A Preliminary Investigation of Skype Traffic Classification Using a Minimalist Feature Set

Author

Angevine, Duffy ; Zincir-Heywood, A. Nur

Author_Institution

Dalhousie Univ., Halifax, NS

fYear

2008

Firstpage

1075

Lastpage

1079

Abstract

In this work, AdaBoost and C4.5, are employed for classifying Skype direct (UDP and TCP) communications from traffic log files. Pre-processing is applied to the traffic data to express it as flows, which is later converted into a descriptive feature set. The aforementioned algorithms are then evaluated on this feature set. Results show that a 98% detection rate with 6% false positive rate for UDP based Skype and a 94% detection rate with 4% false positive rate for TCP based Skype is possible to achieve.

Keywords

Internet telephony; learning (artificial intelligence); peer-to-peer computing; telecommunication computing; telecommunication traffic; transport protocols; AdaBoost algorithm; C4.5 algorithm; Skype traffic classification; TCP communication; UDP communication; machine learning algorithm; minimalist feature set; peer-to-peer VoIP network; traffic log file; Availability; Cryptography; Hidden Markov models; Machine learning algorithms; Payloads; Privacy; Protocols; Telecommunication traffic; Traffic control; Tunneling; encrypted; skype; traffic classification;

fLanguage

English

Publisher

ieee

Conference_Titel

Availability, Reliability and Security, 2008. ARES 08. Third International Conference on

Conference_Location

Barcelona

Print_ISBN

978-0-7695-3102-1

Type

conf

DOI

10.1109/ARES.2008.158

Filename

4529463