AI-Driven Security Solutions for the Internet of Everything

Deepak Puthal, Amit Mishra, Suraj Sharma
2021 IEEE Consumer Electronics Magazine  
& THE INTERNET OF Everything (IoE) integrates the components, such as people, things, data, and process for the smart system building, and where the Internet of Things (IoT) is integrated as one of its architectural components. 1 Individual components are connected to the Internet to build an autonomous system. The IoE application ranges from smart cities, smart healthcare to battlefield monitoring. 2 The openness properties of the Internet welcome several attackers to the systems; thus, the
more » ... plete system gets compromised without meeting the purpose of the applications. To avoid such scenarios, many security solutions exist, including cryptography-based solutions and context-aware security solutions. Above all these techniques, artificial intelligence (AI) has been emerging as a promising approach to secure the IoE-based complex system. 3 Machine learning (ML), deep learning (DL), and these days federated learning (FL) are the subsets of AI to power AI-empowered systems through learning. 2,4 Here, models learn from the historical data to validate current input and predict the future in a dynamic system. This motivates us to secure a system from a different approach by generating an attacker and protector approach. For a dynamic scenario, including IoT, conventional security mechanisms are inefficient to predict future attacks. 3,5 We need an AIenabled security solution that can predict the potential future security threats and strengthen the secure solution accordingly. Generative adversarial networks (GAN) is a generative-AI approach dedicated to learning the real distributions from the data. 2 It has two components, namely, generator and discriminator. The generator generates vague samples to add to the real data. In contrast, the discriminator aims to extricate the fake samples from the original data. 2 The security threats of an IoE system and possible AI-based solutions to mitigate the threats are shown in Figure 1 , which is abstracted from Tahsien et al. 's work. 5 In this Special Section, we have four interesting articles discussing the cutting-edge work on endeavors toward making IoE more secure. IoE provides numerous benefits with connected devices. Federated learning is another recent innovation providing a methodology to have machine learning algorithms for edge-learning for IoE. The article "RR-LADP: A privacyenhanced federated learning scheme for Internet of Everything" proposes privacy-aware federated learning for IoE by adopting two mechanisms, i.e.,
doi:10.1109/mce.2021.3071676 fatcat:663acipcprfavpv5d33paiyjoa