Detecting Malicious DNS Queries Over Encrypted Tunnels Using Statistical Analysis and Bi-Directional Recurrent Neural Networks
Bibliographic record
Abstract
The exponential rise in the number of malicious threats targeting computer networks and digital services puts network infrastructure in jeopardy. Domain name protocol attacks are one of the most pervasive network attacks posing a threat to networks, whereby attackers send harmful information to the network; this type of threat is identified as DNS tunneling. The DNS protocol has recently gained increased attention from cyber-attackers, targeting organizations with a web presence or reliance on e-commerce businesses. Cyber-attackers can subtly exploit the contents of encrypted DNS packets that are sent across covert network tunnels, which are difficult for firewalls and blacklist detection methods to detect. Therefore, efficient methods for detecting DNS intrusions in the network are required. Machine learning (ML), deep learning (DL), and computational intelligence models have proved to be increasingly effective in dealing with these cyber-attacks, especially when using an appropriate dataset. This paper proposes an intrusion detection model to detect malicious DNS over HTTPS (DoH) queries among network covert tunnels, using statistical analysis and Bi-directional Recurrent Neural Network (BRNN) techniques, based on the flow level of the network traffic. The proposed approach was tested and evaluated based on a realistic dataset called CIRA-CIC-DoHBrw-2020, provided by the Canadian Institute for Cybersecurity. Experiments have shown that the robustness of the model is strong, with a detection rate of 100%. Furthermore, the proposed model achieved high performance in terms of the accuracy rate in detecting malicious DoH queries, with low false-negative and false-positive rates. Furthermore, the number of features used is fewer than other approaches, making it perform faster in the training and testing phases.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".