Presentation 2015-09-01
Evaluation of Router Power Reduction By Processing Speed Control Utilizing Support Vector Machines
Hiroshi Kawase, Yojiro Mori, Hiroshi Hasegawa, Ken-ichi Sato,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this paper, we propose a novel router performance control method for realizing power consumption reduction. The proposed method utilizes a machine learning technique, Support Vector Machine, which utilizes traffic normalization with a corresponding performance threshold that enables unified control. This formulation achieves fast adaptability to changing traffic characteristics. Numerical experiments using several real Internet traffic data sets elucidate that the energy efficiency attained by the proposed method is improved from that by the previous method based on frequency decomposition. Moreover, we evaluate the effect of several key parameters such as traffic measurement interval and the number of packet processing engines, and clarify the proper parameter values that can yield significant power reduction while preventing packet loss.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) router / dynamic performance control / power saving / machine learning
Paper # PN2015-16
Date of Issue 2015-08-24 (PN)

Conference Information
Committee PN
Conference Date 2015/8/31(2days)
Place (in Japanese) (See Japanese page)
Place (in English) Asahidake Manseikaku Hotel Bearmonte
Topics (in Japanese) (See Japanese page)
Topics (in English) Photonic network technologies
Chair Eiji Oki(Univ. of Electro-Comm.)
Vice Chair Hiroshi Hasegawa(Nagoya Univ.) / Takehiro Tsuritani(KDDI Labs.) / Haruki Ogoshi(Furukawa Electric)
Secretary Hiroshi Hasegawa(NICT) / Takehiro Tsuritani(Osaka Univ.) / Haruki Ogoshi
Assistant Masahiro Nakagawa(NTT)

Paper Information
Registration To Technical Committee on Photonic Network
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Evaluation of Router Power Reduction By Processing Speed Control Utilizing Support Vector Machines
Sub Title (in English)
Keyword(1) router
Keyword(2) dynamic performance control
Keyword(3) power saving
Keyword(4) machine learning
1st Author's Name Hiroshi Kawase
1st Author's Affiliation Nagoya University(Nagoya Univ.)
2nd Author's Name Yojiro Mori
2nd Author's Affiliation Nagoya University(Nagoya Univ.)
3rd Author's Name Hiroshi Hasegawa
3rd Author's Affiliation Nagoya University(Nagoya Univ.)
4th Author's Name Ken-ichi Sato
4th Author's Affiliation Nagoya University(Nagoya Univ.)
Date 2015-09-01
Paper # PN2015-16
Volume (vol) vol.115
Number (no) PN-204
Page pp.pp.31-35(PN),
#Pages 5
Date of Issue 2015-08-24 (PN)