Blockchain and Reinforcement Neural Network for Trusted Cloud-Enabled IoT Network

• Published in: IEEE transactions on consumer electronics Vol. 70; no. 1; pp. 2311 - 2322
• Main Authors: Samriya, Jitendra Kumar, Kumar, Surendra, Kumar, Mohit, Xu, Minxian, Wu, Huaming, Gill, Sukhpal Singh
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.02.2024; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: 5G mobile communication; Artificial intelligence; Blockchain; Blockchains; Cloud computing; Computer architecture; Consumers; control system; Cryptography; Cybersecurity; Data collection; Denial of service attacks; Fault detection; heterogeneous autonomous network; Internet of Things; Malware; Network security; Neural networks; Recurrent neural networks; reinforcement neural network; Security
• ISSN: 0098-3063; 1558-4127
• DOI: 10.1109/TCE.2023.3347690

The rapid integration of Internet of Things (IoT) services and applications across various sectors is primarily driven by their ability to process real-time data and create intelligent environments through artificial intelligence for service consumers. However, the security and privacy of data have emerged as significant threats to consumers within IoT networks. Issues such as node tampering, phishing attacks, malicious code injection, malware threats, and the potential for Denial of Service (DoS) attacks pose serious risks to the safety and confidentiality of information. To solve this problem, we propose an integrated autonomous IoT network within a cloud architecture, employing Blockchain technology to heighten network security. The primary goal of this approach is to establish a Heterogeneous Autonomous Network (HAN), wherein data is processed and transmitted through cloud architecture. This network is integrated with a Reinforced Neural Network (RNN) called ClouD_RNN, specifically designed to classify the data perceived and collected by sensors. Further, the collected data is continuously monitored by an autonomous network and classified for fault detection and malicious activity. In addition, network security is enhanced by the Blockchain Adaptive Windowing Meta Optimization Protocol (BAW_MOP). Extensive experimental results validate that our proposed approach significantly outperforms state-of-the-art approaches in terms of throughput, accuracy, end-to-end delay, data delivery ratio, network security, and energy efficiency.