Lightweight GenAI for Network Traffic Synthesis: Fidelity, Augmentation, and Classification

Researchers developed lightweight generative AI models for creating synthetic network traffic data to address privacy concerns and data scarcity in network traffic classification. The models achieved up to 87% F1-score when classifiers were trained solely on synthetic data, with transformer-based approaches providing the best balance of accuracy and computational efficiency.

• →Lightweight GenAI models can generate synthetic network traffic that preserves both static and temporal characteristics of real data.
• →Classifiers trained exclusively on synthetic traffic achieved up to 87% F1-score when tested on real network data.
• →In low-data scenarios, GenAI-driven data augmentation improved classification performance by up to 40%.
• →Transformer-based models offered the optimal trade-off between fidelity and computational efficiency.
• →The approach addresses critical privacy requirements while mitigating data scarcity issues in network traffic analysis.