AI-Driven Cybersecurity Threats to Future Networks [From the Guest Editors]
Sidi-Mohammed Senouci,Hichem Sedjelmaci,Jiajia Liu,Mubashir Husain Rehmani,Elias Bou-Harb
DOI: https://doi.org/10.1109/MVT.2020.3007981
IF: 13.609
2020-01-01
IEEE Vehicular Technology Magazine
Abstract:The articles in this special section focus on artificial intelligence-driven cybersecurity threats to future networks. Future-generation networks (5G and beyond 5G) will include a variety of services for various verticals, such as enhanced mobile broadband, health monitoring, Industry 4.0, smart energy distribution, and automotive networks. These vertical services and the critical components that comprise 5G architecture (e.g., radio access and edge and core networks) exhibit several cybersecurity vulnerabilities that attract attackers to use all their capabilities to exploit and hence shut down these networks. Recently, a new generation of smart threats, defined as artificial intelligence (AI ) attacks, has appeared. These smart attacks either turn AI into weapons to attack 5G services or hack the AI algorithms used by 5G components. In the first misbehavior, attackers take advantage of AI’s improved ability to launch lethal and stealthy threats against attractive targets, e.g., autonomous vehicles, drones, or manufacturing machinery. In the second misbehavior, attackers hack machine learning (ML) algorithms by modifying, for instance, the labels of the ML classification functions and altering the training data, causing a decrease in the accuracy of the classification rate.