A Hybrid Deep Learning Model-Based Intrusion Detection System for Emergency Planning Using IoT-Network

The ICT solutions based on Machine Learning (ML) and Internet of Things (IoT) are gaining traction in disaster management systems. They have applications in disaster management, early warning and emergency response, search and rescue missions and many more. All the IoT devices ought to ensure a full proof secured environment. Traditional intrusion detection system (IDS) is unable to detect cyber-attacks. Integrating machine learning with IoT provides a sense of intelligence to ensure security aspects. Basically, the intelligent-IDS framework depends on training & testing of datasets. Using machine learning-based security classifiers, it can easily obtain high accuracy, reduce false alarms, identify illegal data packets and provide optimal solutions to detect malicious activities. This paper presents a hybrid deep neural network model to detect intrusion in IoT-network. Several classifiers using CICIDS2017 and IoTID20 datasets are utilised in a simulation environment. IoT using machine learning and neural networks is a novel approach which feeds data packets to automate the entire network. Data traffic management in IoT networks is properly discussed. More interestingly, the uniqueness of this study provides a detailed analysis of techniques evaluating various parameters/metrics.