基于流数据挖掘的网络流量异常检测及分析研究

Research on Anomaly Detection and Analysis of Network Traffic Based on Data Stream Mining

网络流量异常检测及分析是网络及安全管理领域的重要研究内容。本文探讨了网络流量异常的种类、网络流量异常检测的方法,分析了基于传统检测方法在网络流量异常检测应用中存在的问题。并重点对基于流数据模型的网络流量异常检测进行了研究,综述了已有流数据挖掘研究方法在网络流量异常检测中的研究进展。最后,本文对现有研究工作存在的问题及未来的研究方向进行了探讨。

Anomaly detection and analysis based on network traffic are used for network and security management. In this paper, we give a list of 4 kinds of of traffic anomalies and the methods of traffic anomaly detection are presented. Also we give an analysis of the limitations in traditional anomaly detection methods. After that, the research work focus on the traffic anomaly detection methods based on data stream mining. The available research work of data stream mining methods for anomaly detection and analysis are summarized. Finaly, we discuss the open problems and challenges in this area.