入侵检测系统中动态优化检测器生成方法的研究  

Study of the Generation of the Dynamic Optimization Detector in the Intrusion Detection Systems

作　　者：姜懿庭[1] 姜跃[2] 李静[1] 郑毅[1] 

机构地区：[1]云南师范大学计算机科学与信息技术学院,云南昆明650092 [2]云南财经大学高等职业技术学院,云南昆明650101

出　　处：《云南民族大学学报（自然科学版）》2010年第3期216-219,共4页Journal of Yunnan Minzu University:Natural Sciences Edition

基　　金：国家自然科学基金(60963020)

摘　　要：在基于人工免疫的入侵检测系统中,通常会采用阴性选择算法来生成检测器,这种算法生成的检测器存在缺乏更新、容易饱和、误警率高等缺陷.鉴于此,给出了一种生成检测器的算法思路:以免疫遗传算法作为基础,对子代记忆检测器进行优化,并结合动态自体集演化方法生成正常行为特征,使子代检测器能够在一定程度上继承父代检测器的优良基因,该方案发挥了遗传算法并行操作、全局寻优、自适应优化等特征,避免了免疫病理机制转移进入入侵检测系统造成的安全隐患,增强了免疫入侵检测系统的实时性、健壮性、高效性、并行性和可适应性.In the intrusion detection systems based on artificial immunity,the detectors are usually generated by the negative selective algorithm,but if the detectors are only generated by such algorithm,some functional defects will occur,such as lack of upgrade,limited capacity and high rate of wrong identity.For solving such problems,this research brings up a brand-new idea of algorithm for the generated detector on the basis of the artificial immune genetic algorithm to optimize the son memorial detectors,and uses the construction method of dynamic self-set which makes son detectors inherit the maximal excellent genes from father detectors.Thus,it makes the performance of the invasion detector achieve a high level of diversity and adaptability as well.The new algorithm performs well both in improving the good non-self detection rate and reducing the chance of wrong self detection.

关 键 词：自体集 人工免疫 遗传算法 记忆检测器 Hamming距离匹配 

分 类 号：TP182[自动化与计算机技术—控制理论与控制工程]