关键词: 分布式拒绝服务攻击 流特征分布熵 k&mdash, means聚类算法

Abstract:

Distributed Denial-of-Service （DDoS） attacks are a critical threat to the Internet. This paper analyses the essential characteristics of DDoS attack    s, and presents a novel detection scheme by combining the traffic volume with traffic feature distribution entropy. Furthermore, we improve the k-means    clustering algorithm, which is used to establish the attack detection model. Finally we test our detection model on the LLDOS1.0 dataset. The experiment    al results indicate that the model leads to an improved accuracy of attack detection.

Key words: DDoS attack, traffic feature distribution entropy, k-means clustering algorithm