Real-time traffic classification based on cosine similarity using sub-application vectors

Cihangir Beşiktaş; Haci Ali Mantar

doi:10.1007/978-3-642-28534-9_10

Real-time traffic classification based on cosine similarity using sub-application vectors

Cihangir Beşiktaş^*, Haci Ali Mantar

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

4 Citations (Scopus)

Abstract

Internet traffic classification has a critical role on network monitoring, quality of service, intrusion detection, network security and trend analysis. The conventional port-based method is ineffective due to dynamic port usage and masquerading techniques. Besides, payload-based method suffers from heavy load and encryption. Due to these facts, machine learning based statistical approaches have become the new trend for the network measurement community. In this short paper, we propose a new statistical approach based on DBSCAN clustering and weighted cosine similarity. Our experimental test results show that the proposed approach achieves very high accuracy.

Original language	English
Title of host publication	Traffic Monitoring and Analysis - 4th International Workshop, TMA 2012, Proceedings
Pages	89-92
Number of pages	4
DOIs	https://doi.org/10.1007/978-3-642-28534-9_10
Publication status	Published - 2012
Externally published	Yes
Event	4th International Workshop on Traffic Monitoring and Analysis, TMA 2012 - Vienna, Austria Duration: 12 Mar 2012 → 12 Mar 2012

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	7189 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	4th International Workshop on Traffic Monitoring and Analysis, TMA 2012
Country/Territory	Austria
City	Vienna
Period	12/03/12 → 12/03/12

Funding

This work has been supported by Inforcept Networks corporation.

Funders	Funder number
Inforcept Networks corporation

Keywords

cosine similarity
DBSCAN
packet inspection
Traffic classification

Access to Document

10.1007/978-3-642-28534-9_10

Cite this

Beşiktaş, C., & Mantar, H. A. (2012). Real-time traffic classification based on cosine similarity using sub-application vectors. In Traffic Monitoring and Analysis - 4th International Workshop, TMA 2012, Proceedings (pp. 89-92). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7189 LNCS). https://doi.org/10.1007/978-3-642-28534-9_10

Beşiktaş, Cihangir ; Mantar, Haci Ali. / Real-time traffic classification based on cosine similarity using sub-application vectors. Traffic Monitoring and Analysis - 4th International Workshop, TMA 2012, Proceedings. 2012. pp. 89-92 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{99833d529826436b94a30418e41852b4,

title = "Real-time traffic classification based on cosine similarity using sub-application vectors",

abstract = "Internet traffic classification has a critical role on network monitoring, quality of service, intrusion detection, network security and trend analysis. The conventional port-based method is ineffective due to dynamic port usage and masquerading techniques. Besides, payload-based method suffers from heavy load and encryption. Due to these facts, machine learning based statistical approaches have become the new trend for the network measurement community. In this short paper, we propose a new statistical approach based on DBSCAN clustering and weighted cosine similarity. Our experimental test results show that the proposed approach achieves very high accuracy.",

keywords = "cosine similarity, DBSCAN, packet inspection, Traffic classification",

author = "Cihangir Be{\c s}ikta{\c s} and Mantar, {Haci Ali}",

year = "2012",

doi = "10.1007/978-3-642-28534-9_10",

language = "English",

isbn = "9783642285332",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

pages = "89--92",

booktitle = "Traffic Monitoring and Analysis - 4th International Workshop, TMA 2012, Proceedings",

note = "4th International Workshop on Traffic Monitoring and Analysis, TMA 2012 ; Conference date: 12-03-2012 Through 12-03-2012",

}

Beşiktaş, C & Mantar, HA 2012, Real-time traffic classification based on cosine similarity using sub-application vectors. in Traffic Monitoring and Analysis - 4th International Workshop, TMA 2012, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7189 LNCS, pp. 89-92, 4th International Workshop on Traffic Monitoring and Analysis, TMA 2012, Vienna, Austria, 12/03/12. https://doi.org/10.1007/978-3-642-28534-9_10

Real-time traffic classification based on cosine similarity using sub-application vectors. / Beşiktaş, Cihangir; Mantar, Haci Ali.
Traffic Monitoring and Analysis - 4th International Workshop, TMA 2012, Proceedings. 2012. p. 89-92 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7189 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Real-time traffic classification based on cosine similarity using sub-application vectors

AU - Beşiktaş, Cihangir

AU - Mantar, Haci Ali

PY - 2012

Y1 - 2012

N2 - Internet traffic classification has a critical role on network monitoring, quality of service, intrusion detection, network security and trend analysis. The conventional port-based method is ineffective due to dynamic port usage and masquerading techniques. Besides, payload-based method suffers from heavy load and encryption. Due to these facts, machine learning based statistical approaches have become the new trend for the network measurement community. In this short paper, we propose a new statistical approach based on DBSCAN clustering and weighted cosine similarity. Our experimental test results show that the proposed approach achieves very high accuracy.

AB - Internet traffic classification has a critical role on network monitoring, quality of service, intrusion detection, network security and trend analysis. The conventional port-based method is ineffective due to dynamic port usage and masquerading techniques. Besides, payload-based method suffers from heavy load and encryption. Due to these facts, machine learning based statistical approaches have become the new trend for the network measurement community. In this short paper, we propose a new statistical approach based on DBSCAN clustering and weighted cosine similarity. Our experimental test results show that the proposed approach achieves very high accuracy.

KW - cosine similarity

KW - DBSCAN

KW - packet inspection

KW - Traffic classification

UR - http://www.scopus.com/inward/record.url?scp=84858389149&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-28534-9_10

DO - 10.1007/978-3-642-28534-9_10

M3 - Conference contribution

AN - SCOPUS:84858389149

SN - 9783642285332

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 89

EP - 92

BT - Traffic Monitoring and Analysis - 4th International Workshop, TMA 2012, Proceedings

T2 - 4th International Workshop on Traffic Monitoring and Analysis, TMA 2012

Y2 - 12 March 2012 through 12 March 2012

ER -

Beşiktaş C, Mantar HA. Real-time traffic classification based on cosine similarity using sub-application vectors. In Traffic Monitoring and Analysis - 4th International Workshop, TMA 2012, Proceedings. 2012. p. 89-92. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-642-28534-9_10

Real-time traffic classification based on cosine similarity using sub-application vectors

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this