TY - JOUR
T1 - ESMLB
T2 - Efficient Switch Migration-Based Load Balancing for Multicontroller SDN in IoT
AU - Sahoo, Kshira Sagar
AU - Puthal, Deepak
AU - Tiwary, Mayank
AU - Usman, Muhammad
AU - Sahoo, Bibhudatta
AU - Wen, Zhenyu
AU - Sahoo, Biswa P.S.
AU - Ranjan, Rajiv
N1 - Publisher Copyright:
© 2014 IEEE.
PY - 2020/7
Y1 - 2020/7
N2 - In software-defined networks (SDNs), the deployment of multiple controllers improves the reliability and scalability of the distributed control plane. Recently, edge computing (EC) has become a backbone to networks where computational infrastructures and services are getting closer to the end user. The unique characteristics of SDN can serve as a key enabler to lower the complexity barriers involved in EC, and provide better quality-of-services (QoS) to users. As the demand for IoT keeps growing, gradually a huge number of smart devices will be connected to EC and generate tremendous IoT traffic. Due to a huge volume of control messages, the controller may not have sufficient capacity to respond to them. To handle such a scenario and to achieve better load balancing, dynamic switch migrating is one effective approach. However, a deliberate mechanism is required to accomplish such a task on the control plane, and the migration process results in high network delay. Taking it into consideration, this article has introduced an efficient switch migration-based load balancing (ESMLB) framework, which aims to assign switches to an underutilized controller effectively. Among many alternatives for selecting a target controller, a multicriteria decision-making method, i.e., the technique for order preference by similarity to an ideal solution (TOPSIS), has been used in our framework. This framework enables flexible decision-making processes for selecting controllers having different resource attributes. The emulation results indicate the efficacy of the ESMLB.
AB - In software-defined networks (SDNs), the deployment of multiple controllers improves the reliability and scalability of the distributed control plane. Recently, edge computing (EC) has become a backbone to networks where computational infrastructures and services are getting closer to the end user. The unique characteristics of SDN can serve as a key enabler to lower the complexity barriers involved in EC, and provide better quality-of-services (QoS) to users. As the demand for IoT keeps growing, gradually a huge number of smart devices will be connected to EC and generate tremendous IoT traffic. Due to a huge volume of control messages, the controller may not have sufficient capacity to respond to them. To handle such a scenario and to achieve better load balancing, dynamic switch migrating is one effective approach. However, a deliberate mechanism is required to accomplish such a task on the control plane, and the migration process results in high network delay. Taking it into consideration, this article has introduced an efficient switch migration-based load balancing (ESMLB) framework, which aims to assign switches to an underutilized controller effectively. Among many alternatives for selecting a target controller, a multicriteria decision-making method, i.e., the technique for order preference by similarity to an ideal solution (TOPSIS), has been used in our framework. This framework enables flexible decision-making processes for selecting controllers having different resource attributes. The emulation results indicate the efficacy of the ESMLB.
KW - Controller
KW - software-defined network (SDN)
KW - switch migration
KW - technique for order preference by similarity to an ideal solution (TOPSIS)
UR - http://www.scopus.com/inward/record.url?scp=85089309161&partnerID=8YFLogxK
U2 - 10.1109/JIOT.2019.2952527
DO - 10.1109/JIOT.2019.2952527
M3 - Article
AN - SCOPUS:85089309161
SN - 2327-4662
VL - 7
SP - 5852
EP - 5860
JO - IEEE Internet of Things Journal
JF - IEEE Internet of Things Journal
IS - 7
M1 - 8894397
ER -