Deep Learning Techniques for IoT Security and Privacy (Studies in Computational Intelligence, 997)

✍ Scribed by Mohamed Abdel-Basset, Nour Moustafa, Hossam Hawash, Weiping Ding

• Publisher: Springer
• Year: 2021
• Tongue: English
• Leaves: 273
• Edition: 1st ed. 2022
• Category: Library

✦ Synopsis

This book states that the major aim audience are people who have some familiarity with Internet of things (IoT) but interested to get a comprehensive interpretation of the role of deep Learning in maintaining the security and privacy of IoT. A reader should be friendly with Python and the basics of machine learning and deep learning. Interpretation of statistics and probability theory will be a plus but is not certainly vital for identifying most of the book's material.

✦ Table of Contents

Preface I
Preface II
Acknowledgements
About This Book
Who This Book is for?
What This Book Covers
Contents
1 Introduction Conceptualization of Security, Forensics, and Privacy of Internet of Things: An Artificial Intelligence Perspective
1.1 Cyber-Physical Internet of Things
1.2 Security of Internet of Things
1.3 Forensics of Internet of Things
1.4 Artificial Intelligence for Secure Internet of Things
1.4.1 Machine Learning
1.4.2 Deep Learning
1.5 Taxonomy of Deep Learning Approaches
1.5.1 Deep Supervised Learning
1.5.2 Deep Unsupervised Learning
1.5.3 Deep Weakly Supervised Learning
1.5.4 Deep Reinforcement Learning
1.6 Privacy in Internet of Things
1.7 Summary and Learnt Lessons
1.8 Book Organization
References
2 Internet of Things, Preliminaries and Foundations
2.1 The Architecture of IoT System
2.1.1 Perception/Physical Layer
2.1.2 Connectivity/Network Layer
2.1.3 Fog or Edge Layer
2.1.4 Middleware Layer
2.1.5 Applications Layer
2.1.6 Business Service Layer
2.1.7 Security Layer
2.2 The Cloud Computing Based IoT System
2.3 The Fog Computing Based IoT System
2.4 The Edge Computing Based IoT System
2.5 Summary and Learnt Lessons
References
3 Internet of Things Security Requirements, Threats, Attacks, and Countermeasures
3.1 Security Requirements in Internet of Things
3.1.1 Authentication (S1)
3.1.2 Authorization (S2)
3.1.3 Availability (S3)
3.1.4 Confidentiality (S4)
3.1.5 Data Security (S5)
3.1.6 Integrity (S6)
3.1.7 Non-repudiation (S7)
3.1.8 Network Security (S8)
3.1.9 Maintainability (S9)
3.1.10 Resilience (S10)
3.1.11 Security Monitoring (S11)
3.2 IoT Threats, Attacks, Vulnerabilities, and Risks
3.2.1 IoT Threats
3.2.2 IoT Vulnerabilities
3.2.3 IoT Risks
3.3 Today’s IoT Attacks and Countermeasures
3.3.1 Physical IoT Attacks (Hardware-Based)
3.3.2 Software-Related IoT Attacks
3.3.3 Data-Related IoT Attacks
3.4 IoT Attack Surfaces
3.4.1 Physical Surface
3.4.2 Network Surface
3.4.3 Cloud Attack Surface
3.4.4 Application Surface
3.5 Summary and Learnt Lessons
References
4 Digital Forensics in Internet of Things
4.1 What Is Digital Forensic?
4.2 Digital Evidence
4.2.1 Digital Forensics and Other Related Disciplines
4.2.2 A Brief History of Digital Forensics
4.2.3 Common Sources of Digital Evidence
4.2.4 Understanding Law Enforcement Agency Investigations
4.3 Major Areas of Investigation for Digital Forensics
4.4 Following Legal Processes
4.5 Types of Digital Evidence
4.6 The Cyber Kill Chain
4.7 IoT Forensics
4.8 Summary and Learnt Lessons
References
5 Supervised Deep Learning for Secure Internet of Things
5.1 Convolutional Neural Network
5.1.1 Convolutional Layer
5.1.2 Pooling Layer
5.1.3 Fully Connected Layer
5.1.4 Feature Map and Receptive Field
5.2 Advanced Convolutional Networks
5.2.1 VGG Network
5.2.2 Residual Network
5.2.3 Dense Network
5.3 Temporal Convolutional Network
5.4 Recurrent Neural Networks
5.4.1 Vanilla Recurrent Neural Networks
5.4.2 Longest Short-Term Memory (LSTM)
5.4.3 Gated Recurrent Units
5.5 Graph Neural Networks
5.6 Supervised Datasets and Evaluation Measures
5.6.1 Datasets
5.6.2 Evaluation Metrics
5.7 Taxonomy of Deep Learning Solutions for IoT
5.7.1 Input Scheme
5.7.2 Detection Scheme
5.7.3 Deployment Scheme
5.7.4 Evaluation Scheme
5.8 Summary and Learnt Lessons
References
6 Unsupervised Deep Learning for Secure Internet of Things
6.1 Generative Adversarial Networks
6.2 Autoencoders
6.2.1 Sparse Auto Encoder (SAE)
6.2.2 Denoising Auto Encoder (DAE)
6.2.3 Variational Auto Encoder (VAE)
6.3 Energy-Based Models
6.3.1 Boltzmann Machine (BM)
6.3.2 Restricted Boltzmann Machine (RBM)
6.3.3 Deep Boltzmann Machine (DBM)
6.3.4 Deep Belief Network
6.4 Summary and Learnt Lessons
References
7 Semi-supervised Deep Learning for Secure Internet of Things
7.1 Background and Foundations
7.1.1 Semi-supervised Learning Assumptions
7.1.2 Related Theories
7.1.3 Taxonomization
7.2 Consistency Regularization Approaches
7.2.1 Ladder Network
7.2.2 П-Model
7.2.3 Temporal Ensembling
7.2.4 Mean Teacher
7.2.5 Dual Student
7.3 Semi-supervised Generative Approaches
7.3.1 Categorical Generative Adversarial Network (CatGAN)
7.3.2 Context-Conditional Generative Adversarial Networks (CCGAN)
7.3.3 GoodBadGAN
7.4 Semi-supervised Autoencoder Approaches
7.4.1 Semi-supervised VAEs (SSVAEs)
7.4.2 Infinite VAE
7.4.3 Disentangled Variational Autoencoder
7.5 Semi-supervised Graph-Based Approaches
7.5.1 Baseline GNN
7.5.2 Graph Convolutional Network (GCN)
7.5.3 Graph Attention Network (GAT)
7.6 Pseudo-Labeling Approaches
7.6.1 Disagreement-Centered Methods
7.6.2 Self-Training Methods
7.7 Hybrid Approaches
7.7.1 Interpolation Consistency Training (ICT)
7.7.2 MixMatch
7.7.3 ReMixMatch
7.7.4 DivideMix
7.7.5 FixMatch
7.8 Summary and Learnt Lessons
References
8 Deep Reinforcement Learning for Secure Internet of Things
8.1 Foundations and Preliminaries
8.2 Single-Agent Reinforcement Learning
8.2.1 Markov Decision Process
8.2.2 Partially Observable Markov Decision Process
8.3 Multi-agent Reinforcement Learning
8.3.1 Markov/Stochastic Games
8.3.2 Dec-POMDP
8.3.3 Networked Markov Games
8.4 Taxonomy of Deep Reinforcement Learning
8.4.1 Value-Based DRL
8.4.2 Policy-Based DRL
8.5 Reinforcement Learning Based IoT Applications
8.5.1 IoT-Based Industrial IoT
8.5.2 IoT-Based Intelligent Transportation
8.6 Summary and Learnt Lessons
References
9 Federated Learning for Privacy-Preserving Internet of Things
9.1 Definition of Federated Learning
9.1.1 Federated Training
9.2 Taxonomy of Federated Learning Solutions
9.2.1 Data Partition
9.2.2 Privacy-Preservation Techniques
9.2.3 Communication Architecture
9.2.4 Heterogeneity Handling
9.3 Summary and Learnt Lessons
References
10 Challenges, Opportunities, and Future Prospects
10.1 Internet of Things Security
10.2 Cloud Computing Based Security Solutions
10.3 Fog Computing Based Security Solutions
10.4 Edge Computing Based Security Solutions
10.5 Deep Learning for IoT Security
10.6 Deep Reinforcement Learning
10.7 Privacy-Preserving Federated Learning
References