Securing the IoT System of Smart Cities by Interactive Layered Neuro-Fuzzy Inference Network Classifier With Asymmetric Cryptography

Abstract

Smart environments (SE) aim to improve daily comfort in the form of the internet of things (IoT). It starts many everyday services due to its stable and easy-to-use operations. Any real-world SE based on IoT architecture prioritises privacy and security. Internet of things systems are vulnerable to security flaws, affecting SE applications. To identify attacks on IoT smart cities, an IDS based on an iterative layered neuro-fuzzy inference network (ILNFIN) is presented. Initially the TON-IoT dataset was preprocessed, and the sparse wrapper head selection approach isolates attack-related features. The Iterative stacking neuro-fuzzy inference network classifies attacked data from the normal data. The asymmetric prime chaotic Rivest Shamir Adleman technique ensures the secure transmission of non-attacked data. To show the effectiveness of the suggested secure data transfer techniques, the authors compare their experimental results to existing approaches.