Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Edge Learning for 6G-enabled Internet of Things: A Comprehensive Survey of Vulnerabilities, Datasets, and Defenses

Mohamed Amine Ferrag, Othmane Friha, Burak Kantarci, Norbert Tihanyi, Lucas Cordeiro, Merouane Debbah, Djallel Hamouda, Muna Al-Hawawreh, Kim-Kwang Raymond Choo

TL;DR

This survey analyzes vulnerabilities, datasets, and defenses for edge learning in 6G-enabled IoT, detailing ML threat models across centralized, federated, and distributed paradigms. It uncovers eight attack categories (backdoors, adversarial examples, poisoning, Sybil, Byzantine, inference, and dropping) and benchmarks a wide spectrum of defense strategies spanning training, post-training, and inference phases. The work aggregates a comprehensive dataset taxonomy, testbeds, and practical deployment considerations, highlighting challenges such as non-IID data, privacy, and zero-day threats, while offering a structured framework for designing robust, trustworthy 6G edge-learning systems. Overall, the paper provides a foundational reference for researchers and practitioners to evaluate, compare, and advance defense mechanisms in AI-driven 6G IoT ecosystems, with emphasis on real-world datasets, scalable defenses, and future research directions in edge intelligence and security.

Abstract

The ongoing deployment of the fifth generation (5G) wireless networks constantly reveals limitations concerning its original concept as a key driver of Internet of Everything (IoE) applications. These 5G challenges are behind worldwide efforts to enable future networks, such as sixth generation (6G) networks, to efficiently support sophisticated applications ranging from autonomous driving capabilities to the Metaverse. Edge learning is a new and powerful approach to training models across distributed clients while protecting the privacy of their data. This approach is expected to be embedded within future network infrastructures, including 6G, to solve challenging problems such as resource management and behavior prediction. This survey article provides a holistic review of the most recent research focused on edge learning vulnerabilities and defenses for 6G-enabled IoT. We summarize the existing surveys on machine learning for 6G IoT security and machine learning-associated threats in three different learning modes: centralized, federated, and distributed. Then, we provide an overview of enabling emerging technologies for 6G IoT intelligence. Moreover, we provide a holistic survey of existing research on attacks against machine learning and classify threat models into eight categories, including backdoor attacks, adversarial examples, combined attacks, poisoning attacks, Sybil attacks, byzantine attacks, inference attacks, and dropping attacks. In addition, we provide a comprehensive and detailed taxonomy and a side-by-side comparison of the state-of-the-art defense methods against edge learning vulnerabilities. Finally, as new attacks and defense technologies are realized, new research and future overall prospects for 6G-enabled IoT are discussed.

Edge Learning for 6G-enabled Internet of Things: A Comprehensive Survey of Vulnerabilities, Datasets, and Defenses

TL;DR

This survey analyzes vulnerabilities, datasets, and defenses for edge learning in 6G-enabled IoT, detailing ML threat models across centralized, federated, and distributed paradigms. It uncovers eight attack categories (backdoors, adversarial examples, poisoning, Sybil, Byzantine, inference, and dropping) and benchmarks a wide spectrum of defense strategies spanning training, post-training, and inference phases. The work aggregates a comprehensive dataset taxonomy, testbeds, and practical deployment considerations, highlighting challenges such as non-IID data, privacy, and zero-day threats, while offering a structured framework for designing robust, trustworthy 6G edge-learning systems. Overall, the paper provides a foundational reference for researchers and practitioners to evaluate, compare, and advance defense mechanisms in AI-driven 6G IoT ecosystems, with emphasis on real-world datasets, scalable defenses, and future research directions in edge intelligence and security.

Abstract

The ongoing deployment of the fifth generation (5G) wireless networks constantly reveals limitations concerning its original concept as a key driver of Internet of Everything (IoE) applications. These 5G challenges are behind worldwide efforts to enable future networks, such as sixth generation (6G) networks, to efficiently support sophisticated applications ranging from autonomous driving capabilities to the Metaverse. Edge learning is a new and powerful approach to training models across distributed clients while protecting the privacy of their data. This approach is expected to be embedded within future network infrastructures, including 6G, to solve challenging problems such as resource management and behavior prediction. This survey article provides a holistic review of the most recent research focused on edge learning vulnerabilities and defenses for 6G-enabled IoT. We summarize the existing surveys on machine learning for 6G IoT security and machine learning-associated threats in three different learning modes: centralized, federated, and distributed. Then, we provide an overview of enabling emerging technologies for 6G IoT intelligence. Moreover, we provide a holistic survey of existing research on attacks against machine learning and classify threat models into eight categories, including backdoor attacks, adversarial examples, combined attacks, poisoning attacks, Sybil attacks, byzantine attacks, inference attacks, and dropping attacks. In addition, we provide a comprehensive and detailed taxonomy and a side-by-side comparison of the state-of-the-art defense methods against edge learning vulnerabilities. Finally, as new attacks and defense technologies are realized, new research and future overall prospects for 6G-enabled IoT are discussed.
Paper Structure (119 sections, 13 equations, 15 figures, 19 tables, 2 algorithms)

This paper contains 119 sections, 13 equations, 15 figures, 19 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Review methodology
  3. Objectives and research questions
  4. Search strategy
  5. Data extraction
  6. Quality assessment
  7. 6G–IoT Intelligence: an overview
  8. Architectural Landscape
  9. Physical (PHY) Layer Enhancements
  10. Medium Access Control (MAC)
  11. Core-Network and Services Intelligence
  12. 6G architecture for energy-efficient communication
  13. 5G/6G IoT testbeds
  14. Supported Technological Prospects
  15. In-Network Domains
  16. ...and 104 more sections

Figures (15)

  • Figure 1: Structure of this survey paper.
  • Figure 2: Distributed Edge Learning vs. Federated Edge Learning vs. Centralized Edge Learning.
  • Figure 3: AI as a native component of 6G concerning the operational, defensive, and targeted perspectives: 1) Intelligent Sensing/Edge: This comprises two primary components: the first is the data generation aspect, which includes devices, systems, and processes from which data originates; the second is the edge layer, where certain cloud processing tasks are executed at the network's periphery; 2) Intelligent Control: This pertains to smart network management, primarily at the network core (e.g., Core Network Data Analytics Function (CNDF) chouman2022towards); 3) Smart Applications: These encompass present and future intelligent applications that utilize the network; Additionally, we consider various AI perspectives within the network, such as operational, defensive, and targeted.
  • Figure 4: Perspectives on 6G IoT and Its Transformative Potential
  • Figure 5: Predicted AI involvement in future 6G networks: 1) Architectural landscape: AI as an enabler of native intelligent functionality. 2) Security and privacy: AI as an embedded/additive defender; 3) Technology Prospects: AI as an enabling intelligence service for high-level layers.
  • ...and 10 more figures