List of Contributors xiii Acknowledgements xix Introduction xxi 1 Threat Landscape for 6G-Enabled Massive IoT 1 Maria Papaioannou, Georgios Mantas, Firooz B. Saghezchi, Georgios Kambourakis, Felipe Gil-Castiñeira, Raúl Santos de la Cámara, and Jonathan Rodriguez 1.1 Introduction 1 1.2 6G Vision and Core Values 2 1.3 Emerging Massive IoT Applications Enabled by 6G 3 1.3.1 Enabling Sustainability 4 1.3.

1.1 E-health for All 5 1.3.1.2 Institutional Coverage 6 1.3.1.3 Earth Monitor 6 1.

3.1.4 Autonomous Supply Chains 6 1.3.1.5 Sustainable Food Production 7 1.3.1.

6 Network Trade-Offs for Minimized Environmental Impact 7 1.3.1.7 Network Functionality for Crisis Resilience 8 1.3.2 Massive Twinning 8 1.3.2.

1 Digital Twins for Manufacturing 9 1.3.2.2 Immersive Smart Cities 10 1.3.2.3 Internet of Tags 11 1.3.

3 Telepresence 12 1.3.3.1 Fully Merged Cyber-Physical Worlds 12 1.3.3.2 Mixed Reality Co-design 14 1.3.

3.3 Immersive Sport Event 14 1.3.3.4 Merged Reality Game/Work 15 1.3.4 From Robots to Cobots 15 1.3.

4.1 Consumer Robots 15 1.3.4.2 AI Partners 16 1.3.4.3 Interacting and Cooperative Mobile Robots 16 1.

3.4.4 Flexible Manufacturing 17 1.3.4.5 Situation-Aware Device Reconfiguration 17 1.3.5 Trusted Embedded Networks 18 1.

3.5.1 Human-Centric Communications 18 1.3.5.2 Infrastructure-less Network Extensions and Embedded Networks 19 1.3.5.

3 Local Coverage for Temporary Usage 20 1.3.5.4 Small Coverage, Low-Power Micro-Network in Networks for Production and Manufacturing 20 1.3.6 Hyperconnected Resilient Network Infrastructures 20 1.3.6.

1 Sensor Infrastructure Web 21 1.3.6.2 AI-Assisted Vehicle-to-Everything (V2X) 21 1.3.6.3 Interconnected IoT Micro-Networks 22 1.3.

6.4 Enhanced Public Protection 23 1.4 Overview of a 6G Network Architecture to Enable Massive IoT 23 1.4.1 Space Network 24 1.4.2 Air Network 25 1.4.

3 Ground Network 25 1.4.4 Sea/Underwater Network 26 1.4.4.1 Surface Network 26 1.4.4.

2 Underwater Network 26 1.5 Security Objectives in Massive IoT in 6G 26 1.5.1 Confidentiality 26 1.5.2 Integrity 27 1.5.3 Availability 27 1.

5.4 Authentication 28 1.5.5 Authorization 28 1.6 Security Threats in Massive IoT in 6G 28 1.6.1 Security Threats to Data Confidentiality 29 1.6.

2 Security Threats to Data Integrity 29 1.6.3 Security Threats to Availability 30 1.6.4 Security Threats to Authentication 31 1.6.5 Security Threats to Authorization 31 1.7 Conclusion 32 References 33 2 Secure Edge Intelligence in the 6G Era 35 Tanesh Kumar, Juha Partala, Tri Nguyen, Lalita Agrawal, Ayan Mondal, Abhishek Kumar, Ijaz Ahmad, Ella Peltonen, Susanna Pirttikangas, and Erkki Harjula 2.

1 Introduction 35 2.2 Background 36 2.2.1 Edge Computing and Its Importance 36 2.2.2 Emergence of Edge Intelligence 38 2.2.3 Fusion of Edge Intelligence and 6G 39 2.

2.4 The Need for Secure Edge Intelligence 40 2.3 Security Challenges in 6G EI 40 2.3.1 Computational Offloading 40 2.3.2 Security of Machine Learning 41 2.3.

3 Post-quantum Cryptography 42 2.4 Privacy Challenges in 6G EI 42 2.5 Trust Challenges in 6G EI 44 2.6 Security Standardization for EI and 6G 46 2.7 Conclusion 47 References 47 3 Privacy-Preserving Machine Learning for Massive IoT Deployments 53 Najwa Aaraj, Abdelrahaman Aly, Alvaro Garcia-Banda, Chiara Marcolla, Victor Sucasas, and Ajith Suresh 3.1 Introduction 53 3.2 PPML for IoT 54 3.3 Secure Multiparty Computation 56 3.

3.1 MPC: Security Models and Setups 60 3.3.2 MPC: Privacy and Output Guarantees 61 3.3.3 MPC: Arithmetic 61 3.3.4 MPC: Preprocessing 63 3.

4 Fully Homomorphic Encryption 63 3.4.1 FHE: Bottlenecks and Advantages 64 3.4.2 Comparison with MPC 65 3.4.3 Generations of FHE Schemes 66 3.5 Oblivious Neural Networks (ONN) 67 3.

5.1 Two-Party (2PC) 67 3.5.2 Two-Party with Helper (2PC +) 67 3.5.3 Three-Party (3PC) and Four-Party (4PC) 69 3.5.4 n-Party (MPC) 69 3.

5.5 A Closer Look at PPML Frameworks 69 3.5.5.1 CryptoNets 69 3.5.5.2 SecureML 70 3.

5.5.3 MiniONN 70 3.5.5.4 DeepSecure 72 3.5.5.

5 Gazelle 73 3.5.5.6 Delphi 74 3.5.5.7 Aby 3 74 3.5.

5.8 Fantastic Four 77 3.5.5.9 Crypten 77 3.5.5.10 Fanng-mpc 78 3.

6 Decision Trees 78 3.6.1 Structure of a Decision Tree 79 3.6.2 Training 79 3.6.3 Computational Setups 80 3.6.

4 Inference 81 3.6.4.1 The Three Stages Model 81 3.6.5 Relevant Contributions 83 3.6.5.

1 On Passive Security 83 3.6.5.2 On Active Security 84 3.7 Software and Frameworks 85 3.8 Lightweight FHE and MPC for IoT 86 3.8.1 MPC Outsourcing 86 3.

8.2 Hybrid-FHE 87 3.9 Alternative Solutions 90 3.10 Conclusions 90 References 91 4 Federated Learning-Based Intrusion Detection Systems for Massive IoT 101 Filippos Pelekoudas-Oikonomou, Parya H. Mirzaee, Waleed Hathal, Georgios Mantas, Jonathan Rodriguez, Haitham Cruickshank, and Zhili Sun 4.1 Introduction 101 4.2 Intrusion Detection Systems (IDSs) in IoT 103 4.2.

1 Fundamentals of IDSs 103 4.2.2 Machine Learning Techniques for IDSs in IoT 104 4.2.3 Limitations of ML-Based IDSs in IoT 106 4.3 Federated Learning: A Decentralized ML Approach 107 4.3.1 Definition of Federated Learning 107 4.

3.2 Categories of Federated Learning 108 4.3.3 Federated Learning Architectures 108 4.3.3.1 Horizontal Federated Learning (HFL) 110 4.3.

3.2 Vertical Federated Learning (VFL) 111 4.3.3.3 Federated Transfer Learning (FTL) 111 4.4 Federated Learning-Based IDSs for IoT 113 4.4.1 FL-Based IDSs in IoT Use Cases 113 4.

4.1.1 FL-Based IDSs in Smart Homes 113 4.4.1.2 FL-Based IDSs in Industrial IoT 114 4.4.1.

3 FL-Based IDSs in Agricultural IoT 115 4.4.1.4 FL-Based IDS in Vehicular IoT Networks 116 4.4.2 Cross-layer FL for Lightweight IoT Privacy-Preserving IDS 116 4.5 Model Aggregation Approaches and Algorithms in FL 117 4.5.

1 Model Aggregation Approaches in FL 117 4.5.2 Model Aggregation Algorithms in FL 119 4.6 Challenges and Future Directions in FL-Based IDS for IoT 121 4.6.1 Validation and Standardization 121 4.6.2 Data Heterogeneity and Non-IID Data 121 4.

6.3 Security and Privacy Enhancements 121 4.6.4 Communication Efficiency and Scalability 122 4.6.5 Explainability and Interpretability 122 4.7 Conclusion 122 List of Abbreviations 123 References 124 5 Securing Massive IoT Using Network Slicing and Blockchain 129 Shihan Bao, Zhili Sun, and Haitham Cruickshank 5.1 Introduction 129 5.

2 Background 131 5.2.1 Massive IoT 131 5.2.2 Blockchain 132 5.2.2.1 Blockchain Applications 132 5.

2.2.2 Consensus Mechanisms 133 5.3 Challenges on Massive IoT 135 5.3.1 Security Requirements in Massive IoT 135 5.3.2 Privacy Requirements in Massive IoT 136 5.

3.3 Location Privacy Challenges 138 5.4 Securing IoT Using Network Slicing and Blockchain 138 5.4.1 Network Slicing Standardization 138 5.4.2 Network Slicing Security, Privacy, and Trust Threats for 5G and Beyond 142 5.4.

3 Blockchain-Enabled Secure Network Slicing 144 5.4.3.1 Integration of Distributed Ledger Technology (DLT) with Network Slicing 144 5.4.3.2 Blockchain-Powered Network Slicing in Verticals 144 5.4.

3.3 Multiple Participant Coordination and Trust Management by Blockchain for Network Slicing 145 5.5 Open Challenges 146 5.5.1 Edge Intelligence in Network Slicing 147 5.5.2 Post-quantum Security 147 5.5.

3 Blockchain Scalability 147 5.5.4 RAN Slicing 148 5.6 Conclusion 148 References 148 6 Physical Layer Security for RF-Based Massive IoT 155 Marcus de Ree, Seda Dogan-Tusha, Elmehdi Illi, Marwa Qaraqe, Saud Althunibat, Georgios Mantas, and Jonathan Rodriguez 6.1 Introduction 155 6.2 Physical Layer-Based Key Establishment 156 6.2.1 Introduction 156 6.

2.2 Channel Reciprocity-Based Key Establishment 157 6.2.2.1 Principles and Assumptions 157 6.2.2.2 Evaluation Metrics 159 6.

2.2.3 Key Establishment Model 159 6.2.3 Signal Source Indistinguishability-Based Key Establishment 161 6.2.3.1 Principles and Assumptions 161 6.

2.3.2 Evaluation Metrics 162 6.2.3.3 Key Establishment Model 162 6.2.3.

4 Security and Performance Evaluation 163 6.2.4 Final Remarks 164 6.3 Physical Layer-Based Node Authentication 164 6.3.1 Introduction 164 6.3.2 Key-Based Physical Layer Authentication 165 6.

3.2.1 Tag-Embedding Authentication 166 6.3.3 Channel-Based Keyless Physical Layer Authentication 166 6.3.3.1 RSS-Based Authentication 167 6.

3.3.2 Frequency-Based Authentication 167 6.3.3.3 CIR-Based Authentication 167 6.3.3.

4 Pilot-Based Authentication 168 6.3.3.5 Machine Learning-Based Authentication 168 6.3.4 Device-Based Keyless Physical Layer Authentication 168 6.3.4.

1 Radio Frequency Fingerprinting 168 6.3.4.2 Physically Unclonable Function 169 6.3.5 Final Remarks 170 6.4 Physical.