Research On Detection Method Of Network Abnormal Traffic Based On LSTM

With the emergence of a variety of new network attack methods,the traditional method of establishing detection model based on existing attack types has become inadvisable.Due to its powerful complex function fitting ability,deep learning has been increasingly used in the field of abnormal traffic detection in recent years.At the same time,the abnormal traffic detection method based on deep learning has certain problems,such as poor detection performance and low accuracy.In order to improve the accuracy of anomaly traffic detection,this thesis design an anomaly traffic detection model based on LSTM(long short-term memory)neural network,and uses convolution neural network to improve the model.In order to solve the problem of anomalous traffic detection,a model based on LSTM is proposed to solve the problem of long-distance dependence of traffic characteristics.Based on this model,the VGG-LSTM(Visual Geometry Group)model is designed by combining convolutional neural network with long and short-term memory network,which effectively exerts the learning ability of convolution neural network to local spatial features.Finally,this thesis further improves the model by using the ability of bi-directional long short-term memory network to learn information before and after,and designs a VGG-BILSTM model.In order to verify the performance of the proposed model,experiments are carried out on NSL-KDD data set.The experimental results show that compared with the original model,the accuracy of the VGG-BILSTM model for traffic detection is improved from 91.48% to99.20%.The experiental results show that the feature extraction ability of the VGG-BILSTM model proposed in this thesis is strengthened,and the normal traffic and abnormal traffic can be detected effectively.