基于知识图谱的DDoS攻击源检测研究  

DDoS Attack Detection Based on Knowledge Graph

作　　者：陈佳 Chen Jia(College of Cybersecurity,Sichuan University,Chengdu 610065)

机构地区：[1]四川大学网络空间安全学院

出　　处：《信息安全研究》2020年第1期91-96,共6页Journal of Information Security Research

摘　　要：分布式拒绝服务(DDoS)是黑客常用的网络攻击之一.提出了一种基于知识图谱的DDoS攻击检测方法,针对TCP流量的DDoS攻击,使用知识图谱的形式表述2个主机之间TCP流量的通信过程,在此基础上进行DDoS攻击检测,判断该源主机是否是DDoS攻击的发起者.使用加州大学洛杉矶分校数据集(UCLA CSD packet traces)进行实验,实验表明,提出的检测方法能够有效检测出发起DDoS的攻击源.Distributed denial of service(DDoS)is one of the most common network attacks by hackers.In this paper,a DDoS attack detection method based on knowledge graph is proposed,which is mainly aimed at the DDoS attack of TCP traffic.Knowledge graph is used to express the communication process of TCP traffic between two hosts,DDoS attack detection is carried out to determine whether the source host is the initiator of DDoS attack base on that.Experiments with UCLA CSD Packet Traces have shown that the proposed approach can effectively detect the source of DDoS attacks.

关 键 词：DDOS攻击 知识图谱 入侵检测 TCP流量 源检测 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]