Research On Network Intrusion Detection Algorithm Based On Deep Learning

With the rapid development of information technology,the structure of the Internet has become more and more complex,and it has become a necessity for daily life and the backbone of daily business.However,the abnormal traffic generated in network intrusion also produces many security threats,which bring great harm to computer equipment and user privacy.Network intrusion detection can provide a certain degree of security for the Internet.This paper combines deep learning technology with network intrusion detection,and first proposes a Hybrid Convolutional Neural Network(HYBRIDCNN)method.The algorithm uses Deep Neural Network(DNN)to effectively extract global features of network traffic through one-dimensional data and uses Convolutional Neural Network(CNN)to summarize local features of network traffic through two-dimensional data.Finally,the performance of the proposed method is evaluated through experimental simulations on the UNSW?NB15dataset and KDDCup99 dataset.Experimental results show that our proposed HYBRID-CNN detection algorithm has obvious advantages in accuracy and false alarm rate,which successfully proves that it can effectively solve the problem of abnormal traffic detection.Secondly,this paper proposes an efficient Bidirectional Simple Recurrent Unit(Bi SRU)that combines feature dimensionality reduction for abnormal traffic detection.In order to perform feature reduction on the original highdimensional network traffic,the algorithm uses a Stack Sparse Autoencoder(s SAE)to extract compressed high-level features.In order to achieve efficient parallel computing and accurate feature extraction,a Simple Recurrent Unit(SRU)with a bidirectional structure is used to extract the characteristics of the flow.Finally,the experimental results show that this method achieves better results than other methods in terms of accuracy and training time.This paper uses two public network intrusion detection datasets,KDDCup99 and UNSW?NB15,to evaluate the experimental performance.The research results show that the two network intrusion detection algorithms based on deep learning proposed in this paper have achieved good detection results.