Network Traffic Classification by Common Subsequence Finding

Abstract

The paper describes issues related to network traffic analysis. The scope of this article includes discussion regarding the problem of network traffic identification and classification. Furthermore, paper presents two bioinformatics methods: Clustal and Center Star. Both methods were precisely adapted to the network security purpose. In both methods, the concept of extraction of a common subsequence, based on multiple sequence alignment of more than two network attack signatures, was used. This concept was inspired by bioinformatics solutions for the problems related to finding similarities in a set of DNA, RNA or amino acids sequences. Additionally, the scope of the paper includes detailed description of test procedures and their results. At the end some relevant evaluations and conclusions regarding both methods are presented.