Machine Learning-Based Intrusion Detection Approaches for Secured Internet of Things

Abstract

Nowadays, protecting communication and information for Internet of Things (IOT) has emerged as a critical challenge. Existing systems use firewalls to ensure that they are safe from any unexpected occurrences that may disrupt the desired systems and applications. Intrusion detection systems (IDSs) are an acceptable second line of defence for IOT applications. IDS play a crucial role ensuring that it enhances the IOT security level maintaining sophisticated framework. Attackers have continuously been attempting to determine novel ways to circumnavigate security frameworks that prevent the structures. This paper reviews the security advances, threats and countermeasures for the IOT applications. A state of art review has accomplished using the references from 2009 to 2020 to encompass the real demography of the IOT security research data. This work also highlights the deep learning-based intrusion detection approaches for Internet of Things (IOT) security. With the systematic literature review approach, the review suggests that implementing existing security measures, such as encryption, authentication, access control, network and application security for IoT systems and their intrinsic amenability is ineffective for the IOT systems.