Learning methodologies towards leveraging security resiliency in internet-of-things environment

Abstract

The evolution of artificial intelligence (AI) has faciliated a significant contribution of machine learning and deep learning in order to improvise the security features of large internet-of-things (IoT) environment. Since last decade there has been different variants of learning-based methodologies towards leveraging security improvements among communication in IoT devices; however, it is yet to know the strength and weakness of them. Hence, this paper presents a review of security methodologies adopted in machine learning and deep learning-based techniques in IoT to understand the degree of resiliency and effectiveness of these techniques. The paper further contributes towards highlighting the current methodologies with respect to benefits and limiting factors along with exclusive highlights of research trends while the research gap explored assists in offering these insights. The distinct findings of the study assist in paving the work direction in future by harnessing better form of learning scheme.