Detection of DDoS Attack in IoT Networks Using Sample elected RNN-ELM

IoT, a global technology paradigm, allows a variety of devices to be online at any time and from any location. But the widespread use of IoT devices has made them vulnerable to risks and vulnerabilities. Data breaches within IoT networks, particularly those that exist beyond the fog layer, have escalated as more businesses use IoT applications. In order to address this security concern, deep learning models have gained popularity as threat detection and mitigation techniques. An SSRNN-ELM hybrid approach is suggested in this study for the identification and mitigation of IoT data breaches. The SSRNN-ELM model consists of RNNs for supervised learning and ELMs for unsupervised classification. This hybrid design between the devices and fog layer protects IoT networks. Feature selection is the main innovation of the algorithm. These particular features impart these significant feature behaviour patterns to an RNN. Malicious and benign network data are separated by an ELM classifier in the final layer.We assess our hybrid SSRNN-ELM model's ability to identify DDoS attacks on IoT devices outside of the fog node by utilising the NSL-KDD benchmark dataset. Impressively, our model obtained 99% accuracy on the NSL-KDD dataset. When our strategy is compared with current models, its benefits become clear.To sum up, our hybrid SSRNN-ELM model offers a sophisticated and reliable solution to problems with IoT network security, especially DDoS detection that goes beyond the fog layer. With the help of feature selection, RNN-based learning, and ELM classification, our model effectively mitigates data vulnerabilities. The security of the IoT ecosystem and application integrity across industries are supported by this research.