在线自适应网络异常检测系统模型与算法.pdf

计算机研究与发展 ISSN 1000—1239[CN11—1777／TP Journal Researchand ofComputer Development 在线自适应网络异常检测系统模型与算法 魏小涛1 黄厚宽2 田盛丰2 1(北京交通大学软件学院北京100044) 2(北京交通大学计算机与信息技术学院北京100044) (weixt@bjtu．edu．cn) AnOnline Network Detection and Adaptive Anomaly System。ModelAlgorithm Wei Tian Houkuan2，and Xiaota01，Huang Shengfen92 1(School 100044) ofSoftware。BeijingJiaotongUniversity，Beijing 2(School and Information 100044) ofComputer Technology，BeijingJiaotongUniversity，BeQing AbstractTheextensiveofInternetand networksmakes a criticalissue． usage computer security Thereisan needfornetworkintrusiondetection whichcan defendnetworks urgent systems actively the threats．Inthis online network againstgrowingsecurity paper，alightweightedadaptive anomaly is functionbased detectionmodel relatedinfluence detection system presented．The anomaly algorithm isalso can networktrafficdatastreamin build provided．Thesystemprocess real—time，graduallyup itslocalnormal baseandintrusionbaseunderalittle oftheadministrator， pattern pattern supervising and thecontentsofthe base tothe ofthenetwork dynamicallyupdate knowledgeaccordingchanging can the mode，the detectnot thelearnedintrusion applicationpatterns．Atchecking system only butalsotheunseenintr