基于Web数据流技术的网络入侵检测研究
Study on network intrusion detection based on Web data streams technology
-
摘要: 针对传统多遍扫描数据库的挖掘技术构建的入侵检测模型已不能满足Web数据流高速并且无限到达的需要,根据多维频繁模式的特点,提出了一种新的入侵检测模型和一种新型数据结构SW.Tree,并给出了一种基于滑动窗口树的挖掘频繁项集的新型算法AFP.对不同流量数据的实验结果表明该模型有较高的报警率和较低的误报率.Abstract: Aiming at the problem that intrusion detection model constructed by mining technique with multiscanning to databases has not met the needs of high-speed and unlimited for Web data streams.Based on the characteristic of multi-dimension frequent patterns,a new intrusion detection model and data structure called SW.Tree was proposed,and a new algorithm AFP mining frequent patterns from data streams based on sliding window tree was designed.The different flow experiments data showed that the model had high alarm rate and low false alarm rate.
-
Key words:
- Web data streams /
- network intrusion detection /
- frequent patterns /
- sliding window
计量
- PDF下载量: 32
- 文章访问数: 1101
- 引证文献数: 0