Presentation | 2017-11-16 Virtual Network Reconfiguration Based on Bayesian Attractor Model with Linear Regression Toshihiko Ohba, Shin'ichi Arakawa, Masayuki Murata, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | A typical approach for configuring a virtual network (VN) over an optical network is to design an optimal VN with a knowledge of the end-to-end traffic demand matrix. However, it is difficult to configure the optimal VN using the traffic demand matrix in a changing environment. We have previously proposed a bayesian-approach for VN reconfiguration without using the traffic demand matrix. The approach memorizes a set of ``good" VNs, each of which works well for a certain traffic situation, and identify the current traffic situation using Bayesian inference, and then retrieve one of the VNs suitable for the current traffic situation. We use the amounts of outgoing/incoming traffic at edge routers as the traffic situation. However, this approach has difficulty in dealing with the case when the identification of the traffic situation fails. In this paper, we develop a VN reconfiguration method to deal with this case. In our method, the current traffic situation is fitted by linear regression, and then our method designs a VN using the obtained regression coefficients. Evaluation results show that our method can design a VN suitable for the current traffic demand when the identification of the traffic situation fails. We also discuss how to select and update the set of pre-specified traffic situations, and found that it is effective to select a set of traffic situations to have linear independence. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Virtual Network Reconfiguration / Bayesian Attractor Model / Linear Regression |
Paper # | PN2017-37 |
Date of Issue | 2017-11-08 (PN) |
Conference Information | |
Committee | PN |
---|---|
Conference Date | 2017/11/15(2days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | Kogakuin Univ. |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | Elastic Optical Networks, Flexible Networks, Optical Network Control/Protocol, Transport SDN, IP Backbone, SDM, Mode Division Multiplexing, Photonic Network Devices, JPN Model, EXAT, etc. |
Chair | Hiroshi Hasegawa(Nagoya Univ.) |
Vice Chair | Haruki Ogoshi(Furukawa Electric) / Takehiro Tsuritani(KDDI Research) / Hideaki Furukawa(NICT) |
Secretary | Haruki Ogoshi(NICT) / Takehiro Tsuritani(Mitsubishi Electric) / Hideaki Furukawa |
Assistant | Masahiro Nakagawa(NTT) / Keijiro Suzuki(AIST) |
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) | Virtual Network Reconfiguration Based on Bayesian Attractor Model with Linear Regression |
Sub Title (in English) | |
Keyword(1) | Virtual Network Reconfiguration |
Keyword(2) | Bayesian Attractor Model |
Keyword(3) | Linear Regression |
1st Author's Name | Toshihiko Ohba |
1st Author's Affiliation | Osaka University(Osaka Univ.) |
2nd Author's Name | Shin'ichi Arakawa |
2nd Author's Affiliation | Osaka University(Osaka Univ.) |
3rd Author's Name | Masayuki Murata |
3rd Author's Affiliation | Osaka University(Osaka Univ.) |
Date | 2017-11-16 |
Paper # | PN2017-37 |
Volume (vol) | vol.117 |
Number (no) | PN-298 |
Page | pp.pp.57-63(PN), |
#Pages | 7 |
Date of Issue | 2017-11-08 (PN) |