计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (11): 127-128.

• 网络、通信、安全 • 上一篇    下一篇

基于KQPSO聚类算法的网络异常检测

马汝辉,刘 渊,林 星   

  1. 江南大学 信息工程学院,江苏 无锡 214122
  • 收稿日期:2007-08-01 修回日期:2007-10-19 出版日期:2008-04-11 发布日期:2008-04-11
  • 通讯作者: 马汝辉

Network anomaly detection based on KQPSO clustering algorithm

MA Ru-hui,LIU Yuan,LIN Xing   

  1. School of Information Engineering,Southern Yangtze University,Wuxi,Jiangsu 214122,China
  • Received:2007-08-01 Revised:2007-10-19 Online:2008-04-11 Published:2008-04-11
  • Contact: MA Ru-hui

摘要: 提出一种基于KQPSO聚类算法的网络异常检测模型.该模型利用K-Means聚类算法的结果重新初始化粒子群,聚类过程都是根据数据间的Euclidean(欧几里德)距离。再通过量子粒子群优化算法(QPSO)寻找聚类中心。最后进行仿真模拟,实验结果表明,该模型对网络异常检测是有效的。

关键词: QPSO算法, 网络异常检测, K-Means, KQPSO

Abstract: Model of detecting network anomaly based on KQPSO(K-Means Quantum-behaved Particle Swarm Optimization) clustering algorithms is presented.The authors uses K-Means clustering to seed the initial swam.All the process of clustering is based on the Euclidean distance among data vector.Cluster-centroid is chosen by QPSO clustering algorithm.Finally,the experiment result shows that this model is effective for network anomaly detection.

Key words: QPSO(Quantum-behaved Particle Swarm Optimization)algorithm, network anomaly detection, K-Means, K-Means Quantum-behaved Particle Swarm Optimization(KQPSO)