A Stratified IoT Deep Learning based Intrusion Detection System

Abstract

The Internet of Things (IoT) enables billions of intelligent linked devices to communicate through the IP standard. With the Internet and the Cloud Computing environment, nearly any system can be established and turned smarter. Regardless of their size or form, all IoT devices need processing, security, sensing, and actuation to work successfully; yet, there are currently no IoT-specific security standards. Users and IoT devices security are often neglected by IoT product designers and manufacturers. The functionality of some IoT devices may also be manipulated or handicapped by malicious actors, causing infected IoT devices to behave differently when it comes to defending against these attacks, as well as being a part of these attacks. To address these issues, we propose in this paper a Stratified Deep Learning Based-Intrusion Detection System (SDL-IDS) for the IoT environment at the three levels, Edge, Fog, and Cloud, in order to enhance the security of IoT networks, this proposed SDL-IDS is composed of blocks that act in collaboration.