An Attribution Method for Alerts in an Educational Cyber Range based on Graph Database

Abstract: Finding the source of events in a network is a critical problem in network security, and this process is called network attribution. This thesis develops a system to attribute alerts from the cyber range of the Ethical Hacking Course at KTH to students. The cyber range is an essential component of the Ethical Hacking course. It provides a platform for students to practice their hacking knowledge and skills while recording their actions for research or course purposes. To use the alerts generated by the cyber range to study student behavior, it is necessary to find which student triggered the alert. In this thesis, the system uses a method based on a graph database to attribute alerts to students. The system designed in this thesis use the log data recorded by the cyber range to create nodes and relationships and use the data related to the traffic between hosts to attribute the traffic. After the attribution is complete, the system uses the attributed student traffic to reconstruct the path from the student to the host that captured the alert. If the path from the student to the host is successfully constructed, the attribution of the alert is considered successful. In the end, the system was able to attribute 94% of the student traffic. Using the student traffic that is successfully attributed, the system can attribute 79.75% of the bad-samba alerts to students and build the path from the student to the host. The system designed in this thesis is helpful for understanding and managing the students’ hacking behavior in the cyber range of the Ethical Hacking course.