首页|基于CNN-LSTM模型的入侵检测算法研究

基于CNN-LSTM模型的入侵检测算法研究

扫码查看
随着云技术的广泛应用,网络入侵检测系统越来越受到人们的欢迎.一般的入侵检测数据集都具有空间性和时序性,因此利用卷积神经网络(CNN)以及长短期记忆网络(LSTM)的强大特点,构建了一个深度学习模型 CNN-LSTM来学习数据的时空属性.首先使用卷积神经网络对流量数据做特征选择,并在卷积层后加入了 Dropout层防止过拟合.然后利用长短期记忆网络的时间序列学习特点对卷积神经网络筛选后的特征进行学习与分类,以提高网络检测的效率和准确率.经实验证明使用的模型的检测结果优于单一 CNN和LSTM模型.
Research on intrusion detection algorithm based on CNN-LSTM model
With the wide application of cloud technology,network intrusion detection systems are be-coming more and more popular.However,general intrusion detection datasets are spatial and tem-poral,in this paper,we utilize the powerful features of Convolutional Neural Networks(CNN)as well as Long Short-Term Memory Networks(LSTM)to construct a deep learning model,CNN-LSTM,to learn the spatiotemporal properties of data.Firstly,a convolutional neural network is used to do feature selection on the traffic data,and a Dropout layer is added after the convolutional layer to prevent over fitting.Then the time series learning characteristics of the Long Short Term Memory Network are used to learn and classify the features screened by the Convolutional Neural Network in order to improve the efficiency and accuracy of the network detection.It is proved ex-perimentally that the detection results of the model used in this paper are better than single CNN and LSTM models.

Network IntrusionDeep LearningConvolutional Neural NetworkLong Short Term Mem-ory Network

高鑫泽、吕国、杨宵、张建成

展开 >

河北建筑工程学院,河北 张家口 075000

辽宁省铁岭市昌图县第一高级中学,辽宁 铁岭 112500

网络入侵 深度学习 卷积神经网络 长短期记忆网络

2024

河北建筑工程学院学报
河北建筑工程学院

河北建筑工程学院学报

影响因子:0.502
ISSN:1008-4185
年,卷(期):2024.42(3)