Presentation 2011-04-21
Low-cost Traffic Classification Method for Large-scale ISP
Takayuki GOTO, Chikara SASAKI, Atsuo TACHIBANA, Shigehiro ANO,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Quantifying the traffic volume of each protocol is useful for effective traffic engineering and network designing. A straight forward approach for protocol identification is to first capture full traffic and apply DPI (Deep Packet Inspection) technology. However, it is difficult to deploy the dedicated DPI hardware to large ISP networks because the hardware is expensive and should be deployed to multiple measurement points. In this paper, we propose a cost-effective traffic classification method which is composed from packet sampling and flow features-based traffic classification which uses supervised learning. We verify the effectiveness of our proposed method in an experiment using traffic traces that are publicly available.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Packet sampling / Traffic classification / Deep Packet Inspection
Paper # CQ2011-1
Date of Issue

Conference Information
Committee CQ
Conference Date 2011/4/14(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Communication Quality (CQ)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Low-cost Traffic Classification Method for Large-scale ISP
Sub Title (in English)
Keyword(1) Packet sampling
Keyword(2) Traffic classification
Keyword(3) Deep Packet Inspection
1st Author's Name Takayuki GOTO
1st Author's Affiliation KDDI R&D Labs.()
2nd Author's Name Chikara SASAKI
2nd Author's Affiliation KDDI R&D Labs.
3rd Author's Name Atsuo TACHIBANA
3rd Author's Affiliation KDDI R&D Labs.
4th Author's Name Shigehiro ANO
4th Author's Affiliation KDDI R&D Labs.
Date 2011-04-21
Paper # CQ2011-1
Volume (vol) vol.111
Number (no) 11
Page pp.pp.-
#Pages 4
Date of Issue