A comparative study for fast-flux service networks detection

One of the most active threats we meet on the Internet is cyber-crime. Fast-flux is a kind of DNS technique used by botnets to hiding the malicious activities. In this paper we use data mining techniques to detect the fast-flux service network (FFSN) which is newly emerging and still not perceiving widely. From the data mining perspective, the detection of cyber-crime is viewed as kind of imbalanced class problem. In this paper we analysis the feature attributes which can distinguish fast-flux domains from benign ones by observing system/network performance. Then we present the solution approach and comparative study based on data mining techniques for fast-flux networks detection. The experiment results show our approach is effective and efficient.