摘要
针对如何进一步提高网络入侵检测性能,提出一种基于改进深度神经网络的网络入侵检测方法。首先,对无监督稀疏自编码器(SAE)进行L1正则化以增强数据自动编码器的稀疏性;然后,将无监督SAE引入深度神经网络建立入侵检测网络入侵模型,采用深度神经网络完成对网络攻击入侵的预测和分类,通过分类完成对入侵攻击的特征提取。最后,为了验证模型在检测率和低误报率方面的优越性,论文分别采用了KDDCup99、NSL-KDD等数据集进行验证。结果表明与传统方法相比,新提出的方法在准确率、检测率有约10%的提升。
In order to further improve the performance of network intrusion detection,this paper proposes a network intrusion detection method based on improved deep neural network.First,the unsupervised sparse self-encoder(SAE)was regularized by L1 to enhance the sparsity of the automatic data encoder;Then,the unsupervised SAE was introduced into the deep neural network to establish the intrusion detection network intrusion model.The deep neural network was used to complete the prediction and classification of network attack intrusion,and the feature extraction of intrusion attack was completed through classification.Finally,in order to verify the superiority of the model in terms of detection rate and low false positive rate,the paper used KDDCup99,NSL-KDD and other data sets for validation.The results show that compared with the traditional methods,the accuracy and detection rate of the proposed method are improved by about 10%.
作者
宋吉飞
郭金雷
王蓉
孙成
SONG Ji-fei;GUO jin-lei;WANG Rong;SUN Cheng(Ningxia Zhongwei New Internet Exchange Center Co.,Ltd,Zhongwei Ningxia 755001,China;Shanghai Institute of Mechanical and Electrical Engineering,Shanghai 201100,China)
出处
《计算机仿真》
2024年第5期405-409,共5页
Computer Simulation
基金
宁夏回族自治区产业创新重点任务揭榜公关项目(2021020301)。
关键词
深度神经网络
网络预警
入侵检测
Deep neural network
Network early warning
Intrusion detection
