基于人工免疫模型的入侵检测系统中检测器生成算法的分析与改进  

Research and Improvement of Detector Generation Algorithm in Intrusion Detection System Based on Artificial Immune Model

作　　者：胡亮[1] 王程明[1] 赵阔[1] 努尔布力[1] 姜千[1] 

机构地区：[1]吉林大学计算机科学与技术学院,长春130012

出　　处：《吉林大学学报（理学版）》2010年第1期67-72,共6页Journal of Jilin University:Science Edition

基　　金：国家自然科学基金(批准号:60873235);吉林省科技支撑计划重点项目基金(批准号:20080318);吉林大学"985工程"创新计划项目(批准号:20080244)

摘　　要：对检测器生成算法中的逐级反向选择算法进行改进,通过对检测器表示方式和检测器生成算法的修改,解决了原算法中检测器数量过多的问题,提高了检测器的生成效率和检测效率.算法的测试结果表明,改进后的算法在检测精度不变的前提下降低了描述相同非自体空间需要的检测器数量,提高了算法执行效率,使其更接近实时监测的要求.The Multi-level Negative Selection Algorithm was improved by introducing a new detector model and generating method.The new algorithm can solve the problem of a high number of detectors.Therefore,both the generation efficiency and the detecting efficiency are improved as well.The experimental results prove that both the number of detectors and the time for generating detectors are decreased greatly by the improved algorithm on the premise that the precision of detection is unchanged,making it approach the real-time detection.

关 键 词：入侵检测 人工免疫 检测器 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]