邻域空间下检测器的生成及分布算法研究

发布时间：2018-04-20 09:49

本文选题：入侵检测 + 检测器　；参考：《哈尔滨理工大学》2014年硕士论文

【摘要】：随着网络技术的飞速发展，越来越多的人开始关注网络安全，免疫入侵检测作为一种非常有效的入侵检测技术也吸引了越来越多的研究人员。免疫入侵检测将人工免疫原理引入到入侵检测，使系统能够实时的检测异常行为作出判断并发出警报。本文介绍分析了入侵检测的分类以及相关技术，阐述了入侵检测的国内外的研究现状，在对比了主流的入侵检测算法后，选择邻域形态空间作为入侵检测研究的形态空间。在邻域否定选择算法中重点放在了增强的自适应性，提高检测效率上。并且针对领域空间下检测器分布问题进行分析研究，提出改进算法。对于以上提出的问题，通过分析邻域空间的特点，本文采用一种基于有效信息比率的离散化方法对邻域否定选择算法进行改进，增强其自适应性。针对邻域空间下的检测器分布的问题，在分析了邻域空间的特点和匹配规则后，采用两级检测器集合和变阈值的分布方法，能够用更少的检测器达到更好的分布效果。通过本文所述算法的研究，可以有效的改善邻域否定选择算法的自适应性，提高检测率。邻域空间下的检测器分布更有效率，减少资源的消耗。
[Abstract]:With the rapid development of network technology, more and more people begin to pay attention to network security. As a very effective intrusion detection technology, immune intrusion detection has attracted more and more researchers. Immune intrusion detection introduces artificial immune principle into intrusion detection, which enables the system to detect abnormal behavior and issue alarm in real time. This paper introduces and analyzes the classification of intrusion detection and related technologies, expounds the current research situation of intrusion detection at home and abroad, and after comparing the mainstream intrusion detection algorithms, selects the neighborhood morphological space as the morphological space of intrusion detection research. In the neighborhood negative selection algorithm, emphasis is placed on the enhancement of self-adaptability and the improvement of detection efficiency. An improved algorithm is proposed to solve the problem of detector distribution in domain space. For the above problems, by analyzing the characteristics of neighborhood space, this paper uses a discretization method based on effective information ratio to improve the neighborhood negative selection algorithm to enhance its self-adaptability. In order to solve the problem of detector distribution in neighborhood space, after analyzing the characteristics and matching rules of neighborhood space, a two-stage detector set and variable threshold distribution method are adopted to achieve better distribution effect with fewer detectors. Through the research of the algorithm mentioned in this paper, the self-adaptability of neighborhood negative selection algorithm can be improved effectively and the detection rate can be improved. The detector distribution in neighborhood space is more efficient and resource consumption is reduced.
【学位授予单位】：哈尔滨理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.08

【参考文献】