Flow traffic classification with support vector machine by using payload length

Masayoshi Kohara, Yoshiaki Hori, Kouichi Sakurai, Heejo Lee, Jae Cheol Ryou

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

Classifying traffic is an important task for effective network planning and design, and monitoring the trends of the applications in operational networks. In this paper, we propose flow traffic classification methods using support vector machine. Classifying traffic is an important task for effective network planning and design, and monitoring the trends of the applications in operational networks. The proposals satisfy the following three requirements. Using to only flow information, not using port numbers, automatic making of traffic models. In this paper, we provide an empirical evaluation of our proposals using datasets of MIT Lincoln Laboratory, which illustrates that our proposals can classify network traffic flow over 90 % precision.

Original languageEnglish
Title of host publicationProceedings of the 2009 2nd International Conference on Computer Science and Its Applications, CSA 2009
DOIs
Publication statusPublished - 2009 Dec 1
Event2009 2nd International Conference on Computer Science and Its Applications, CSA 2009 - Jeju Island, Korea, Republic of
Duration: 2009 Dec 102009 Dec 12

Other

Other2009 2nd International Conference on Computer Science and Its Applications, CSA 2009
CountryKorea, Republic of
CityJeju Island
Period09/12/1009/12/12

Fingerprint

Support vector machines
Planning
Monitoring

Keywords

  • Machine learning
  • Network
  • Traffic analysis

ASJC Scopus subject areas

  • Computational Theory and Mathematics
  • Computer Science Applications

Cite this

Kohara, M., Hori, Y., Sakurai, K., Lee, H., & Ryou, J. C. (2009). Flow traffic classification with support vector machine by using payload length. In Proceedings of the 2009 2nd International Conference on Computer Science and Its Applications, CSA 2009 [5404180] https://doi.org/10.1109/CSA.2009.5404180

Flow traffic classification with support vector machine by using payload length. / Kohara, Masayoshi; Hori, Yoshiaki; Sakurai, Kouichi; Lee, Heejo; Ryou, Jae Cheol.

Proceedings of the 2009 2nd International Conference on Computer Science and Its Applications, CSA 2009. 2009. 5404180.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Kohara, M, Hori, Y, Sakurai, K, Lee, H & Ryou, JC 2009, Flow traffic classification with support vector machine by using payload length. in Proceedings of the 2009 2nd International Conference on Computer Science and Its Applications, CSA 2009., 5404180, 2009 2nd International Conference on Computer Science and Its Applications, CSA 2009, Jeju Island, Korea, Republic of, 09/12/10. https://doi.org/10.1109/CSA.2009.5404180
Kohara M, Hori Y, Sakurai K, Lee H, Ryou JC. Flow traffic classification with support vector machine by using payload length. In Proceedings of the 2009 2nd International Conference on Computer Science and Its Applications, CSA 2009. 2009. 5404180 https://doi.org/10.1109/CSA.2009.5404180
Kohara, Masayoshi ; Hori, Yoshiaki ; Sakurai, Kouichi ; Lee, Heejo ; Ryou, Jae Cheol. / Flow traffic classification with support vector machine by using payload length. Proceedings of the 2009 2nd International Conference on Computer Science and Its Applications, CSA 2009. 2009.
@inproceedings{f6989ec921704e5bba87ca0768ce4f7d,
title = "Flow traffic classification with support vector machine by using payload length",
abstract = "Classifying traffic is an important task for effective network planning and design, and monitoring the trends of the applications in operational networks. In this paper, we propose flow traffic classification methods using support vector machine. Classifying traffic is an important task for effective network planning and design, and monitoring the trends of the applications in operational networks. The proposals satisfy the following three requirements. Using to only flow information, not using port numbers, automatic making of traffic models. In this paper, we provide an empirical evaluation of our proposals using datasets of MIT Lincoln Laboratory, which illustrates that our proposals can classify network traffic flow over 90 {\%} precision.",
keywords = "Machine learning, Network, Traffic analysis",
author = "Masayoshi Kohara and Yoshiaki Hori and Kouichi Sakurai and Heejo Lee and Ryou, {Jae Cheol}",
year = "2009",
month = "12",
day = "1",
doi = "10.1109/CSA.2009.5404180",
language = "English",
isbn = "9781424449460",
booktitle = "Proceedings of the 2009 2nd International Conference on Computer Science and Its Applications, CSA 2009",

}

TY - GEN

T1 - Flow traffic classification with support vector machine by using payload length

AU - Kohara, Masayoshi

AU - Hori, Yoshiaki

AU - Sakurai, Kouichi

AU - Lee, Heejo

AU - Ryou, Jae Cheol

PY - 2009/12/1

Y1 - 2009/12/1

N2 - Classifying traffic is an important task for effective network planning and design, and monitoring the trends of the applications in operational networks. In this paper, we propose flow traffic classification methods using support vector machine. Classifying traffic is an important task for effective network planning and design, and monitoring the trends of the applications in operational networks. The proposals satisfy the following three requirements. Using to only flow information, not using port numbers, automatic making of traffic models. In this paper, we provide an empirical evaluation of our proposals using datasets of MIT Lincoln Laboratory, which illustrates that our proposals can classify network traffic flow over 90 % precision.

AB - Classifying traffic is an important task for effective network planning and design, and monitoring the trends of the applications in operational networks. In this paper, we propose flow traffic classification methods using support vector machine. Classifying traffic is an important task for effective network planning and design, and monitoring the trends of the applications in operational networks. The proposals satisfy the following three requirements. Using to only flow information, not using port numbers, automatic making of traffic models. In this paper, we provide an empirical evaluation of our proposals using datasets of MIT Lincoln Laboratory, which illustrates that our proposals can classify network traffic flow over 90 % precision.

KW - Machine learning

KW - Network

KW - Traffic analysis

UR - http://www.scopus.com/inward/record.url?scp=80655134693&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80655134693&partnerID=8YFLogxK

U2 - 10.1109/CSA.2009.5404180

DO - 10.1109/CSA.2009.5404180

M3 - Conference contribution

AN - SCOPUS:80655134693

SN - 9781424449460

BT - Proceedings of the 2009 2nd International Conference on Computer Science and Its Applications, CSA 2009

ER -