Predictive Mobility-Aware Analysis of Controller Overhead in SDN-Enabled IoT

The rapid growth and increasing mobility of Internet of Things (IoT) mobile devices significantly intensify flow setup requests in Software-Defined Networking (SDN). Frequent movement of mobile devices across access points triggers repeated flow rule installations in data-plane switches, leading to excessive controller overhead under limited flow-table memory capacity. Inaccurate mobility prediction in existing approaches further aggravates this challenge by causing unnecessary rule placement and redundant flow setup requests. This study investigates the impact of mobility-aware prediction on controller overhead with experimental scenarios using different groups of mobile IoT devices, while considering controller CPU and memory utilization. A proactive mobility prediction model is employed to reduce reactive controller interactions. The proposed prediction model demonstrated approximately 100% accuracy compared to the benchmark model, significantly reducing unnecessary flow setup requests and minimising controller overhead. Furthermore, sensitivity analysis reveals that lower prediction error reduces controller load, whereas higher prediction error leads to increased controller overhead, highlighting the importance of accurate mobility prediction for scalable SDN-enabled IoT networks.