Padding Ain't Enough: Assessing the Privacy Guarantees of Encrypted DNS
July 02, 2019 ยท Declared Dead ยท ๐ FOCI @ USENIX Security Symposium
"No code URL or promise found in abstract"
Evidence collected by the PWNC Scanner
Authors
Jonas Bushart, Christian Rossow
arXiv ID
1907.01317
Category
cs.CR: Cryptography & Security
Citations
58
Venue
FOCI @ USENIX Security Symposium
Last Checked
3 months ago
Abstract
DNS over TLS (DoT) and DNS over HTTPS (DoH) encrypt DNS to guard user privacy by hiding DNS resolutions from passive adversaries. Yet, past attacks have shown that encrypted DNS is still sensitive to traffic analysis. As a consequence, RFC 8467 proposes to pad messages prior to encryption, which heavily reduces the characteristics of encrypted traffic. In this paper, we show that padding alone is insufficient to counter DNS traffic analysis. We propose a novel traffic analysis method that combines size and timing information to infer the websites a user visits purely based on encrypted and padded DNS traces. To this end, we model DNS sequences that capture the complexity of websites that usually trigger dozens of DNS resolutions instead of just a single DNS transaction. A closed world evaluation based on the Alexa top-10k websites reveals that attackers can deanonymize at least half of the test traces in 80.2% of all websites, and even correctly label all traces for 32.0% of the websites. Our findings undermine the privacy goals of state-of-the-art message padding strategies in DoT/DoH. We conclude by showing that successful mitigations to such attacks have to remove the entropy of inter-arrival timings between query responses.
Community Contributions
Found the code? Know the venue? Think something is wrong? Let us know!
๐ Similar Papers
In the same crypt โ Cryptography & Security
R.I.P.
๐ป
Ghosted
R.I.P.
๐ป
Ghosted
Membership Inference Attacks against Machine Learning Models
R.I.P.
๐ป
Ghosted
The Limitations of Deep Learning in Adversarial Settings
R.I.P.
๐ป
Ghosted
Practical Black-Box Attacks against Machine Learning
R.I.P.
๐ป
Ghosted
Distillation as a Defense to Adversarial Perturbations against Deep Neural Networks
R.I.P.
๐ป
Ghosted
Extracting Training Data from Large Language Models
Died the same way โ ๐ป Ghosted
R.I.P.
๐ป
Ghosted
Language Models are Few-Shot Learners
R.I.P.
๐ป
Ghosted
PyTorch: An Imperative Style, High-Performance Deep Learning Library
R.I.P.
๐ป
Ghosted
XGBoost: A Scalable Tree Boosting System
R.I.P.
๐ป
Ghosted