Fair Dynamic Switch Migration and Scheduling in SDN for Next Generation Internet of Things (NG-IoT)

Due to the advancements in smart devices, leading to the emergence of Next Generation Internet of Things (NG-IoT), there has been a remarkable increase in the number of users and clients choosing to use these devices for personal or business purposes. To ensure optimal performance, Software Defined Networking (SDN) is employed to adapt the NG-IoT network by effectively managing traffic patterns to reduce delays and ensure sufficient bandwidth. This research explores and quantify the role of SDN in NG-IoT, specifically focusing on the identification of bottlenecks related to migration, load balancing, and scheduling issues within the provisioned resources.

An aggregation module is proposed, which operates based on variable thresholds. By aggregating controller data with the same destination into bursts, is transmitted through the data plane, the communication overhead and control-related tasks can be greatly reduced. Each burst consists of multiple packets, with a single control packet, effectively streamlining the process. A novel framework is proposed for the optimized selection of controllers, utilizing the Non-dominated Sorting Genetic Algorithm - II (NSGA-II) algorithm integrated with network simulator - 3 (ns-3). The selection of the most suitable controller for migration is determined by evaluating parameters, like performance and cost, through mathematical model using the NSGA-II algorithm.

Additionally, a novel method is included to ensure fair switch selection, enabling the transfer of workload to underutilize controllers to meet quality of service preservation and fairness in the network. Experimental analysis confirms the effectiveness of the proposed frameworks in terms of cost, performance, delay, throughput, and other relevant performance metrics. This framework employs mathematical model to determine the optimal controller selection, thereby facilitating efficient migration processes. By combining optimized controller selection with aggregation techniques, the proposed framework incorporates migration processes and compares the results with state-of-the-art methods. The proposed approach is able to ensure seamless migration while maintaining high Quality of Service (QoS) standards and promoting equitable resource allocation.