基于直觉模糊核匹配追踪的网络入侵意图识别方法研究

项目名称： 基于直觉模糊核匹配追踪的网络入侵意图识别方法研究

项目编号： No.61309022

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 自动化技术、计算机技术

项目作者： 雷阳

作者单位： 中国人民武装警察部队工程大学

项目金额： 27万元

中文摘要： 随着网络攻击事件的增多和攻击行为的复杂化，入侵检测系统暴露出诸如误报漏报严重、海量数据处理能力弱等缺陷。网络入侵意图识别通过处理不确定性告警信息，探寻隐藏在多个单独攻击事件背后的逻辑关系，有效解决上述问题。针对这一典型的不确定性难题，本项目拟利用直觉模糊集、核匹配追踪及免疫克隆选择理论，研究基于特征加权直觉模糊c均值聚类的网络入侵意图识别方法，研究直觉模糊集与核匹配追踪理论的融合方法及其在网络入侵意图识别中的应用，进而研究其优化方法--直觉模糊c均值聚类核匹配追踪方法及其在入侵意图识别中的应用，研究核匹配追踪与免疫克隆理论的融合方法及其在网络入侵意图识别中的应用，进一步研究其改进方法--自适应多克隆聚类的核匹配追踪方法及其在入侵意图识别中的应用。本项研究通过汲取上述理论优势，突破其相互融合的难点，形成有效的网络入侵意图识别方法，可望提高网络入侵意图识别的能力，推进我国网络防御系统的发展。

中文关键词： 网络安全；直觉模糊集；核匹配追踪；免疫克隆选择；入侵意图识别

英文摘要： As to the increasing network attacks and more complicated intrusion actions, intrusion detection systems expose lots of defects, such as mass duplications, huge amount alerts. Intrusion plan recognition could solve these problems effectively by fusing lots of uncertain alert information to recognize logic relations concealed in single attack incident. To the above typical problem with uncertainty, the project intends to adopt intuitionistic fuzzy sets, kernel matching pursuit and immune clonal selection algorithm. Technique for intrusion plan recognition based on feature weighted intuitionistic fuzzy c-means clustering will be firstly investigated. Secondly, technique for intrusion plan recognition based on intuitionistic fuzzy kernel matching pursuit will be deliberated as the key part of this project, with well-established theoretical model. In further research, the improvement of second technique, namely technique for intrusion plan recognition based on intuitionistic fuzzy c-means clustering and kernel matching pursuit. Finally, the algorithm for kernel matching pursuit syncretized with immune clonal selection will be deeply researched, and how to apply to intrusion plan recognition will be respectively investigated. Furthermore, the improvement of the former one, namely technique for intrusion plan recognit

英文关键词： Network security；Intuitionstic fuzzy set；Kernel matching pursuit；Immune clonal selection；Intrusion Intension Recognition