Dynamic SDN Multiple Nodes Migration Using SARSA Reinforcement Learning

This article addresses the problem of software-defined Networking (SDN) migration process under dynamic network conditions and present a SARSA-based approach to tackle the challenges associated with it. Traditional SDN migration techniques often fail to adapt effectively to dynamic network environments, leading to suboptimal performance. To overcome these limitations, this article explores the design of State–Action–Reward–State–Action (SARSA) reinforcement learning in SDN migration process. The article presents a comprehensive analysis of the existing literature on SDN migration process and the design of reinforcement learning, highlighting the limitations of current approaches in dynamic network scenarios. It then presents a SARSA-based SDN migration system that utilizes reinforcement learning to adaptively migrate multiple network nodes in response to changing network conditions. Overall, this article gives insightful information into the design of SARSA reinforcement learning in SDN migration process.