An Intelligent Deep Learning Framework For Intrusion Detection In Iot Environment

Abstract

Software-Defined Networking (SDN) introduces a paradigm shift in the architecture of modern networks by decoupling the control and data planes, enhancing scalability, flexibility, and programmability. However, this architectural transformation also exposes new attack surfaces, especially in the Internet of Things (IoT) ecosystem, where SDN is increasingly adopted for traffic control and resource optimization. The SDN controller, acting as the brain of the network, becomes a prime target for cyberattacks such as Denial of Service
(DoS) and Distributed Denial of Service (DDoS). This research proposes a robust and intelligent hybrid deep learning framework combining Random Forest (RF) and Long Short-Term Memory (LSTM) networks for intrusion detection in SDN-based IoT environments. A specialized SDN-focused dataset (InSDN) is utilized to capture flow-level anomalies and system-specific behavior. To improve generalization and reduce overfitting, L2 regularization and Dropout techniques are applied. Feature selection methods are used to optimize the model and reduce computational complexity.
The proposed model is evaluated using performance metrics such as accuracy, precision, recall, F1-score, and detection time. Experimental results demonstrate that the hybrid RF-LSTM approach significantly outperforms conventional machine learning and standalone deep learning models in detecting a wide range of attacks with high accuracy and minimal latency. The lightweight nature of the proposed model makes it suitable for real-time deployment in resource-constrained IoT-SDN networks.