Enhancing IoT (Internet of Things) Security Through Intelligent Intrusion Detection Using ML Models
Mukund Sai Vikram Tyagadurgam, Venkataswamy Naidu Gangineni, Sriram Pabbineedi, Mitra Penmetsa, Jayakeshav Reddy Bhumireddy, Rajiv Chalasani · International Journal of Emerging Research in Engineering and Technology · 2021
Through IoT technology wireless communications experience a fundamental transformation that reshapes various industries. Multiple cyberattacks exploit the limited capacity and broad exposure among IoT networks. IDS systems require advanced technologies because existing security systems fail to detect new potential threats. This research proposes Long Short-Term Memory (LSTM)-based deep learning models to develop an intelligent intrusion detection system (IDS) that improves IoT security. The LSTM model performs training on the ToN-IoT dataset data after applying multiple preparation steps, including cleaning and normalization and encoding different features. The model's remarkable detection skills are demonstrated by a number of evaluations, such as its 99.41% detection accuracy, 99.35% precision, 99.32% recall, and 99.33% F1-score. By employing an implemented LSTM model researcher could achieve higher classification success rates than a DBN model serves as validation for monitoring threat detection and temporal pattern measurement. The suggested method provides a strong, scalable, and flexible IoT intrusion detection solution, enhancing security for IoT settings that are becoming more intricate and networked