Holistic Feistel Authenticated Learning-`Based Authorization for Protecting the Internet of Things from Cyber Attacks

As Artificial Intelligence (AI) and the Internet of Things (IoT) applications become more eminent, business and government organizations have started thinking aggressively about processing and analyzing data. Despite efficient data processing and analyzing, business organizations are faced with several issues, because, with data becoming an organization’s most valuable, data tampering creates the threat of data breaches, and cyberattacks nullify the provisioning of a smarter and more continuous working world. Hence, protecting the advantages that AI applications and IoT devices provide must remain the primary concern for today’s organizations. In this article, Holistic Feistel Authentication and Learning-based Authorization (HFA-LA) method present this for protecting IoT applications against cyberattacks. More specifically, a lightweight authentication approach is designed by exploring the use of Feistel Block Message Authentication by the cloud service manager to identify the authenticated devices. A holistic authentication and authorization approaches is designed. In which Artificial Intelligent Feed Forward Learning and trust management via error measurement are adopted to analyze the IoT applications to access the cloud resources. This new HFA-LA method establish the connections between IoT devices and cloud services and combat against cyberattacks so that the communication latency (i.e., communication time and overhead) is improved, and security risks are well controlled for IoT applications. Simulation is performed with different metrics, such as, communication time, communication overhead, true positive rate, and false-positive rate. The observed results show that the HFA-LA method efficiently improves the true positive rate and reduces the communication time, the overhead and false positive rate more than the state-of-the-art methods.