关键词: 信息熵 子空间方法 大规模网络流量 异常分类

Abstract:

This paper presents an entropy-based large-scale network traffic anomaly classification method for the integrated use of the subspace method and the k -means clustering method.And classifying network traffic anomalies is realized in the experimental environment of campus networks.The experimental resul  ts show that large-scale traffic anomaly classification based on entropy not only realizes simple and has a small computation quantity,but also has a high classification precision.

Key words: （entropy,subspace method,large-scale network traffic,anomaly classification）