From traffic classes to content: A hierarchical approach for encrypted traffic classification

Li, Ying; Huang, Yung-Fu; Seneviratne, Suranga; Thilakarathna, Kanchana; Cheng, A.; Jourjon, Guillaume; Webb, Darren; Smith, David B.; Xu, Richard Yi Da

doi:10.1016/j.comnet.2022.109017

Cited by 7 publications

(7 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We evaluate the proposed W-VAE algorithm on a synthetic multi-layer signature dataset. The synthetic dataset consists of a real-world video traffic dataset [48] paired with a simulated channel state information (CSI). A sample of the video traffic dataset has 200 binary sequences of the uplink and downlink packet lengths, and we pre-process them into traffic states (0: idle, 1: non-zero packet lengths).…”

Section: Discussionmentioning

confidence: 99%

“…We combine the two datasets with the video traffic sequences as the first multi-layer feature and the CSI as the second one. There are 10 videos traffic sequences collected from YouTube as detailed in [48,Sec 3.2]. After pairing each video sample with a simulated CSI, there are 2557 training samples and 638 testing samples with each set uniformly distributed across the 10 classes.…”

Section: Discussionmentioning

confidence: 99%

“…The loss function for the supervised variant of the W-VAE follows (19) and we denote it as W-VAE (Supervised). We compared the W-VAE (Supervised) with a baseline approach [48]. This compared method implemented a deep neural network (DNN)based classifier and empirically demonstrated the state-of-theart classification accuracy with the video traffic dataset.…”

Section: A Supervised Device Identificationmentioning

confidence: 99%

“…This baseline only uses traffic states as the inputs and is denoted as Traffic Only baseline. The DNN consists of two fully connected layers with dropout [50], the details are referred to [48,Fig. 5].…”

Section: A Supervised Device Identificationmentioning

confidence: 99%

“…For the W-VAE (supervised), 8 mini-batch size is used with a fixed learning rate 10 −3 for the standard ADAM optimizer [51]. As for the baseline, the configurations follow [48,Section 4.2.3]. We report the model with the maximum accuracy from 25 trained ones.…”

Section: A Supervised Device Identificationmentioning

confidence: 99%

See 4 more Smart Citations

The Wyner Variational Autoencoder for Unsupervised Multi-Layer Wireless Fingerprinting

Huang¹,

Dahanayaka²,

Thilakarathna³

et al. 2023

Preprint

View full text Add to dashboard Cite

Wireless fingerprinting refers to a device identification method leveraging hardware imperfections and wireless channel variations as signatures. Beyond physical layer characteristics, recent studies demonstrated that user behaviours could be identified through network traffic, e.g., packet length, without decryption of the payload. Inspired by these results, we propose a multi-layer fingerprinting framework that jointly considers the multi-layer signatures for improved identification performance. In contrast to previous works, by leveraging the recent multiview machine learning paradigm, i.e., data with multiple forms, our method can cluster the device information shared among the multi-layer features without supervision. Our informationtheoretic approach can be extended to supervised and semisupervised settings with straightforward derivations. In solving the formulated problem, we obtain a tight surrogate bound using variational inference for efficient optimization. In extracting the shared device information, we develop an algorithm based on the Wyner common information method, enjoying reduced computation complexity as compared to existing approaches. The algorithm can be applied to data distributions belonging to the exponential family class. Empirically, we evaluate the algorithm in a synthetic dataset with real-world video traffic and simulated physical layer characteristics. Our empirical results show that the proposed method outperforms the state-of-the-art baselines in both supervised and unsupervised settings.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%