Introducing Deep Learning for IoT Security
Mohamed Abdel‐Basset, Nour Moustafa, Hossam Hawash · 2022
This chapter emphasizes the attributes of typical Internet of Things (IoT) systems and presents the most significant security risks that such systems may be subjected to in their operation. A lot of research efforts have been devoted to studying and categorizing IoT vulnerabilities and attacks. Given the attacker's activity as a categorization criterion, IoT attacks can be broadly classified as passive and active attacks. IoT risks or malicious behaviors can be detected early on in the IoT system's lifecycle by aggregating and analyzing data provided by various IoT sectors, which can then be studied by one or some artificial intelligence (AI) approaches to distinguish regular behaviors from malicious behaviors. Machine learning (ML) is regarded as a widely applied AI paradigm. Deep learning is an ML technology that creates deeper versions of neural networks that imitate the composition and functionality of the human brain.