BoTGen: A new approach for in-lab generation of botnet datasets

Abstract

Although datasets represent a critical part of research and development activities, botnet research suffers from a serious shortage of reliable and representative datasets. In this paper, we explain a new approach to build a botnet experimentation platform completely from off-the-shelf open sources. This work aims to fill the gap in botnet research due to the lack of representative datasets. The proposed approach provides a flexible way to experiment with botnets freely in a controlled environment. Moreover, various botnet scenarios can be generated and carried out automatically, which allows producing rich datasets with diverse botnet scenarios.