Document Type

Dissertation

Rights

This item is available under a Creative Commons License for non-commercial use only

Disciplines

1.2 COMPUTER AND INFORMATION SCIENCE, Computer Sciences, Information Science

Abstract

DNS is one of the most widely used protocols on the internet and is used in the translation of domain names into IP address in order to correctly route messages between computers. It presents an attractive attack vector for criminals as the service is not as closely monitored by security experts as other protocols such as HTTP or FTP. Its use as a covert means of communication has increased with the availability of tools that allow for the creation of DNS tunnels using the protocol. One of the primary motivations for using DNS tunnels is the illegal extraction of information from a company’s network. This can lead to reputational damage for the organisation and result in significant fines – particularly with the introduction of General Data Protection Regulations in the EU. Most of the research into the detection of DNS tunnels has used anomalies in the relationship between DNS requests and other protocols, or anomalies in the rate of DNS requests made over specific time periods. This study will look at the characteristics of an individual DNS requests to see how effective different classification techniques are at identifying tunnels. The different techniques selected are Logistic Regression (LR), Decision Tree (DT), Random Forest (RF), and Support Vector Machine (SVM). The effectiveness of the different techniques will be measured and compared to see if there are statistically significant differences between them using a Cochran’s Q test. The results will indicate that DT, RF and SVM, are the most effective techniques at categorising DNS requests, and that they are significantly different to the other models. Key Words: DNS Tunnel, Logistic Regression, Support Vector Machine, Decision Tree, Random Forest, Cochran’s Q Test.

DOI

https://doi.org/10.21427/D7SC04


Share

COinS