Related Experiment Videos
Collection of datasets with DNS over HTTPS traffic
Kamil Jeřábek1, Karel Hynek2,3, Tomáš Čejka3
1Faculty of Information Technology BUT, Bozetechova 1/2, Brno 612 00, Czech Republic.
Data in Brief
|June 9, 2022
Summary
A new large-scale dataset of DNS over HTTPS (DoH) and non-DoH HTTPS traffic was created. This dataset supports the evaluation and development of network traffic classifiers for privacy-aware applications.
Area of Science:
- Computer Science
- Network Security
- Data Science
Background:
- DNS over HTTPS (DoH) is increasingly adopted for enhanced user privacy.
- Network monitoring and traffic classification are crucial for understanding DoH dissemination.
- Existing datasets may not fully represent real-world DoH traffic complexities.
Purpose of the Study:
- To introduce a comprehensive, large-scale dataset for evaluating DNS over HTTPS (DoH) traffic classifiers.
- To provide a benchmark for comparing the performance of various network monitoring tools.
- To facilitate the development of new methods for classifying DoH and non-DoH HTTPS traffic.
Main Methods:
- Collected network traffic data encompassing both DoH and non-DoH HTTPS connections.
- Included traffic from multiple DoH providers and clients to capture implementation variations.
- Incorporated traffic from controlled environments and real ISP networks for diverse scenarios.
Main Results:
- A novel, large-scale dataset comprising two distinct traffic classes (DoH and non-DoH HTTPS) has been generated.
- The dataset captures the nuances of various DoH implementations and configurations.
- It includes a wide range of non-DoH HTTPS traffic from diverse network applications.
Conclusions:
- The created dataset offers valuable real-world network traffic data.
- It is suitable for evaluating existing classifiers and developing novel network monitoring techniques.
- This resource will advance research in privacy-aware network applications and traffic analysis.