A Mathematical Model for SDN Control Plane Scalability Evaluation Based on Controller Utilization

Abstract

The increasing number of users on internet, clouds, and data centers should be followed by more efforts to change the behavior of traditional networks. Software-defined networks give more flexibility to the network, but the scalability of SDN is still a problem for network designers and administrators about which architecture is suitable to be applied in their network. As more studies study the scalability problem, controller utilization as a scalability metric for different SDN control planes still doesn’t have a deep analysis which is the novelty of this work. In this paper, we propose a mathematical model for SDN control plane scalability evaluation based on controller utilization as a metric. This paper employs mathematical techniques to examine and contrast various architectures. The numerical results conclude that the decentralized global view has the worst utilization, while the hierarchical root controller has the best, and the local decentralized control plan has almost the same utilization as the hierarchal leaf controller. These findings have significant implications for network designers and administrators as they suggest that implementing a hierarchical control plane can optimize the network scalability, ensuring efficient network management and resource utilization. This paper not only presents a mathematical model for evaluating SDN scalability, but also provides valuable insights into the implications of the findings. Network professionals can leverage this knowledge to make informed decisions and design more efficient and scalable SDN. The validity of the findings presented in this study is further substantiated through numerical assessments and achieved by proposing a novel method.

DOI: 10.61416/ceai.v25i4.8783