Session recognition and Bandwidth Guarantee for Encrypted Internet Voice Traffic : Case Study of Skype

Author

Wang, Jian-Hong ; Pan, J.-Y. ; Cheng, Yi-Chi

Author_Institution

Dept. of Commun. Eng., Nat. Chung Cheng Univ.

fYear

2007

fDate

March 1 2007-April 5 2007

Firstpage

384

Lastpage

389

Abstract

Under limited bandwidth, how to improve quality of service in VoIP (Voice over Internet Protocol) is a significant problem of the network traffic today. However, some protocol of VoIP traffic such as Skype is encrypted; with this problem, the network administrator can´t manage the VoIP bandwidth and result in time delay or loss of voice packets. The purpose of this study is to improve VoIP bandwidth guarantee by identifying encrypted Skype session. First, we use the Pcap (packet capture library) to capture session characteristics from the information about established connections in the network. Next, K-means is used to do clustering analysis and pattern recognition. After then, we can recognize the Skype session. Finally, we can feed the information of encrypted Skype session which has been identified into the bandwidth management system. So that, we can improve quality of bandwidth guarantee in Skype.

Keywords

Internet telephony; bandwidth allocation; cryptography; pattern clustering; quality of service; telecommunication traffic; K-means method; Skype session identification; VoIP bandwidth guarantee; VoIP traffic; Voice over Internet Protocol; bandwidth management system; clustering analysis; encrypted Internet voice traffic; network traffic; packet capture library; pattern recognition; quality of service; session recognition; Bandwidth; Cryptography; Delay effects; Internet telephony; Libraries; Pattern analysis; Protocols; Quality of service; Speech recognition; Telecommunication traffic; Encrypted VoIP; K-means; Skype session identification; pattern recognition;

fLanguage

English

Publisher

ieee

Conference_Titel

Computational Intelligence and Data Mining, 2007. CIDM 2007. IEEE Symposium on

Conference_Location

Honolulu, HI

Print_ISBN

1-4244-0705-2

Type

conf

DOI

10.1109/CIDM.2007.368900

Filename

4221324