基于生物记忆原理的入侵预警频度异常分析方法  

An Approach to Frequency Anomaly Analysis of Intrusion Early Warning Based on Biological Memory Theory

作　　者：黄光球[1] 刘冠[1] 石昌文[1] 

机构地区：[1]西安建筑科技大学管理学院

出　　处：《微计算机信息》2007年第04X期67-69,共3页Control & Automation

基　　金：陕西自然科学基金项目(2005F38);校基础研究基金项目(JC0616)

摘　　要：通过分析生物记忆的基本原理,建立了基于生物记忆原理的入侵预警频度异常分析模型。在该模型中,应用短时记忆容量限制理论和遗忘理论来确保信息更合理地新旧更替,从而可节省入侵检测系统大量的存储空间;应用感应阈技术灵活地调整频度异常灵敏度,使入侵分析更有敏感性和针对性,使入侵检测系统摆脱大量无用的分析消耗,而把注意力集中到更有可能是攻击行为的数据中去。实践检验发现该分析方法能有效的对异常访问行为产生警报,变化形式灵活,对于短时间的异常访问行为可给出比较准确的警报。Through studying on biological memory theory, a model of frequency anomaly analysis of intrusion early warning based on biological memory theory is represented. In the model, the short-period memory capacity restricting theory is used to create optimum policies of data saving and updating in accessing process so that a large amount of system＇s storage space is saved; the inductive valve controlling technique is used to flexibly adjust sensitivity of frequency anomaly analysis so as to make intrusion analysis to be more sensitivity and pertinence. Therefore network intrusion detection system can cast off a large amount of useless analysis so that more attention of system is paid to some data with great attacking possibility. An application shows that the method can effectively throw warning signals on anomalous access, and has advantages of flexible adjustment and high accurate rate of warning on short-period anomalous access actions.

关 键 词：网络安全 记忆原理 遗忘原理 入侵检测 预警 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]