A Smart NIDS Design for SDN-based Cloud IoT Network using Optimal Path Selection and RNN

The Internet of Things (IoT) has greatly increased the possibility for developing intelligent connections and applications in many facets of daily life. Traditional security solutions frequently fail to solve security issues in cloud-based IoT systems, proving ineffective and insufficient. Software Defined Networking (SDN), which effectively detects and monitors network security vulnerabilities thanks to its programmable characteristics, presents a viable solution to these problems. In order to strengthen computer systems and fend off threats to network security, machine learning (ML) techniques have recently been included into SDN-Network Intrusion Detection Systems (NIDS). Deep learning (DL), one of these cutting-edge ML techniques, has become more popular in the SDN environment. For SDN-based cloud IoT networks, this paper design an efficient path selection system in this study utilizing DL approaches. The strategy comprises employing a Recurrent Neural Network (RNN) to identify threats and a selection mechanism to pick the best route for data delivery. This study performed tests on two datasets, KDD and UNSWNB15, and the results show that the suggested RNN outperformed existing DL methods by achieving an amazing accuracy ranging from 91% to 95% on both datasets.