University of Twente Student Theses


Detecting botnets using file system indicators

Wagenaar, P. (2012) Detecting botnets using file system indicators.

[img] PDF
Abstract:Botnets, large groups of networked zombie computers under centralised control, are recognised as one of the major threats on the internet. There is a lot of research towards ways of detecting botnets, in particular towards detecting Command and Control servers. Most of the research is focused on trying to detect the commands that these servers send to the bots over the network. For this research, we have looked at botnets from a botmaster's perspective. First, we characterise several botnet enhancing techniques using three aspects: resilience, stealth and churn. We see that these enhancements are usually employed in the network communications between the C&C and the bots. This leads us to our second contribution: we propose a new botnet detection method based on the way C&C's are present on the file system. We define a set of file system based indicators and use them to search for C&C's in images of hard disks. We investigate how the aspects resilience, stealth and churn apply to each of the indicators and discuss countermeasures botmasters could take to evade detection. We validate our method by applying it to a test dataset of 94 disk images, 16 of which contain C&C installations, and show that low false positive and false negative ratio's can be achieved. Approaching the botnet detection problem from this angle is novel, which provides a basis for further research.
Item Type:Essay (Master)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science
Programme:Computer Science MSc (60300)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page