Statistical Texture Analysis Methods for Network Traffic Classification

T. Kisner, A. Essoh, and F. Kaderali (Germany)


Traffic Engineering, Network Traffic Classification, Intru sion Detection, Anomaly Detection


Traffic modeling and classification is important in many areas like intrusion detection, anomalous traffic detection, network planning or bandwidth management. A novelty in the work presented in this paper is the use of texture clas sification methods from the domain of digital image pro cessing for network traffic classification. We use strate gies based on Co-occurence Matrices to derive statistical properties for network traffic classification. Using the well known kNN-Classificator we are able to distinguish differ ent classes with a high probability.

Important Links:

Go Back