Research And Implementation Of Network Intrusion Detection Based On CNN-GRU

Network intrusion detection technology mainly refers to the use of existing behavior models,security logs or other available information to detect malicious use of the system when there is malicious use of computer and network resources.Network intrusion detection technology is one of the important technologies to ensure the safety of industrial production,civil and other fields.In recent years,the amount of Internet data has grown rapidly,and the way of network intrusion has also evolved.Traditional network security technology can no longer fully guarantee the security of computer systems.An intrusion detection algorithm with strong detection ability,fast response,good universality and active defense function is designed and developed,which will have broad application and development space in protecting the safe operation of the system.In the process of intrusion detection,this paper maps the features of the data to be detected into a graphical way,uses deep learning methods to extract features,and obtains a classification model.Firstly,the basic framework of network intrusion detection is analyzed.The data imbalance problem of the classic data set NSL-KDD in the field of network intrusion detection is analyzed and the solutions are discussed.The deep learning network model involved in this paper is described.Secondly,aiming at the data imbalance problem of NSL-KDD dataset,a balanced dataset construction method based on mixed sampling is designed.This method constructs a balanced training set by using OSS algorithm to undersample most class samples and Borderline SMOTE algorithm to oversample minority class samples.According to the characteristics of network traffic data,an intrusion detection model based on convolutional neural network and gated recurrent unit is designed,which can fully learn the characteristics of data in space dimension and time dimension.Simulation results show that the model has high classification accuracy.Finally,in order to further increase the generalization ability of the model and improve the classification performance of the model,an intrusion detection model based on feature fusion is proposed.The model uses densely connection network and convolutional neural network to extract features and fuse features.Using two features at the same time can make full use of the original information of the data and prevent overfitting.The attention mechanism is introduced into the model to accelerate the convergence speed of the model.Simulation results show that the model has good generalization ability and higher classification accuracy.