Defending Against IoT Threats: A Comprehensive Framework with Advanced Models and Real-Time Threat Intelligence for DDoS Detection

The incorporation of Software Defined Networking (SDN) with IoT emerges as an auspicious strategy for enhancing security and access control mechanisms. Nevertheless, substantial threats from DDoS attacks persist in IoT networks, with the potential for execution over botnet or zombie attack, incorporating a machine learning-based detection scheme enables establishing it a framework have been introduced to analyze the performance of IoT devices. Profiles that inform decision-making processes are amassed by these frameworks, ultimately protecting the security of the IoT devices. This work presents a machine learning-based approach for detecting DDoS attacks in an SDN-WISE IoT controller. The incorporation of a machine learning-based detection scheme enables the establishment of a testbed environment for simulating the traffic of DDoS attacks. A logging mechanism within the SDN-WISE controller detects the traffic, generating network logs that are pre-processed and transformed into a dataset. The machine learning DDoS detection module entrenched in the SDN-WISE controller uses Naive Bayes (NB), Decision Tree (DT), and Support Vector Machine (SVM) algorithms to categorize SDN-IoT data. The proposed framework's performance is assessed across various traffic simulation scenarios, and the results from the machine learning DDoS detection module are analyzed. The proposed framework exhibits an accuracy rate of 98.2%, 97.1%, and 97.18% for NB, SVM, and DT, respectively. The attack detection module consumes up to 36% of memory and CPU usage, saving approximately 64% of memory while maintaining up to 71% CPU availability to analyze SD-IoT network traffic. The proposed framework achieves an accuracy of 98.2% which proves the efficiency of the work over state of the work models.