一种模糊神经网络检测下的入侵容忍系统设计

Design of Intrusion Tolerant System Based on Flurry Neural Network

为了与网络系统中的入侵行为进行抗衡,可以通过安装入侵检测系统来监控系统中的各种网络行为,入侵容忍可以作为入侵检测系统的互补系统与之一起工作。漏报与误报问题是入侵检测系统面临的根本性问题,加入神经网络技术后并不能完全解决此问题。入侵容忍技术的提出很好地平衡了"失误"与"恶意"、"入侵"与"非有意"之间的冲突与矛盾。入侵容忍本身就是一种"模糊"概念,为模糊数学在计算机安全中的应用提供了有力的理论支持。提出了一种基于模糊神经网络的入侵检测系统应用于入侵容忍的设计方案,并将入侵容忍与入侵检测、防火墙结合,作为一个独立于应用的系统进行设计。

In order to confront the invading behavior of the net system,the invading detection system could be set up to inspect and control the various net perform. The invading abide could work with the invading detection system as a compensatory system. Fail to report and misreport a problem is ultimacy problem facing the intrusion detect system, can not resolve this problem completely after adding the nerve network technology. Intrusion tolerant technology suggesting that has balanced conflict and contradiction between the ＂fault＂ and ＂evil intention＂,＂ has invaded＂ and ＂must have a mind to＂. Intrusion tolerant technology is one kind of ＂flurry＂ concept,be that the fuzzy mathematics has provided forceful theory support in application in computer safety. This article brought forward a design project ,which is one kind of the invading detecting application based on flurry neural network,and combined the invading abide and the invading detection with the firewall ,which designed as a system isolated from the application.