A Smart Home System based on Internet of Things

TL;DR

This paper proposes a multi-factor authentication system for IoT smart homes, integrating face recognition and liveness detection to enhance security against unauthorized access.

Contribution

It introduces a novel software architecture that incorporates face recognition and liveness detection into IoT smart home authentication, addressing gaps in existing systems.

Findings

Enhanced security through multi-factor authentication

Inclusion of face recognition and liveness detection

Improved user registration and login processes

Abstract

The Internet of Things (IoT) describes a network infrastructure of identifiable things that share data through the Internet. A smart home is one of the applications for the Internet of Things. In a smart home, household appliances could be monitored and controlled remotely. This raises a demand for reliable security solutions for IoT systems. Authorization and authentication are challenging IoT security operations that need to be considered. For instance, unauthorized access, such as cyber-attacks, to a smart home system could cause danger by controlling sensors and actuators, opening the doors for a thief. This paper applies an extra layer of security of multi-factor authentication to act as a prevention method for mitigating unauthorized access. One of those factors is face recognition, as it has recently become popular due to its non-invasive biometric techniques, which is easy to use with cameras attached to most trending computers and smartphones. In this paper, the gaps in existing IoT smart home systems have been analyzed, and we have suggested improvements for overcoming them by including necessary system modules and enhancing user registration and log-in authentication. We propose software architecture for implementing such a system. To the best of our knowledge, the existing IoT smart home management research does not support face recognition and liveness detection within the authentication operation of their suggested software architectures.