Identifying differentiating characteristics of internet applications using Principal Component Analysis

Abstract

An accurate mapping between traffic and their corresponding applications can be very important for a wide range of network management and measurement tasks including traffic engineering, service differentiation, performance/failure monitoring, and security. Having the ability to identify the differentiating characteristics of IP applications is a crucial issue that can form the basis for the development of an efficient application identification tool. This paper presents an approach, based on principal component analysis, that is able to identify differentiating characteristics of different Internet applications, including several peer-to-peer file sharing protocols. These applications have gained tremendous popularity in the past few years and, today, traffic generated by P2P systems accounts for a major fraction of the Internet traffic and is bound to increase.