A new traffic flow behavior analysis method has been proposed to detect botnets irrespective of their control structures.

Benchmark datasets collected from various sources, such as ISOT Botnet dataset from University of Victoria, Conficker dataset from CAIDA, dataset from CVUT University, dataset from Dalhousie University and dataset from Centro University.

It can successfully detect the various types of botnets with a high detection rate and a low false positive rate.