The two greatest weaknesses of Intrusion Detection Systems (IDS) are the ease of which they may be evaded and their tendency to generate vast amounts of false alarms. Sophisticated attackers are able to easily avoid detection, maintaining a low profile by spreading out the attack both in time and (network) space. Meanwhile alerts are generated by normal user activity. IDS have not yet reached a level where they can reliably detect and assess advanced attacks while being able to separate normal user activities.
This presentation discusses the use of Information Warfare theory, combined with multiple target tracking algorithms to generate a higher level of knowledge from current IDS. Instead of looking at IDS as the final stage in attack determination, it becomes the first stage. The IDS are treated as sensors on our network gathering information that is fed into a data fusion engine. By gathering information from different types of IDS and other sensors distributed throughout one or more networks, we aim to generate a higher level of knowledge, a situational awareness, that paints a much clearer picture of the activity on out networks.
By combining and fusing data gathered from many independent networks, it is possible to move away from the traditional defensive posture of network security. In its place we are given more of bird's eye view of the scene, and are able to see the activity of individual attackers spread out across many networks.
This presentation is based on research being conducted at the Institute for Security Technology Studies (ISTS), a federally funded research institute housed at Dartmouth College. A demonstration of the data fusion / target tracking system will be provided during the presentation.