基于卷积神经网络的网络安全态势感知研究  

Research on Network Security Situation Awareness Based on Convolutional Neural Networks

作　　者：袁华刚[1] YUAN Huagang(Communication Information Center,Hubei Provincial Department of Transport,Wuhan Hubei 430022,China)

机构地区：[1]湖北省交通运输厅通信信息中心,湖北武汉430022

出　　处：《信息与电脑》2024年第7期206-208,共3页Information & Computer

摘　　要：本文提出了一种基于网络攻击影响的态势量化评估指标,旨在精确衡量网络安全状况。基于此,设计了相应的态势值计算方法,并通过新小型学习-知识发现和数据挖掘(New Smaller Learning-Knowledge Discovery and Data Mining,NSL-KDD)数据集的验证,充分证明了所提出的选择性卷积神经网络模型在攻击识别方面拥有更高的准确度。实验结果显示,其态势评估结果与真实网络环境中的情况更为一致。The article proposes a situational quantification index based on the impact of network attacks and designs a corresponding situational value calculation method.The validation of the NSL(New Small Learning)KDD(Knowledge Discovery and Data Mining)dataset confirms that the proposed selective convolutional neural network model has higher accuracy in attack recognition,and its situational assessment results are more consistent with the actual situation.

关 键 词：卷积神经网络 网络安全 态势感知评估 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]