At present, deep learning has achieved excellent performance in the field of network traffic classification. However, deep learning relies on massive data-driven classification models. When the data set is small, it is usually hindered. In order to solve this problem, few-shot network traffic classification technology based on deep learning has been gradually studied. In this paper, we conducted a comprehensive survey to fully understand the few-shot traffic classification techniques. We first define the few-shot traffic classification problem. Based on how to solve the contradiction between the few-shot data set and the large number of parameters to be trained in the model, we classify the current few-shot traffic classification method based on deep learning from two perspectives: (1) Data augmentation method, which uses the method of expanding the data set to enhance the supervision experience. In this paper, the methods based on data enhancement are divided into two categories based on GAN and feature transformation.(2) Model-based methods. Model-based methods are divided into three categories: transfer learning, metric learning and meta-learning. And discuss the advantages and disadvantages of each classification method. Finally, the results are summarized and the future development direction of few-shot network traffic classification technology based on deep learning is prospected.
Network traffic classification plays an important role in network resource management and security. The application of encryption techniques and the rapid increase in the size of network traffic have placed higher demands on traffic classification. In this paper, we design multi-headed attention (MHA) and deep metric learning (DML) in our model for network traffic classification. In addition, MHA-DML also extracts more subtle and highly differentiated features through the improved triplet measurement loss. Experimental results demonstrate that the model achieves the best classification on all three publicly available web traffic datasets. The MHA-DML guarantees detection accuracy even when facing a classification task with many categories.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.