Abnormal Traffic Detection Based on Data Augmentation and Hybrid Neural Network

Aiming at the problem of low detection rate caused by class imbalance and limited detection ability of single model in abnormal traffic detection, an abnormal traffic detection model based on hybrid sampling and deep learning has been proposed. On the basis of using Tomek-Link method to filter noise samples in data, CTGAN technology is used to generate minority samples that conform to the original data distribution, make the samples tend to balance. By combining the advantages of convolutional neural network and bidirectional long short term memory network, the spatial and temporal features of network data are extracted in depth to further improve the overall detection performance of the model. The multi-classification experimental results on the NSL-KDD dataset show that the accuracy and recall of the proposed algorithm. The rates were increased by 12.06% and 11.53%, respectively, which verified the effectiveness of the model.