Using internet traffic self-similarity for detection of network anomalies

The goal of this paper is to prove by simulations in an virtual environment the potential of fractal analysis techniques in evaluation of network characteristics, especially in detection of anomalies. In this aim there are proposed several original research results: an Internet model based on scale free network graphs, a method to reveal self-similarities in generated traffic, a method to compare the behavior of two sub-networks and a powerful experimental test-bed which allows parallel processing running pdns (parallel/distributed network simulator) package on different sub-networks having scale free topology.