TY - JOUR
T1 - Efficient OpenFlow based Inbound Load Balancing for Enterprise Networks
AU - Wang, Guodong
AU - Liu, Peng
AU - Zhao, Yanxiao
AU - Li, Jun
AU - Song, Min
N1 - Publisher Copyright:
© 2018 Elsevier Ltd. All rights reserved.
PY - 2018
Y1 - 2018
N2 - Inbound load balancing plays an important role in improving the quality of experience for enterprise networks. Traditional traffic engineering approaches fail to achieve fine-grained inbound load balancing because they have limited capability of obtaining detailed network status and lack flexibility for traffic flow scheduling. In this paper, we adopt OpenFlow to achieve effective inbound load balancing. The potential of implementing OpenFlow in an enterprise network is explored to schedule traffic among different inbound links. We also investigate how to avoid frequent scheduling and the feasibility when applying this approach in real networks. In order to evaluate the proposed approach, an emulation platform is developed based on Mininet, POX and Open vSwitch. Emulation is conducted based on an actual network topology. Emulation results verify that the proposed approach is able to achieve effective load balancing for enterprise networks and avoid unnecessary flow scheduling.
AB - Inbound load balancing plays an important role in improving the quality of experience for enterprise networks. Traditional traffic engineering approaches fail to achieve fine-grained inbound load balancing because they have limited capability of obtaining detailed network status and lack flexibility for traffic flow scheduling. In this paper, we adopt OpenFlow to achieve effective inbound load balancing. The potential of implementing OpenFlow in an enterprise network is explored to schedule traffic among different inbound links. We also investigate how to avoid frequent scheduling and the feasibility when applying this approach in real networks. In order to evaluate the proposed approach, an emulation platform is developed based on Mininet, POX and Open vSwitch. Emulation is conducted based on an actual network topology. Emulation results verify that the proposed approach is able to achieve effective load balancing for enterprise networks and avoid unnecessary flow scheduling.
KW - Balancing
KW - Inbound Links
KW - OpenFlow
KW - SDN
KW - Traffic Load
UR - http://www.scopus.com/inward/record.url?scp=85047072034&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85047072034&partnerID=8YFLogxK
U2 - 10.1016/j.procs.2018.03.082
DO - 10.1016/j.procs.2018.03.082
M3 - Conference article
AN - SCOPUS:85047072034
SN - 1877-0509
VL - 129
SP - 319
EP - 323
JO - Procedia Computer Science
JF - Procedia Computer Science
T2 - International Conference on Identification,Information and Knowledgein The Internet of Things, 2017
Y2 - 19 October 2017 through 21 October 2017
ER -