Abstract
The fast spread of Internet of Things (IoT) devices over many different fields has made network security even more crucial. Conventional security systems can fail to handle the dynamic and complex character of contemporary cyber threats aiming at IoT systems. This paper suggests a novel security framework combining blockchain technology, machine learning (ML), and a centralized iOS application to get past these constraints. The suggested approach guarantees privacy, integrity, and immutability of shared Cyber Threat Intelligence (CTI) data by using smart contracts and the Ethereum blockchain. Fundamentally, a hybrid deep learning model CNNTransLSTM is used to highly precisely detect and categorize threats in real-time. Combining Transformer encoders, Long Short-Term Memory (LSTM) networks, and Convolutional Neural Networks (CNN), this model efficiently records spatial and temporal aspects of IoT network data. By allowing users to report hazards and get alerts, the iOS app serves as an interactive hub improving human-machine cooperation. CNNTransLSTM model beats conventional approaches in terms of accuracy, sensitivity, and loss rate according to experimental evaluations. Moreover, the distributed blockchain architecture enables among stakeholders safe, open, and cooperative threat intelligence sharing. This all-encompassing strategy enables users and cloud providers to make quick, well-informed decisions to reduce risks, hence greatly improving the resilience of IoT ecosystems.
Keywords: Blockchain, Convolutional Neural Network Transformer (CNNTrans), Cyber Threat Intelligence (CTI), Internet of Things (IoT), Long Short-Term Memory (LSTM), Threat Intelligence (TI).