An Advanced System for Detecting Malware in the Internet of Things Using Artificial Intelligence and Blockchain

Abstract

The Internet of Things (IoT) has witnessed significant and rapid expansion in various industrial and service sectors, causing increased security challenges related to malware detection and classification. Traditional models suffer from poor handling of temporal relationships and nonlinear patterns in data, which negatively impacts detection accuracy and efficiency in detecting complex malware. This study relied on artificial intelligence techniques to automatically detect malware in Internet of Things networks by analyzing abnormal temporal patterns and behaviors. Three deep learning models, namely recurrent neural networks (RNNs), long- and short-term memory units (GRU), and neural networks (NNs), have been applied to analyze long-term temporal relationships and achieve accurate and efficient data classification. The results indicated that the GRU algorithm outperformed both the RNN (99.54%) and the NN (99.36%) in terms of accuracy, achieving a score of 99.61%. To enhance system security, blockchain technology has been integrated to provide an additional layer of security, integrity, and data reliability through its decentralized features. Transparency and secure documentation of transactions. It also contributed to recording the results of detection and transactions transparently and securely and achieved an average confidence level of 93.7%, which enhances the reliability of the system in IoT environments. The system was tested using the Edge-IoT dataset, and the results showed high efficiency in detection, classification, and protection tasks. The results indicate that the proposed model represents an effective and reliable solution for enhancing the security of Internet of Things (IoT) systems.