🤖 AI Summary
Existing performance evaluations of SDN controllers lack comprehensive coverage of realistic scenarios—such as asymmetric topologies, dynamic topology reconfigurations, and high-concurrency flow bursts—limiting insights into practical bottlenecks. Method: We design a programmable emulation platform based on Mininet and OpenFlow 1.3, introducing a standardized experimental framework and automated performance measurement scripts to enable reproducible, multi-controller comparative analysis. Contribution/Results: Our systematic evaluation reveals that Ryu exhibits up to a 37% throughput degradation in tree- and ring-shaped topologies; control-path latency scales quadratically (O(n²)) with network size. Crucially, we quantitatively identify previously unreported deficiencies: drastic flow-table update latency spikes under dynamic topology changes and imbalanced handling of asymmetric links. These findings provide empirically grounded guidance for SDN controller selection, architectural refinement, and protocol enhancement, along with a publicly replicable benchmarking methodology.
📝 Abstract
Software-defined networking (SDN) represents a revolutionary shift in network technology by decoupling the data plane from the control plane.}In this architecture, all network decision-making processes are centralized in a controller, meaning each switch receives routing information from the controller and forwards network packets accordingly. This clearly highlights the crucial role that controllers play in the overall performance of SDN. Ryu is one of the most widely used SDN controllers, known for its ease of use in research due to its support for Python programming. This makes Ryu a suitable option for experimental and academic studies. In this research, we evaluate the performance of the Ryu controller based on various network metrics and across different network topologies. For experimental analysis, we use Mininet, a powerful network emulation tool that enables the creation of diverse network structures and the connection of switches to controllers. To facilitate the experiments, we developed a Python-based script that executes various network scenarios, connects to different controllers, and captures and stores the results. This study not only provides a comprehensive performance evaluation of the Ryu controller but also paves the way for evaluating other SDN controllers in future research.