An Encrypted Traffic Classification Model Based on Improved Dense Network and Bidirectional Gated Recurrent Unit
Yansen Zhou, Songyan Zhou, Jianquan Cui · 2024
Aiming at the problems of incomplete data feature extraction and long training time in the current encrypted traffic classification model, an encrypted traffic classification model based on improved densely connected network and bidirectional gated recurrent unit is proposed. The model first uses one-dimensional convolution and LeakyReLU in the dense connection part to improve the learning ability of neurons, extract the spatial features of the data and reduce the training time of the model; then a bidirectional gated recurrent unit is added to extract the complete time series features of the data; Finally, the data features extracted by the two modules are fused and input into the classifier to complete the traffic classification. The experimental results show that the classification accuracy of the improved model is 96.4%, the accuracy rate is 96.8%, and the recall rate is 96.6% in the classification of complex types of encrypted traffic. Compared with the existing models, the performances of improved model are improved.