Abstract
The proliferation of Software-Defined Networking (SDN) and Internet of Things (IoT) has introduced new security challenges, necessitating effective Intrusion Detection Systems (IDS) tailored to the unique characteristics of SDN-IoT environments. This study presents a comprehensive evaluation of a newly generated IDS dataset specifically designed for SDN-IoT networks. The dataset is benchmarked against three widely used IDS datasets—InSDN, BoT-IoT, and ToN-IoT—using four deep learning architectures: Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), Recurrent Neural Network (RNN), and Deep Neural Network (DNN). Each model was trained and tested on all four datasets to assess performance across key metrics, including accuracy, precision, recall, F1-score, and computational efficiency. The results highlight the strengths and limitations of the proposed dataset in comparison to existing benchmarks, demonstrating the suitability of various deep learning models for anomaly detection in SDN-IoT contexts. This comparative analysis provides valuable insights for researchers and practitioners aiming to design robust, intelligent security systems in evolving network architectures.