基于广义神经网络的网络攻击检测与分类方法  

Detection and Classification Method of Network Attacks Based on Generalized Neural Networks

作　　者：张明明 刘凯 李贤慧 许梦晗 顾颖程 张见豪 程环宇 王永利[4] Zhang Mingming;Liu Kai;Li Xianhui;Xu Menghan;GuYingcheng;Zhang Jianhao;Cheng Huanyu;Wang Yongli(Information Communication Branch,State Grid Jiangsu Electric Power Co.,Ltd.,Nanjing 210024;Jiangsu Ruizhong Data Co.,Ltd.,Nanjing 210012;State Grid Electric Power Research Institute Co.,Ltd.,Nanjing 211106;School of Computer Science and Technology,Nanjing University of Science and Technology,Nanjing 210094)

机构地区：[1]国网江苏省电力有限公司信息通信分公司,南京210024 [2]江苏瑞中数据股份有限公司,南京210012 [3]国网电力科学研究院有限公司,南京211106 [4]南京理工大学计算机科学与工程学院,南京210094

出　　处：《信息安全研究》2023年第6期593-601,共9页Journal of Information Security Research

基　　金：国网公司科技指南项目(5700-202218185A-1-1-ZN);国家自然科学基金重点项目(61941113)。

摘　　要：如今虚拟世界日趋复杂,网络攻击和新出现的安全威胁逐步增加,因此需要研究针对网络攻击的智能化检测和分类方法,以全面地观察网络活动,阻止恶意行为.提出了一种基于广义回归神经网络(generalized regression neural networks,GRNN)建立的入侵检测模型,对恶意网络攻击进行智能化检测和分类,并使用主流的NSL-KDD数据集进行了测试.实验结果表明,所提出的技术相较于目前的其他攻击检测技术,能够更加有效地对恶意行为进行识别与分类.Nowadays,the virtual world is becoming more and more complex,and network attacks and emerging security threats are gradually increasing.Therefore,it is necessary to study intelligent detection and classification methods for network attacks to comprehensively observe network activities and prevent malicious behaviors.It is proposed that an intrusion detection system based on generalized regression neural network(GRNN)in this paper,which can intelligently detect and classify malicious network attacks,and be tested by using today’s mainstream NSL-KDD dataset.The experimental results show that the technology proposed in this paper can identify and classify malicious behaviors more effectively than other current attack detection technologies.

关 键 词：智能化攻击检测 入侵检测系统 广义回归神经网络 恶意行为 检测 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]