Methods of improvement effectveness for high-speed packet classifying

Abstract

The article describes methods and approaches chosen for development of network traffic classifiers. These tools usable as detectors of abnormal activities are based on Aho-Corasick algorithm implementation. The main part of the work is devoted to the overview of ways to improve the classifiers efficiency as well as to minimization of processing time needed to detect traffic class the particular network packet belongs to. Obtained conclusions can be used to create distributed classification system with optimal architecture.