IoT Security and Authentication schemes Based on Machine Learning: Review

TL;DR

This review paper analyzes various machine learning-based IoT authentication schemes, focusing on behavioral biometrics and physical layer methods, discussing their advantages, challenges, and potential for enhancing device security.

Contribution

It provides a comprehensive survey of current authentication models in IoT, highlighting their architectures, benefits, and limitations to guide future research.

Findings

Various authentication schemes have different accuracy levels.

Usability and security trade-offs are common in current models.

Physical layer authentication shows promising security features.

Abstract

With the latest developments in technology, extra and extra human beings depend on their private gadgets to keep their touchy information. Concurrently, the surroundings in which these gadgets are linked have grown to grow to be greater dynamic and complex. This opens the dialogue of if the modern day authentication strategies being used in these gadgets are dependable ample to preserve these user's records safe. This paper examines the distinct consumer authentication schemes proposed to make bigger the protection of exceptional devices. This article is break up into two one of a kind avenues discussing authentication schemes that use both behavioral biometrics or physical layer authentication. This survey will talk about each the blessings and challenges that occur with the accuracy, usability, and standard protection of computing device getting to know strategies in these authentication systems. This article targets to enhance in addition lookup in this subject via exhibiting the more than a few present day authentication models, their schematics, and their results.