Enhancement of Network Management in Software Defined Network Using Mini Batch K-Mean Method

Software Defined Network (SDN) is a significant and well-known technology in the realm of networking that has resolves the issue of typical network encounters. Nevertheless, it may have bottlenecks and get overcrowded. To address this issue, several research works executed but they are based on two or three parameters to achieve load balancing and are either static or dynamic. The rapid growth of cloud computing as well as related services has resulted in increased network transmission and delay needs. Because of the inherent scattered nature of traditional networks, Machine Learning (ML) technique is challenging to apply and implement in network control. The introduction of SDN technology creates new possibilities as well as challenges for the use of ML in network management. A load balancing technique for the network controller with respect to Data Center Networks (DCNs) is accomplished using proposed SDN architecture. Furthermore, ML has assisted in automating software testing and maintenance processes, leading to reduced human effort and increased reliability of software systems. Hence, the proposed an intelligent clustering technique that forwards the packet or node traffic based on several parameters. The Mini Batch k-mean (MBK-mean) with hamely algorithm is used to minimize the iteration and improve the performance of k-mean cluster that determine the optimal nodes for executing the exact load balancing decision. This convergence of SDN and it established ushers in the next phase of intelligent, highly accurate and scalable DCNs that require additional investigation as well as investments in preparation for the future DCNs.