Efficient encrypted traffic identification plays an indispensable role in the cybersecurity ecosystem. With the rapid advancement of AI technologies, a growing number of machine learning-based and deep learning-based approaches have emerged. Among them, the pretraining-finetuning paradigm has become increasingly popular, as labeling traffic data is costly while large-scale unlabeled traffic data are easily accessible. However, existing pretraining frameworks largely rely on reconstruction tasks, which limit the model’s ability to focus on critical information within traffic data, thereby hindering recognition performance. To address this issue, we propose PaT, an enhanced pretraining framework for encrypted traffic analysis. Specifically, we introduce a novel contrastive learning task to complement the reconstruction objective and strengthen representation learning. Meanwhile, we integrate the Mamba-2 module into both encoder and decoder designs and develop a new representation extraction paradigm to expand the model’s global receptive field. Extensive experiments demonstrate the effectiveness of PaT, achieving superior encrypted traffic recognition performance while maintaining a lightweight architecture. • We propose a new pretraining architecture, PaT, for encrypted traffic analysis, introducing a self-supervised contrastive learning task to complement reconstruction learning and enhance the models focus on key traffic content. • We are the first to introduce Mamba-2 into the encrypted traffic recognition task to build the encoder and decoder, achieving improved performance while maintaining a lightweight design. • We optimize the representation extraction paradigm to enhance the models understanding of global traffic information. • We conduct extensive experiments demonstrating that PaT achieves improved encrypted traffic classification performance while maintaining a lightweight design.
Song et al. (Wed,) studied this question.