应用深度自编码网络的局域网空间入侵监测系统设计  

Design of LAN Spatial Intrusion Monitoring System Based on Deep Self-coding Network

作　　者：李芳 LI Fang(Beijing Haidian District Staff University,Beijing 100083 China)

机构地区：[1]北京市海淀区职工大学,北京100083

出　　处：《自动化技术与应用》2024年第3期91-95,共5页Techniques of Automation and Applications

摘　　要：为有效追赶恶意代码更新速度,实时监测多种网络攻击,设计应用深度自编码网络的局域网空间入侵监测系统。先采用库函数以及FHW抓包方法抓取局域网各监测点的数据包,提取数据包数据特征;然后结合支持向量回归预测算法构建入侵监测模型,完成入侵数据监测;最后将入侵监测结果保存至数据库中,判定是否需向用户发出预警,由此完成局域网空间入侵监测。经实验验证该系统能够监测到多种攻击手段,且监测完成时间较短,能够快速实现监测,且在24 h内监测到的入侵攻击强度与实际攻击强度未存在较大差距。In order to effectively catch up with the update speed of malicious code and monitor various network attacks in real time,a LAN space intrusion monitoring system using deep self-coding network is designed.Firstly,the library function and FHW packet capture method are used to capture the data packets of each monitoring point in the local area network,and the data characteristics of the data packets are extracted,then the intrusion monitoring model is constructed combined with the support vector regression prediction algorithm to complete the intrusion data monitoring.Finally,the intrusion monitoring results are saved to the database,determine whether to issue an early warning to the user,thereby completing the LAN space intrusion monitoring.Experiments verifiy that the system can monitor a variety of attack methods,and the monitoring completion time is short,which can quickly realize monitoring,and there is no large gap between the intrusion attack intensity and the actual attack intensity monitored within 24 hours.

关 键 词：深度自编码网络 局域网 空间入侵监测 数据包捕获 攻击强度 响应模块 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]