Using machine learning to provide differentiated services in SDN-like publish/subscribe systems for  IoT

Yulong Shi; Yang Zhang; Hans Arno Jacobsen; Bo Han; Mengxi Wei; Runyuan Li; Junliang Chen

doi:10.1007/978-3-030-03596-9_39

Using machine learning to provide differentiated services in SDN-like publish/subscribe systems for IoT

Yulong Shi
, Yang Zhang
, Hans Arno Jacobsen
, Bo Han
, Mengxi Wei
, Runyuan Li
, Junliang Chen

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

2 Scopus citations

Abstract

At present, most publish/subscribe systems assume that all participants have the same Quality of Service (QoS) requirements. However, in many real-world IoT service scenarios, different users may have different delay requirements. How to provide differentiated services has become an urgent problem. The rise of Software Defined Networking (SDN) provides endless possibilities for meeting customized services due to greater programmability. In this paper, we first propose two new methods to predict the queuing delay of switches. One is an improvement of the traditional Random Early Detection (RED) algorithm; the other is a machine learning method using the eXtreme Gradient Boosting (XGBoost) model. Then we describe an SDN-like publish/subscribe system architecture and priority queues supported by OpenFlow switches to realize differentiated services. In order to guarantee QoS, we present a two-layer queue management mechanism based on user requirements. In the end, we compare our delay prediction methods with the RED method and verify the effectiveness of the two-layer queue management mechanism. Experimental results show that our solution is effective.

Original language	English
Title of host publication	Service-Oriented Computing - 16th International Conference, ICSOC 2018, Proceedings
Editors	Qi Yu, Claus Pahl, Maja Vukovic, Jianwei Yin
Publisher	Springer Verlag
Pages	532-540
Number of pages	9
ISBN (Print)	9783030035952
DOIs	https://doi.org/10.1007/978-3-030-03596-9_39
State	Published - 2018
Externally published	Yes
Event	16th International Conference on Service-Oriented Computing, ICSOC 2018 - Hangzhou, China Duration: 12 Nov 2018 → 15 Nov 2018

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11236 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	16th International Conference on Service-Oriented Computing, ICSOC 2018
Country/Territory	China
City	Hangzhou
Period	12/11/18 → 15/11/18

Keywords

Machine learning
Publish/Subscribe
Quality of Service
Queue management
Software Defined Networking

Access to Document

10.1007/978-3-030-03596-9_39

Cite this

Shi, Y., Zhang, Y., Jacobsen, H. A., Han, B., Wei, M., Li, R., & Chen, J. (2018). Using machine learning to provide differentiated services in SDN-like publish/subscribe systems for IoT. In Q. Yu, C. Pahl, M. Vukovic, & J. Yin (Eds.), Service-Oriented Computing - 16th International Conference, ICSOC 2018, Proceedings (pp. 532-540). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11236 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-03596-9_39

Shi, Yulong ; Zhang, Yang ; Jacobsen, Hans Arno et al. / Using machine learning to provide differentiated services in SDN-like publish/subscribe systems for IoT. Service-Oriented Computing - 16th International Conference, ICSOC 2018, Proceedings. editor / Qi Yu ; Claus Pahl ; Maja Vukovic ; Jianwei Yin. Springer Verlag, 2018. pp. 532-540 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{8fe9dca85ea34311a3a97490032151df,

title = "Using machine learning to provide differentiated services in SDN-like publish/subscribe systems for IoT",

abstract = "At present, most publish/subscribe systems assume that all participants have the same Quality of Service (QoS) requirements. However, in many real-world IoT service scenarios, different users may have different delay requirements. How to provide differentiated services has become an urgent problem. The rise of Software Defined Networking (SDN) provides endless possibilities for meeting customized services due to greater programmability. In this paper, we first propose two new methods to predict the queuing delay of switches. One is an improvement of the traditional Random Early Detection (RED) algorithm; the other is a machine learning method using the eXtreme Gradient Boosting (XGBoost) model. Then we describe an SDN-like publish/subscribe system architecture and priority queues supported by OpenFlow switches to realize differentiated services. In order to guarantee QoS, we present a two-layer queue management mechanism based on user requirements. In the end, we compare our delay prediction methods with the RED method and verify the effectiveness of the two-layer queue management mechanism. Experimental results show that our solution is effective.",

keywords = "Machine learning, Publish/Subscribe, Quality of Service, Queue management, Software Defined Networking",

author = "Yulong Shi and Yang Zhang and Jacobsen, \{Hans Arno\} and Bo Han and Mengxi Wei and Runyuan Li and Junliang Chen",

note = "Publisher Copyright: {\textcopyright} Springer Nature Switzerland AG 2018.; 16th International Conference on Service-Oriented Computing, ICSOC 2018 ; Conference date: 12-11-2018 Through 15-11-2018",

year = "2018",

doi = "10.1007/978-3-030-03596-9\_39",

language = "English",

isbn = "9783030035952",

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

publisher = "Springer Verlag",

pages = "532--540",

editor = "Qi Yu and Claus Pahl and Maja Vukovic and Jianwei Yin",

booktitle = "Service-Oriented Computing - 16th International Conference, ICSOC 2018, Proceedings",

}

Shi, Y, Zhang, Y, Jacobsen, HA, Han, B, Wei, M, Li, R & Chen, J 2018, Using machine learning to provide differentiated services in SDN-like publish/subscribe systems for IoT. in Q Yu, C Pahl, M Vukovic & J Yin (eds), Service-Oriented Computing - 16th International Conference, ICSOC 2018, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11236 LNCS, Springer Verlag, pp. 532-540, 16th International Conference on Service-Oriented Computing, ICSOC 2018, Hangzhou, China, 12/11/18. https://doi.org/10.1007/978-3-030-03596-9_39

Using machine learning to provide differentiated services in SDN-like publish/subscribe systems for IoT. / Shi, Yulong; Zhang, Yang; Jacobsen, Hans Arno et al.
Service-Oriented Computing - 16th International Conference, ICSOC 2018, Proceedings. ed. / Qi Yu; Claus Pahl; Maja Vukovic; Jianwei Yin. Springer Verlag, 2018. p. 532-540 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11236 LNCS).

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

TY - GEN

T1 - Using machine learning to provide differentiated services in SDN-like publish/subscribe systems for IoT

AU - Shi, Yulong

AU - Zhang, Yang

AU - Jacobsen, Hans Arno

AU - Han, Bo

AU - Wei, Mengxi

AU - Li, Runyuan

AU - Chen, Junliang

N1 - Publisher Copyright: © Springer Nature Switzerland AG 2018.

PY - 2018

Y1 - 2018

N2 - At present, most publish/subscribe systems assume that all participants have the same Quality of Service (QoS) requirements. However, in many real-world IoT service scenarios, different users may have different delay requirements. How to provide differentiated services has become an urgent problem. The rise of Software Defined Networking (SDN) provides endless possibilities for meeting customized services due to greater programmability. In this paper, we first propose two new methods to predict the queuing delay of switches. One is an improvement of the traditional Random Early Detection (RED) algorithm; the other is a machine learning method using the eXtreme Gradient Boosting (XGBoost) model. Then we describe an SDN-like publish/subscribe system architecture and priority queues supported by OpenFlow switches to realize differentiated services. In order to guarantee QoS, we present a two-layer queue management mechanism based on user requirements. In the end, we compare our delay prediction methods with the RED method and verify the effectiveness of the two-layer queue management mechanism. Experimental results show that our solution is effective.

AB - At present, most publish/subscribe systems assume that all participants have the same Quality of Service (QoS) requirements. However, in many real-world IoT service scenarios, different users may have different delay requirements. How to provide differentiated services has become an urgent problem. The rise of Software Defined Networking (SDN) provides endless possibilities for meeting customized services due to greater programmability. In this paper, we first propose two new methods to predict the queuing delay of switches. One is an improvement of the traditional Random Early Detection (RED) algorithm; the other is a machine learning method using the eXtreme Gradient Boosting (XGBoost) model. Then we describe an SDN-like publish/subscribe system architecture and priority queues supported by OpenFlow switches to realize differentiated services. In order to guarantee QoS, we present a two-layer queue management mechanism based on user requirements. In the end, we compare our delay prediction methods with the RED method and verify the effectiveness of the two-layer queue management mechanism. Experimental results show that our solution is effective.

KW - Machine learning

KW - Publish/Subscribe

KW - Quality of Service

KW - Queue management

KW - Software Defined Networking

UR - https://www.scopus.com/pages/publications/85056875240

U2 - 10.1007/978-3-030-03596-9_39

DO - 10.1007/978-3-030-03596-9_39

M3 - Conference contribution

AN - SCOPUS:85056875240

SN - 9783030035952

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

SP - 532

EP - 540

BT - Service-Oriented Computing - 16th International Conference, ICSOC 2018, Proceedings

A2 - Yu, Qi

A2 - Pahl, Claus

A2 - Vukovic, Maja

A2 - Yin, Jianwei

PB - Springer Verlag

T2 - 16th International Conference on Service-Oriented Computing, ICSOC 2018

Y2 - 12 November 2018 through 15 November 2018

ER -

Shi Y, Zhang Y, Jacobsen HA, Han B, Wei M, Li R et al. Using machine learning to provide differentiated services in SDN-like publish/subscribe systems for IoT. In Yu Q, Pahl C, Vukovic M, Yin J, editors, Service-Oriented Computing - 16th International Conference, ICSOC 2018, Proceedings. Springer Verlag. 2018. p. 532-540. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-03596-9_39