【作者】 任大勇；

【导师】 黄光球；

【作者基本信息】 西安建筑科技大学 ， 管理科学与工程， 2008， 硕士

【摘要】 随着网络攻击技术和安全防御技术的不断发展,攻击和防御已经成为网络安全的两个密切相关的侧面,不深入研究攻击理论和技术就不能有效地保护网络信息系统的安全。网络攻击研究的一个关键问题是对攻击的认识和描述。攻击模型能对整个攻击过程进行结构化和形象化的描述,有助于分析和充分利用已知的攻击行为研究成果,进一步提高攻击检测和安全预警的效率。本文以双枝模糊集和模糊Petri网为理论基础,以既有攻击因素、又有防御因素的网络攻击形式为研究对象,将网络攻击中对攻击起促进与抑制作用的两方面进行综合考虑与分析,提出了一系列对网络攻击进行定性定量描述的方法。首先,以双枝模糊决策和模糊Petri网理论为基础,定义了一种全新的网络攻击模型BBFPAN(Both-Branch Fuzzy Petri-net Attack net)。并在对变迁输入库所集中因素进行双枝模糊决策分析的基础上,结合一种不确定推理方法,给出了相应的BBFPAN推理算法。其次,在深入研究分析了集对分析理论和双枝模糊集的关系之后,发现了以双枝模糊集为基础的网络攻击模型BBFPAN中多层次的集对关系,从而首次将集对分析理论应用于攻击模型的分析研究,提出了BBFPAN双枝集对分析模型。随后,本文以双枝模糊逻辑理论为依据,提出了BBFPAN模型推理的基本规则,并结合BBFPAN模型的定义,给出其相应基于双枝模糊逻辑的推理算法,进一步完善了攻击模型BBFPAN的推理方法。最后,针对以模糊Petri网为理论基础的新型网络攻击模型BBFPAN,自学习能力差的缺点。本文就此做了初步的探索研究,提出了一种适用于BBFPAN进行层次式划分的分层算法和相应的推理算法。本文将对攻击实施起正反两方面作用的因素一起考虑,提出了一系列对网络攻击进行定性定量描述的方法,并就神经网络在攻击模型BBFPAN中的应用做了初步探索研究,有利于日后的进一步研究。更多还原

【Abstract】 As the developing of the attack and defend technology,attack and defend also become two sides that closely related of network security.Further investigate, attack theory and technology can protect network security of information system effectively.It is a key problem studied about that understanding and description to the network attacks.Attack models can help in describing and analyzing the course of attack structurally and pictorially,facilitate analyzing and utilizing the known attack,further more that can help in improving the efficiency of attacking measuring and security prediction.Based on both-branch fuzzy sets and fuzzy Petri net(FPN),the network attack with the promoting and suppressive factors will be researched in this article and the promoting and suppressive factors are analyzed together.A series of qualitative and quantitative approach to network attack are put forward.At the beginning,based on both-branch fuzzy decision-making and FPN,a new network attack model named both-branch fuzzy Petri-net attack net(BBFPAN)is put forward.Through the both-branch fuzzy decision-making analysis to the factors of input place set,a reasoning algorithm is proposed which applies incertitude consequence way.For another,delved into the both-branch fuzzy sets and set pair analysis,more levels set pair within the BBFPAN based on the both-branch fuzzy sets can be discovered.A both-branch set pair analysis about BBFPAN is put forward which originally applies set pair analysis into the attack modeling.After this problem,based on the both-branch fuzzy logic,the basic inference rules about BBFPAN are put forward.A reasoning algorithm is proposed which use the basic inference rules and it can enrich the reasoning method of the attack model BBFPAN.At last,the lack of learning mechanism is the weakness of the fuzzy systems and is also the weakness to the new network attack model BBFPAN based on the fuzzy Petri net.We have done some preliminary exploratory research to the lack.A delaminating algorithm which can partition the BBFPAN into several levels and a fuzzy reasoning algorithm are presented.This paper considers together the promoting and suppressive factors that have effect to the attacks.Kinds of qualitative and quantitative method about network attack are proposed.At the same time we have done some preliminary exploratory study about the BBFPAN using the neural network and have forecast the future further research direction.更多还原