Simulative Evaluation of KPIs in SDN for Topology Classification and Performance Prediction Models

In recent years Software-defined Networking (SDN) has gained increased popularity by reducing the complexity of management operations and increasing the performance of net-works. As a result, the number of purchasable SDN devices and their deployment in real world network constantly rises, making a thorough understanding of their behaviors and interactions even more important. In this work we analyze distributed controller architectures via simulation, identify Key Performance Indicators (KPIs), classify various network topologies with respect to their impact on SDN, as well as create a prediction model to estimate the overall performance of the overall SDN ecosystem, which can be used for network planning.