| 摘 要: 针对传统的入侵检测技术在大容量网络数据时存在检测性能不足的缺点，研究了一种基于聚类分析算法的 新型入侵检测模型，通过聚类分析算法对多维数据进行分析，当不满足聚类要求时，归并邻近数据再次聚类。最后，设 计了与K-means算法的对比仿真实验，实验结果表明，基于聚类分析的模型能够有效检测出异常序列，能够抵抗异常攻击。 |
| 关键词: 入侵检测；聚类分析；网络安全 |
|
中图分类号: TP393.08
文献标识码: A
|
|
| A Intrusion Detection System Based on Cluster Analysis |
|
FU Mingbai
|
( School of Information Science and Technology, Zhaotong University, Zhaotong 657000, China)
|
| Abstract: Since conventional intrusion detection systems can't meet high demands of the network security,a new intrusion detection method based on clustering algorithm for intrusion detection system is designed in order to cluster analysis high dimensional data,and merge data nearly if cluster condition is not qualified.After stimulate experiment compared with K-means algorithm,the result shows this detection model can detect abnormal attack effectively. |
| Keywords: intrusion detection;cluster analysis;network security |
