摘要
Abstract
With the rapid development of Internet technology,the types of network applications are becoming more and more diversified and complex,and the traditional traffic identification methods based on port number and statistical features are facing serious challenges in accuracy and real-time.To address this problem,this study proposes a real-time network traffic identification method based on deep feature learning.The method designs a two-channel convolutional neural network architecture,which can automatically extract deep features from raw traffic data and realize classification.The experimental results show that the method achieves a high accuracy of 95.8%and an average recognition delay of only 1.2 ms in the recognition test of 10 mainstream network applications,which provides a strong technical support for network security monitoring and traffic management.关键词
深度学习/网络流量识别/实时处理/特征提取/卷积神经网络/注意力机制Key words
deep learning/network traffic identification/real-time processing/feature extraction/convolutional neural network/attention mechanism分类
信息技术与安全科学
