Artificial Intelligence approaches for IoT Security: State of Art

Abstract

With an estimation of more than 35 billion interconnected smart devices by the end of 2021, Internet of Things (IoT) is one of the most rapidly growing technologies in the last decade. However, the complexity nature of IoT systems and the exponential amount of data collected and exchanged between Things relieve a big challenge in terms of security and privacy. Implementing classical security measures, such as encryption, authentication, access control, network and application security for IoT devices is no more effective against sophisticated Cyber-attacks. Artificial intelligence (AI) approaches such as Machine Learning (ML), Deep Learning (DL) and Blockchain can be leveraged to enhance the security of IoT and deal with its various problems. In this paper, we will describe the IoT technology and its domain of application, the protocols used to communicate between smart devices, security issues and existing AI solution.