Adversarial Attacks and Defenses in 6G Network-Assisted IoT Systems

TL;DR

This survey addresses the vulnerability of AI-enabled 6G network-assisted IoT to adversarial attacks and surveys a comprehensive taxonomy of attacks and defenses. It combines theoretical background with Monte Carlo simulations that compare adversarial perturbations to traditional jamming, and analyzes attack pathways across emerging 6G technologies such as RIS, Massive MIMO, satellites, Metaverse, and SemCom. The key contributions include definitions, classifications, and architectures of attacks/defenses, plus practical defense strategies (adversarial training, denoising, detection, and certified approaches) and open challenges for robust AIoT security. The work provides a practical roadmap for securing next-generation IoT deployments, guiding researchers and practitioners toward adaptable defense mechanisms in the face of evolving adversarial threats.

Abstract

The Internet of Things (IoT) and massive IoT systems are key to sixth-generation (6G) networks due to dense connectivity, ultra-reliability, low latency, and high throughput. Artificial intelligence, including deep learning and machine learning, offers solutions for optimizing and deploying cutting-edge technologies for future radio communications. However, these techniques are vulnerable to adversarial attacks, leading to degraded performance and erroneous predictions, outcomes unacceptable for ubiquitous networks. This survey extensively addresses adversarial attacks and defense methods in 6G network-assisted IoT systems. The theoretical background and up-to-date research on adversarial attacks and defenses are discussed. Furthermore, we provide Monte Carlo simulations to validate the effectiveness of adversarial attacks compared to jamming attacks. Additionally, we examine the vulnerability of 6G IoT systems by demonstrating attack strategies applicable to key technologies, including reconfigurable intelligent surfaces, massive multiple-input multiple-output (MIMO)/cell-free massive MIMO, satellites, the metaverse, and semantic communications. Finally, we outline the challenges and future developments associated with adversarial attacks and defenses in 6G IoT systems.

Figures (5)

• Figure 1: Organization of the paper
• Figure 2: IoT devices
• Figure 3: Adversarial attacks based on the adversary's knowledge of the victim models.
• Figure 4: Standard Adversarial Defense Classification
• Figure 5: BLER of the auto-encoder system under Jamming and Adversarial attacks.