【作者】 袁硕；

【导师】 乔佩利；

【作者基本信息】 哈尔滨理工大学 ， 计算机应用技术， 2010， 硕士

【摘要】 信息的共享和网络的开放给我们带来了便利,同时也对信息安全提出了更高的要求。网络与信息安全已经成为信息系统的关键问题。本文研究生物免疫系统的学习与检测机制,入侵检测系统的技术和构成,基于传统的GA理论模型,拟利用均匀设计抽样的理论,对遗传算法中的交叉遗传算子进行了重新设计、对染色体的相似度及其相关内容进行了改进,并结合进局部搜索策略给出一种改进的混合遗传算法HGA;引入检测器冗余优化这一概念及其相关内容,使染色体的相似度与检测器冗余建立关联,针对检测器算法收敛速度慢并且容易陷入早熟的缺点,拟采用上述改进的HGA对成熟检测器进行进化,进而给出一种检测器优化新方案和一个基于HGA的免疫入侵检测模型。最后拟通过对“算法”和“模型”进行仿真实验,验证了算法和模型的可行性和有效性。首先,研究入侵检测和生物免疫的相关原理与技术。从入侵检测的概念、功能、作用、基本结构入手,并进一步研究入侵检测系统分析方法、发展趋势和发展方向。同时,分析人工免疫系统的工作原理和机制,并将生物免疫系统与入侵检测进行类比。研究基于免疫机制的入侵检测系统的模型。然后,研究基于改进混合遗传算法的免疫入侵检测模型。拟讨论对混合遗传算法的改进,对遗传算子进行讨论和设计,并将改进的混合遗传算法应用于对检测器的改进当中。进而给出改进的入侵检测系统模型的体系结构,并介绍各模块的工作流程。最后,通过对仿真实验及实验结果的分析。对前文给出的改进策略和改进结果进行检测和实验,得出结论。更多还原

【Abstract】 Information sharing and networks opening bring us the convenience, but also put higher information security requirements at the same time. Network and information security has become a key issue of information systems.This paper studies the learning and detection mechanisms of biological immune system, the technologies and composition of intrusion detection system, based on the traditional GA theoretical models, to be intended use the theory of uniform design sampling(UDS), has been redesigned the crossover genetic operator in genetic algorithm, improved the chromosome similarity and its related content, incorporated into the local search strategy, gives an improved hybrid genetic algorithm(HGA); pull in the conception of detector redundancy and its relevant content, so that, associate detector redundancy with chromosome similarity. Using the HGA to evolution the mature detectors, then gives a new program of detector optimization and an immuno-intrusion detection model based on improved HGA. The final simulation experiment of the "algorithm" and "model" which is mentioned above test and verify the feasibility and validity of the "algorithm" and "model". This program maintains the diversity and efficiency of detector, can be validate to have the better performance in searching velocity, global optimal search ability and improving detection rate and reducing false alarm rate to a certain degree.First, study the related principles and technologies of intrusion detection and biological immune. From the concept of intrusion detection, function, effect, the basic structure to start, and further to study the analysis methods of intrusion detection systems, trends of development. At the same time, analysis the principle and mechanism of artificial immune system working compare the biological immune system with intrusion detection. Research the intrusion detection system model based on immune mechanisms.Then, research the immune intrusion detection model based on improved hybrid genetic algorithm. Prepared to discuss the improvements of the hybrid genetic algorithm, discuss and design the genetic operators, and apply the improved hybrid genetic algorithm in the detector improvements. Thus gives an improved model of intrusion detection system architecture and describes the working flow of each module.Finally, through the analysis of simulation and experimental results, test the improvement strategy, to reach a conclusion.更多还原