计算机工程

• 人工智能及识别技术 • 上一篇    下一篇

基于协同进化的免疫检测器分布优化算法

刘海龙1,2,张凤斌1,席 亮1   

  1. (1. 哈尔滨理工大学计算机科学与技术学院,哈尔滨 150080;2. 哈尔滨师范大学计算机科学与技术学院,哈尔滨 150025)
  • 收稿日期:2012-10-16 出版日期:2013-11-15 发布日期:2013-11-13
  • 作者简介:刘海龙(1976-),男,讲师、博士研究生,主研方向:网络与信息安全,模式识别;张凤斌(通讯作者),教授、博士生导师;席 亮,讲师、博士
  • 基金项目:
    国家自然科学基金资助项目(60671049, 61172168)

Immune Detector Distribution Optimization Algorithm Based on Co-evolution

LIU Hai-long 1,2, ZHANG Feng-bin 1, XI Liang 1   

  1. (1. College of Computer Science and Technology, Harbin University of Science and Technology, Harbin 150080, China; 2. College of Computer Science and Technology, Harbin Normal University, Harbin 150025, China)
  • Received:2012-10-16 Online:2013-11-15 Published:2013-11-13

摘要: 为解决免疫实值检测器的黑洞问题,分析检测器规模对检测性能的影响,提出一种基于协同进化的免疫实值检测器分布优化算法。将检测器集分成不同子集,寻找每个子集的最优个体,利用各子集间的相互作用与影响对各子集进行优化处理,取并集构成完整检测器集。实验结果表明,与否定选择算法相比,该算法不仅可以有效减少黑洞的产生,并且能以较少的检测器精确地覆盖非自体空间,从而提高检测器性能。

关键词: 入侵检测, 人工免疫, 检测器, 分布优化, 否定选择算法, 协同进化

Abstract: In order to avoid lots of holes among mature immune detectors in intrusion detection, analyzing the relationship with number of detector and detection performance, an immune detector distribution optimization algorithm based on co-evolution is proposed in this paper. It divides the detectors into different subsets, looks for the best individual of each subset, optimizes the subsets by using the interaction between different subsets, and forms a complete detector set by taking union set. Experimental results demonstrate that this algorithm not only can decrease the holes, but also can achieve a more precise coverage of the nonself space with fewer detectors, and can increase the detector performance.

Key words: intrusion detection, artificial immunity, detector, distribution optimization, Negative Selection Algorithm(NSA), co- evolution

中图分类号: