Only Header: a reliable encrypted traffic classification framework without privacy risk

Encrypted traffic classification plays a critical role in network management, providing appropriate Quality-of-Service and Network Intrusion Detection. Conventional port-based deep packet inspection approaches cannot classify encrypted effectively. Methods based on machine learning can by extracting statistical features of the flow. However, they require manual extraction features. Recent studies show that are compelling for task. They automatically learn raw without feature extraction. these still take payload as model input, which may cause privacy risks. Besides, massive causes great storage pressure classification. In this paper, we propose reliable framework only using flow header called Only Header, avoids risks achieves lightweight storage. Firstly, introduce twice segmentation mechanism to dilute interference increase weight effective traffic. Then, use capsule neural networks (CapsNet) spatial byte header. The Header’s effectiveness is compared with other methods two public datasets, including ISCX VPN-nonVPN Tor-nonTor datasets. experimental results demonstrate Header outperforms state-of-the-art methods.