摘要
提出了一种网络流量异常状态统计模型——混合二次网络状态模型MQNSM-G(DKS,DKKS,DAKS)。该模型从动态性原则以及降低误检率和漏检率思想出发,改进原有统计模型,建立了可以动态设定描述网络流量状态参数的加权统计模型。基于混合二次网络状态模型MQNSM-G(DKS,DKKS,DAKS)的入侵检测系统进一步证明了该模型可以更大程度上提高异常检测性能,降低其误检率和漏检率。
A statistical raodel for detecting abnormal network traffic - mixed quadratic network states model MQNSM- G( DKS, DKKS, DAKS )is presented. Based on principles of developments and reducing FNP and FPP, this paper builds up a statistical model with wrights that can dynamically .set parameters of network traffic states, which improves on former statistical models. It has proved that performances of anomaly detection can be improved to a great degree and the FPP and FNP can be cut down prominently in an IDS based on the mixed quadratic network states model MQNSM - G( DKS, DKKS, DAKS).
出处
《计算机技术与发展》
2007年第3期153-155,共3页
Computer Technology and Development
基金
国家自然科学基金(60573141)
华为基金资助
关键词
分布式拒绝服务攻击
入侵检测
误检率
漏检率
dlstributed denial of service
intrusion detection
false positive probability
false negative probability