Efficient semi-supervised learning bittorrent traffic detection - an extended summary

Authors:
Raymond Siulai Wong;Teng-Sheng Moh;Melody Moh
Affiliations:
Computer Science Department, San Jose State University, San Jose, CA;Computer Science Department, San Jose State University, San Jose, CA;Computer Science Department, San Jose State University, San Jose, CA
Venue:
ICDCN'12 Proceedings of the 13th international conference on Distributed Computing and Networking
Year:
2012

Citing 4
Cited 0

Introduction to Data Mining, (First Edition)

Introduction to Data Mining, (First Edition)
Offline/realtime traffic classification using semi-supervised learning

Performance Evaluation
P2P Applications Identification Based on the Statistics Analysis of Packet Length

IEEC '09 Proceedings of the 2009 International Symposium on Information Engineering and Electronic Commerce
P2P Traffic Identification Based on Double Layer Characteristics

ITCS '09 Proceedings of the 2009 International Conference on Information Technology and Computer Science - Volume 02

Quantified Score

Hi-index	0.00

Visualization

Abstract

The peer-to-peer (P2P) technology has been well developed over the Internet. BitTorrent (BT) is one of the most popular P2P sharing protocols. BT network traffic detection has become increasingly important since ISP and enterprise networks often want to detect and limit P2P traffic for other critical applications. We propose a new detection method that is based on an intelligent combination of Deep Packet Inspection (DPI) and Deep Flow Inspection (DFI) with semi-supervised learning. Comparing with existing methods, the new method has achieved equally high accuracy with shorter classification time.