Presentation 2008-01-25
Congestion Control of TCP Using Neural Network Model
Akihiro SHIOZU, Koki ABE,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We proposed TCP Salt, a congestion control scheme which predicts congestions by machine learning to obtain an optimum window size. It was found that the proposed method can utilize the bandwidth better than conventional methods. However, evaluation of the fairness with other connections is not sufficient. In this paper, we describe simulation results that TCP Salt can realize good fairness with other connections not only in the boost state but also in the congestion avoidance state. We also describe that using the previous value of congestion window as an input to machine learning improves the fairness. We further show that it is reasonable to use the time when the variation of RTT is zero within measurement accuracy as the tradition condition from the congestion avoidance state to the boost state.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) TCP / Congestion Control / Neural Network
Paper # NS2007-126
Date of Issue

Conference Information
Committee NS
Conference Date 2008/1/17(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 Network Systems(NS)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Congestion Control of TCP Using Neural Network Model
Sub Title (in English)
Keyword(1) TCP
Keyword(2) Congestion Control
Keyword(3) Neural Network
1st Author's Name Akihiro SHIOZU
1st Author's Affiliation Department of Computer Science, The University of Electro-Communications()
2nd Author's Name Koki ABE
2nd Author's Affiliation Department of Computer Science, The University of Electro-Communications
Date 2008-01-25
Paper # NS2007-126
Volume (vol) vol.107
Number (no) 443
Page pp.pp.-
#Pages 6
Date of Issue