A Novel Method of Filtering Internet Background Radiation Traffic

Internet Background Radiation (IBR) traffic is a kind of abnormal traffic, which is ubiquitous on the Internet. Detecting or filtering out IBR traffic from all traffic is benefit to ensure network security. This paper proposes a novel IBR traffic filtering method based on relative uncertainty theory. It directly filters out IBR traffic from collected Net Flow data without specific configurations or building extra collection system. The key assumption of our approach is that the traffic source of normal users is relatively certain rather than random. The IBR traffic filtered by our approach also includes the part that sent to normal user hosts besides the traffic to unsigned IP, which is more valuable and practical to network security.