A Lightweight Multi-Classification Model for Identifying Network Application Traffic Using Knowledge Distillation
Future Internet · 07.04.2026 · 18 (4) · 197
DOI: 10.3390/fi18040197 ↗Kurz gefasst
Fachartikel in der Zeitschrift Future Internet, online erschienen am 07.04.2026. Verfasst von Zhiyuan Li und 1 weiteren Autor:innen. Die Arbeit nennt 30 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
To address the limitations of insufficient feature representation, large model size, and high deployment cost in network traffic classification, a lightweight classification framework based on multi-teacher knowledge distillation is proposed. The framework consists of two heterogeneous teacher networks and a compact student network to enable end-to-end traffic classification under constrained computational resources. The teacher networks incorporate complementary spatio-temporal modeling strategies, including a bidirectional temporal convolutional network (BiTCN) enhanced with attention mechanisms and convolutional neural network (CNN), and a parallel spatio-temporal fusion architecture integrating bidirectional long short-term memory (BiLSTM) and CNN. Knowledge from the teacher ensemble is distilled into a lightweight CNN-based student network through soft-target supervision, leading to improved generalization capability with significantly reduced model complexity. Experimental results demonstrate that effective knowledge transfer is achieved while reducing model parameters by more than 80%, and performance gains of about 1–3% are obtained compared with baseline methods. These results indicate strong potential for practical deployment in resource-constrained network environments.
Zitieren
APA
Li, Z., & Feng, Y. (2026). A Lightweight Multi-Classification Model for Identifying Network Application Traffic Using Knowledge Distillation. Future Internet, 18(4), 197. https://doi.org/10.3390/fi18040197
Vancouver
Li Z, Feng Y. A Lightweight Multi-Classification Model for Identifying Network Application Traffic Using Knowledge Distillation. Future Internet. 2026;18(4):197. doi:10.3390/fi18040197
BibTeX
@article{Li2026,
author = {Li, Zhiyuan and Feng, Yonghao},
title = {A Lightweight Multi-Classification Model for Identifying Network Application Traffic Using Knowledge Distillation},
journal = {Future Internet},
year = {2026},
volume = {18},
number = {4},
pages = {197},
doi = {10.3390/fi18040197}
}
EUVIMED