融合GSDE与GRU的计算机网络安全入侵检测模型研究  

Computer Network Security Intrusion Detection Model Integrating GSDE and GRU

作　　者：姚孝生 YAO Xiaosheng(Hefei University of Economics,Government Office,Hefei 230011,China)

机构地区：[1]合肥经济学院党政办公室,合肥230011

出　　处：《重庆科技大学学报(自然科学版)》2025年第1期99-106,共8页Journal of Chongqing University of Science and Technology(Natural Sciences Edition)

基　　金：安徽省省级科研重点项目(自然科学类)“基于集成学习的多模态虚假新闻检测技术”(2024AH051545)。

摘　　要：为了提升计算机网络安全防护能力,提出了一种融合广义状态相关探索(GSDE)与门控循环单元(GRU)的计算机网络安全入侵检测模型。模拟网络入侵过程,设定计算机网络安全入侵检测标准。动态采集计算机网络流量、用户行为及网络日志等运行数据,通过GSDE与GRU算法的融合和迭代学习提取计算机网络运行特征,采用特征匹配的方法检测计算机网络入侵状态。测试结果显示,不同工况下的优化设计模型入侵检测正确率比传统模型高,且其入侵强度检测误差平均值为0.004,说明该模型的检测性能更佳。In order to enhance the protection capability of computer network security,this paper proposes a computer network security intrusion detection model integrating GSDE and GRU,and establishes intrusion detection standards for computer network security through simulating the process of network intrusion.It dynamically collects computer network traffic,logs,and other operational data,extracts computer network operational features through the fusion and learning iteration of GSDE and GRU,and detects the intrusion status of the current network through feature matching.The result shows that the optimized design model has a higher accuracy in intrusion state detection than the traditional intrusion detection model under static and dynamic experimental conditions with the intrusion intensity detection error 0.004,indicating that the optimized design model has better performance of intrusion detection.

关 键 词：GSDE算法 GRU算法 计算机网络 安全入侵 入侵检测模型 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]