TRACING OF VOIP TRAFFIC IN THE RAPID FLOW INTERNET BACKBONE

A.Jenefa, Blessy Selvam

Abstract: VoIP traffic application gaining a terrific admiration in the recent couple of years. VoIP Traffic Classification has concerned for network management and it comes to be more complicating because of modern applications behaviors and it has attracted the research community to develop and propose various classification techniques which don’t depend on ‘well known UDP or TCP port numbers. To overcome the problem of unknown flow classification and achieve effective network classification, a new innovative novel work called Multi Stage Fine-Grained classifier is proposed in this research for classifying the VoIP traffic flow with high accurate classification. The datasets of VoIP network traffic measurements taken from our campus WI-FI and the experimental results shows that the proposed work outstrips the existing approaches in the Rapid flow Internet Backbone. Without investigate the packet payloads, our proposed Fine-Grained classifier effectively classifies the Peer-to-Peer encrypted traffic in the real time network. Our experimental results shows high accuracy and small error rate in classifying the Peer-to-Peer network traffic

Keywords: Multi Stage Fine-Grained Classifier, Rapid VoIP traffic Flow (SKYPE, VoIP, GAMING, Other) classification, Machine Learning

DOI: https://doi.org/10.15623/ijret.2015.0403051