Abstract: For traditional intrusion detection algorithms, the lack of self-adaptive and intelligent has become increasingly prominent when they cope with unknown attacks. This paper presents a new unsupervised, adaptive detection algorithm, Clustering Quantum Genetic Algorithm(CQGA). This algorithm employs the Euclidean distance between the samples as the standard of similarity measurement. Furthermore, it can auto-classify the sample sets in the unsupervised condition through quantum genetic algorithm finding the cluster center. Experimental result indicates that the algorithm can classify the test data set accurately, and solve the problem of self-adapting and intelligent effectively.

Key words: intrusion detection, quantum genetic algorithm, clustering quantum genetic algorithm

摘要： 针对传统入侵检测算法当面临未知攻击时所缺乏的自适应性和智能化日益突出的问题，提出一种新的无监督、自适应的检测算 法——量子遗传聚类算法(CQGA)。该算法利用各实例之间的欧氏距离作为相似度量标准，通过量子遗传算法寻找聚类中心以达到在无监督的条件下对数据集自动分类的目的。实验仿真结果显示，该算法能较为准确地对测试数据集进行分类，有效地解决自适应性和智能化问题。

关键词: 入侵检测, 量子遗传算法, 量子遗传聚类算法

CLC Number: 

• TP301.6