面向网络空间防御的越权漏洞对抗机器学习检测系统  

Machine Learning Detection System of Unauthorized Vulnerability Antagonism for Cyberspace Defense

作　　者：金磊[1] JIN Lei(Department of Primary Education,Aksu Education Institute,Aksu 843000,China)

机构地区：[1]阿克苏教育学院,初等教育系,新疆阿克苏843000

出　　处：《微型电脑应用》2025年第2期292-296,共5页Microcomputer Applications

摘　　要：为了提高网络空间越权漏洞检测的准确性,设计面向网络空间防御的越权漏洞对抗机器学习检测系统。在系统感知层的数据采集模块应用互联网控制消息协议(ICMP)扫描技术采集网络安全漏洞数据,聚类分析清洗数据;由网络通信层将数据发送给存储层存储;在应用层的越权漏洞检测模块中,以数据调用模块调度数据输入条件残差生成对抗网络,训练后输出越权漏洞识别结果。实验结果表明,所设计系统可有效识别与检测网络空间环境中存在的越权漏洞,检测准确性较高。To improve the accuracy of detecting unauthorized vulnerabilities in cyberspace,a machine learning detection system for defending unauthorized vulnerabilities in cyberspace is designed.In the data acquisition module of the system perception layer,ICMP scanning technology is used to collect network security vulnerability data,and cluster analysis is used to clean the data.The data is sent to the storage layer for storage by the network communication layer.In the unauthorized vulnerabilities detection module of the application layer,the adversarial network is generated by scheduling data input condition residuals using the data call module,and the unauthorized vulnerabilities identification results are output after training.The experimental results show that the system can effectively identify and detect unauthorized vulnerabilities in the cyberspace environment,with high detection accuracy.

关 键 词：网络空间防御 越权漏洞 漏洞检测 机器学习 对抗网络 残差单元 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]