网络注入式攻击检测方案的研究与改进  

Research and Improvement of Network Injection Attack Detection Scheme

作　　者：刘芬 余铮 廖荣涛 徐焕 代荡荡 LIU Fen;YU Zheng;LIAO Rongtao;XU Huan;DAI Dangdang(Information Communication Company State Grid Hubei Electric Power Company,Wuhan 430077)

机构地区：[1]国网湖北省电力有限公司信息通信公司,武汉430077

出　　处：《计算机与数字工程》2020年第4期914-917,951,共5页Computer & Digital Engineering

摘　　要：目前各种网络攻击行为与日俱增,给网络和数据安全造成了严重的影响。在所有的攻击行为中,注入式攻击由于其隐蔽性高,防火墙拦截效果较差,已经造成了大量的经济损失。论文针对该类型网络攻击行为展开研究,提出对文本分类技术中的KNN算法进行针对性的改进,用以实现对注入式攻击行为的快速检测。随后以Web日志为检测对象,设计了相关的检测模型,最终通过仿真实验论证了本方法的准确性与可靠性。At present,various kinds of network attacks are increasing day by day,which has seriously affected the network and data security.Among all kinds of attacks,injection attack has caused a lot of economic losses because of its high concealment and poor firewall interception effect.This paper aims at the research of this type of network attack,and proposes a specific improvement on the KNN algorithm in the text classification technology,so as to realize the rapid detection of the injecting attack behavior.Subsequently,the Web log is taken as the detection object,and the relevant detection model is designed.Finally,the accuracy and reliability of this method are demonstrated by simulation experiments.

关 键 词：注入式攻击 KNN 样本区域裁剪 WEB日志 检测模型 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]