Controller Placement in Software-defined Networking Using Silhouette Analysis and Gap Statistic
Sanaz Honarpazhooh · Turkish Journal of Computer and Mathematics Education (TURCOMAT) · 2021
Software-defined Networks (SDNs) are an emerging network architecture that has been introduced as a way to facilitate network development in such a way that network control is separate from traffic transmission and is directly programmed. Typically, an SDNs relies on a centralized logic control that uses the network's global information to perform operations within the network. Therefore, the SDNs have scalability problems. In this case, the proper placement of the controllers in the SDN increases the quality of service. In this dissertation, a DBSCAN-based clustering method is proposed with the characteristic of non-dependence on determining the number of clusters in advance, which can be used to properly locate the controllers based on the number clusters. The proposed method is simulated and the results show that the proposed method in the worst case load on SDNs works better than similar methods and has improved performance compared to the base method.