11 protocols.

关 键 词：人工免疫 退避时间 特征序列 自私攻击 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]