From Edge to Control: Benchmarking Software Defined Network Performance in IoT Networks

Abstract

Software Defined Networking (SDN) is an emerging technology that addresses networking complexity and management by separating the control plane from the data plane in switches and routers. This approach introduces key features such as centralized control and global network visibility, which support automation. In the context of the Internet of Things (IoT), SDN helps manage the wide variety of participating devices. By combining SDN controllers, Open vSwitch, and network virtualization, it offers significant operational and management benefits for IoT architectures. However, despite these advantages, existing studies have largely focused on conceptual models or simulated environments, providing limited empirical data on SDN controller performance within real IoT network conditions. This lack of experimental benchmarking makes it difficult to determine how different controllers behave under varying network loads and scales. Addressing this gap is essential to ensure reliable controller selection and deployment in IoT ecosystems. This study investigates SDN controller implementation in IoT networks, focusing on performance evaluation using two controllers, OpenDayLight and Floodlight. Performance is measured in terms of throughput and latency across three scenarios. Unlike prior comparative studies, this work provides a practical benchmarking framework using a real IoT testbed built on Raspberry Pi and Open vSwitch, offering new insights into controller behavior from the network edge to the control layer.