摘要
分析了分布式拒绝服务(Distributed Denial of Service,DDoS)攻击原理及其攻击特征,从提高检测响应时间和减少计算复杂性的角度提出了一种新的DDoS攻击检测方法。该方法基于DDoS攻击的固有特性,从IP连接数据的统计分析中寻找能够描述系统正常行为的分布规律,建立基于统计分析的DDoS攻击检测模型。实验结果表明,该方法能快速有效地实现对DDoS攻击的检测,并对其他网络安全检测具有指导作用。
Distributed Denial of service （DDoS） attacks are a major threat to security of computer network.This paper analyzes DDoS attack scenario and attack signature.Then,a novel scheme for early detection of DDoS attacks is proposed,which uses the probability distributions of normal behavior based on statistical character of＇ IP connections on the computer network.The experi- mental results show the effectiveness of our scheme in early detecting DDoS attacks.Also,this scheme can be applied to other network security detection research.
出处
《计算机工程与应用》
CSCD
北大核心
2007年第33期167-169,210,共4页
Computer Engineering and Applications
基金
河南科技大学博士科研启动基金(06-6)
关键词
分布式拒绝服务
统计分析
攻击检测
Distributed Denial of Service
statistical analysis
attack detection
