Learning Techniques for the Internet of Things

Preface
Acknowledgements
Contents
Editors and Contributors
	About the Editors
	Contributors
1 Edge Computing for IoT
	1.1 Introduction
	1.2 Computing Paradigms for IoT
		1.2.1 Cloud Computing
		1.2.2 Edge Computing
		1.2.3 Fog Computing
	1.3 Edge Computing Paradigms
		1.3.1 Cloudlet
		1.3.2 Mobile Edge Computing
	1.4 Architecture of Edge Computing-Based IoT
	1.5 Advantages of Edge Computing-Based IoT
	1.6 Enabling Edge Computing-Based IoT Technologies
		1.6.1 Edge Intelligence
		1.6.2 Lightweight Virtualization
	1.7 Edge Computing in IoT-Based Intelligent Systems: Case Studies
		1.7.1 Edge Computing in IoT-Based Healthcare
		1.7.2 Edge Computing in IoT-Based Manufacturing
		1.7.3 Edge Computing in IoT-Based Agricultural
		1.7.4 Edge Computing in IoT-Based Transportation
	1.8 Challenges and Future Research Directions
	1.9 Conclusion
	References
2 Federated Learning Systems: Mathematical Modelling and Internet of Things
	2.1 Introduction
	2.2 Federated Learning
		2.2.1 Definition of Federated Learning
		2.2.2 The Different Forms of Federated Learning
	2.3 Mathematical Modeling
		2.3.1 Architecture
	2.4 Internet of Things
		2.4.1 Introduction
		2.4.2 Link Between IoT and Federated Learning
	2.5 Conclusion
	References
3 Federated Learning for Internet of Things
	3.1 Introduction
	3.2 Federated Learning and Internet of Things: Preliminaries
		3.2.1 Federated Learning
			3.2.1.1 Fundamental FL Concept
			3.2.1.2 The Typical Process of FL Training for IoT
			3.2.1.3 The Architecture of Federated Learning for IoT Networks
		3.2.2 Types of Federated Learning for IoT
			3.2.2.1 Types of FL for IoT Based on Networking Structure
			3.2.2.2 Types of Centralized Federated Learning
			3.2.2.3 Types of Federated Learning for IoT Based on Participating Clients
		3.2.3 FL Framework for IoT
	3.3 Federated Learning for IoT Applications
		3.3.1 FL for Smart Healthcare
		3.3.2 FL for Vehicular IoT
		3.3.3 FL for Smart City
		3.3.4 FL for Smart Industry
		3.3.5 FL for Cybersecurity
	3.4 Research Challenges and Directions
		3.4.1 Heterogeneity of IoT Devices
		3.4.2 Limited Computational Resources
		3.4.3 Communication and Bandwidth Limitations
		3.4.4 Privacy and Security Concerns
		3.4.5 Scalability and Management
		3.4.6 Federated Domain Generalization
	3.5 Conclusion
	References
4 Machine Learning Techniques for Industrial Internet of Things
	4.1 Introduction
		4.1.1 Evolution of IoT to IIoT
		4.1.2 Significance of ML in IIoT
		4.1.3 Computational Offloading in ML for IIoT Application
		4.1.4 Objective of This Chapter
		4.1.5 Contributions
	4.2 Fundamental Concept of Machine Learning
		4.2.1 Key Machine Learning Technique for IIoT
		4.2.2 Experiment Analysis of Machine Learning Methods
		4.2.3 State-of-the-Art Research Initiatives
			4.2.3.1 Supervised Learning
			4.2.3.2 Unsupervised Learning
			4.2.3.3 Reinforcement Learning
	4.3 Machine Learning in IIoT Applications
		4.3.1 Predictive Maintenance
		4.3.2 Smart Healthcare
		4.3.3 Smart Manufacturing
		4.3.4 Supply Chain Optimization
		4.3.5 Ultralow Latency Data Transmission
	4.4 Challenges and Future Research Opportunities
		4.4.1 Data Collection and Quality
		4.4.2 Interoperability
		4.4.3 Real-Time Processing
	4.5 Conclusion
	References
5 Exploring IoT Communication Technologies and Data-Driven Solutions
	5.1 Introduction
		5.1.1 Evolution of Communication Protocol
		5.1.2 Standard IoT Architecture
		5.1.3 Data-Driven Technologies for IoT
		5.1.4 Features of IoT Communication Protocols
			5.1.4.1 Low-Power Consumption
			5.1.4.2 Scalability
			5.1.4.3 Security
			5.1.4.4 Interoperability and Standardization
		5.1.5 Contributions
	5.2 Classification of Communication Protocols
		5.2.1 Overview of Short-Range IoT Communication Technologies
			5.2.1.1 Bluetooth
			5.2.1.2 Wi-Fi
			5.2.1.3 Zigbee
			5.2.1.4 RFID
		5.2.2 Overview of Long-Range IoT Communication Protocols
			5.2.2.1 LoRaWAN
			5.2.2.2 NB-IoT
			5.2.2.3 Sigfox
			5.2.2.4 LTE-M
		5.2.3 Literature
	5.3 Emerging Use Cases of IoT
		5.3.1 Industry 5.0
		5.3.2 Smart Healthcare
		5.3.3 Smart Agriculture
		5.3.4 Intelligent Transportation System
	5.4 Challenges and Future Opportunities
		5.4.1 Interoperability
		5.4.2 Energy-Optimized Data Transmission
		5.4.3 Zero-Touch IoT Automation
		5.4.4 Security and Trust
		5.4.5 Scalability
	5.5 Conclusion
	References
6 Towards Large-Scale IoT Deployments in Smart Cities: Requirements and Challenges
	6.1 Introduction
	6.2 Requirements for IoT Deployment in Smart Cities
		6.2.1 Reliable Network Connection
		6.2.2 Infrastructure Deployment
	6.3 Key Aspects of Sensor Deployment and Data Management in Smart Cities
		6.3.1 Sensor Deployment and Placement
		6.3.2 Data Collection
		6.3.3 Data Transmission
		6.3.4 Data Services
		6.3.5 Data Quality
	6.4 Case Study: Air Quality Monitoring with IoT for Smart Cities
		6.4.1 IoT Installation
		6.4.2 Air Quality IoT Monitoring for a Smart City
	6.5 Role of AI and Emerging Technologies in Future Smart Cities
	6.6 Conclusion
	References
7 Digital Twin and IoT for Smart City Monitoring
	7.1 Introduction
		7.1.1 Background and Related Works
		7.1.2 Research Gap and Motivation
		7.1.3 Contributions
	7.2 Proposed System Model
		7.2.1 Twin Time Step
		7.2.2 Twin Reward Function
		7.2.3 Twin Representations
		7.2.4 Twin-State Model
		7.2.5 Twin Message Transmission
		7.2.6 Twin Communications
		7.2.7 Twin Transmission Delay
		7.2.8 Objective Function
	7.3 Twin Protocol Integration
		7.3.1 Optimization Algorithm
	7.4 Results and Discussions
		7.4.1 Discussions
			7.4.1.1 Scenario 1: Analysis of State Model
			7.4.1.2 Scenario 2: Twin Communications
			7.4.1.3 Scenario 3: Monitoring Inactive Twins
			7.4.1.4 Scenario 4: Success Rate
			7.4.1.5 Scenario 5: Number of Message Transmissions
	7.5 Conclusion
	References
8 Multi-Objective and Constrained Reinforcement Learning for IoT
	8.1 Introduction
	8.2 Objectives and Problems in IoT Networks
	8.3 Multi-Objective Optimization
		8.3.1 Pareto Front
		8.3.2 Preference Vector
		8.3.3 Traditional Approaches for MOO in IoT
	8.4 Reinforcement Learning
	8.5 Multi-Objective and Constrained Reinforcement Learning in IoT Networks
		8.5.1 Single-Policy Approaches
		8.5.2 Multiple-Policy Approaches
		8.5.3 Approaches Based on Dynamic Preferences
	8.6 Future Scope and Challenges in MORL
	8.7 Conclusion
	References
9 Intelligence Inference on IoT Devices
	9.1 Introduction
	9.2 Inference on IoT Devices: Preliminaries
	9.3 Promising Intelligence Applications
		9.3.1 Real-Time Video Analytic
		9.3.2 Autonomous Driving
		9.3.3 Smart Manufacturing
		9.3.4 Smart City and Home
	9.4 Commodity Hardware for IoT Devices
	9.5 Model Optimization for IoT Devices
		9.5.1 Lightweight Model Design
		9.5.2 Model Pruning
		9.5.3 Model Quantization
		9.5.4 Knowledge Distillation
	9.6 Inference Library for IoT Devices
	9.7 Inference Systems for IoT Devices
		9.7.1 Edge Cache-Based Inference
		9.7.2 Computing Offloading-Based Inference
	9.8 Challenges and Opportunities of Inference
	9.9 Conclusion
	References
10 Applications of Deep Learning Models in Diverse Streams of IoT
	10.1 Introduction
		10.1.1 Internet of Things
		10.1.2 Automation
		10.1.3 Deep Learning
		10.1.4 The Synergy
	10.2 Applications of DL in IoT Paradigms
		10.2.1 Data Analysis
			10.2.1.1 Overview of Data Analysis in IoT
			10.2.1.2 DL Techniques for IoT Data Analysis
			10.2.1.3 Predictive Analytics in IoT
			10.2.1.4 Data Mining and Pattern Recognition
			10.2.1.5 Visualisation and Interpretability of IoT Data
			10.2.1.6 Case Studies and Applications
		10.2.2 Smart Cities and Development
			10.2.2.1 DL Techniques for IOT in Smart Cities\' Development
		10.2.3 Home Automation
			10.2.3.1 Overview of IOT in Home Automation
			10.2.3.2 Challenges and Opportunities for IOT in Home Automation
			10.2.3.3 DL Techniques for IOT in Home Automation
		10.2.4 Energy-Efficient IoT
		10.2.5 Malware Detection
			10.2.5.1 Overview of Malware Detection in IOT
			10.2.5.2 DL Technique for Malware Detection in IOT
		10.2.6 DL for IOT Healthcare and Telemedicine
			10.2.6.1 Overview of DL in Healthcare and Telemedicine
			10.2.6.2 DL Techniques for IOT in Healthcare and Telemedicine
		10.2.7 Security and Privacy
			10.2.7.1 The Significance of Security and Privacy in IoT
			10.2.7.2 DL for IoT Security
			10.2.7.3 DL-Based Intrusion Detection
			10.2.7.4 DL for Privacy Preservation in IoT
			10.2.7.5  DL for Authentication and Access Control in the IoT
			10.2.7.6 Secure Communication in IoT Using DL
		10.2.8 Transportation and Autonomous Vehicles
			10.2.8.1 Intelligent Transportation Systems and DL
			10.2.8.2 DL for Transportation Object Identification and Recognition
			10.2.8.3 Vehicle Localisation and Mapping Based on DL
			10.2.8.4 DL for Predictive Behaviour and Trajectory Planning
			10.2.8.5 DL for ADAS
			10.2.8.6 DL for Autonomous Vehicle Control
		10.2.9 Environmental Monitoring and Conservation
			10.2.9.1 Environmental Monitoring with DL
			10.2.9.2 Biodiversity Conservation
			10.2.9.3 Water Resource Management
		10.2.10 Industrial Internet of Things
			10.2.10.1 Foundations of IIoT and DL
			10.2.10.2 Applications of DL in IIoT
			10.2.10.3 Energy Optimisation
			10.2.10.4 Challenges and Future Perspectives
	10.3 Conclusion
	References
11 Quantum Key Distribution in Internet of Things
	11.1 Introduction
		11.1.1 Cryptography and Involvement of Quantum Physics
		11.1.2 Security in IOT
	11.2 Fundamentals of Quantum Key Distribution
		11.2.1 Quantum and Classical Channels
		11.2.2 Quantum Phenomena and Security in QKD
		11.2.3 Light as a Medium
	11.3 BB84 Protocol
		11.3.1 Introduction
		11.3.2 Polarization
		11.3.3 QKD Procedure
		11.3.4 Eavesdropping
			11.3.4.1 Information Gain, Error Rate, Key Length
			11.3.4.2 Selective Intercept-Resend Attack
	11.4 Generic QKD Protocols
		11.4.1 Classical and Quantum Channels
		11.4.2 Processing Schemes
		11.4.3 Classical Processing
		11.4.4 Secret Key Rate
	11.5 Types of Protocols
		11.5.1 Discrete-Variable Coding: The Pioneering Approach
		11.5.2 Continuous-Variable Protocols
		11.5.3 Distribute-Phase-Reference Protocols
			11.5.3.1 Differential-Phase-Shift (DPS) Protocol
			11.5.3.2 Coherent-One-Way (COW)
	11.6 Sources
		11.6.1 Lasers
		11.6.2  Sub-Poissonian Sources
		11.6.3 Sources of Entangled Photons
	11.7 Hacking in QKD
		11.7.1 Trojan Horse Attack
		11.7.2 Other Hacking Attacks
			11.7.2.1 Faked State Attacks
			11.7.2.2 Phase-Remapping Attacks
			11.7.2.3 Time-Shift Attacks
	11.8 The ``Uncalibrated-Device Scenario\'\'
	11.9 Conclusion
	References
12 Quantum Internet of Things for Smart Healthcare
	12.1 Introduction
	12.2 Quantum IoT: Fundamentals and Components
		12.2.1 Quantum Computing and Its Relevance to Healthcare
		12.2.2 Quantum Communication for Secured Healthcare Data Transmission
		12.2.3 Quantum Sensing and Imaging in Healthcare Applications
		12.2.4 Integration with Traditional IoT in Healthcare
	12.3 Smart Healthcare Applications of Quantum IoT
		12.3.1 Quantum IoT in Diagnostics and Imaging
		12.3.2 Quantum IoT for Drug Discovery and Development
		12.3.3 Quantum IoT-Enabled Wearable Health Monitoring Devices
		12.3.4 Quantum-Enhanced Telemedicine and Remote Healthcare
	12.4 Advantages and Challenges of Quantum IoT in Smart Healthcare
		12.4.1 Advantages of Quantum IoT for Healthcare Applications
		12.4.2 Security and Privacy Considerations in Quantum IoT
		12.4.3 Technological and Implementation Challenges
		12.4.4 Regulatory and Ethical Implications
	12.5 Current Advances and Case Studies
		12.5.1 Research Initiatives and Collaborations
		12.5.2 Case Studies of Quantum IoT Applications in Healthcare
			12.5.2.1 Quantum Encryption for Secure Medical Data Transmission
			12.5.2.2 Quantum-Enhanced Imaging for Improved Diagnostics
			12.5.2.3 Quantum Algorithms for Drug Discovery
		12.5.3 Implementations and Real-World Deployments
	12.6 Future Directions and Emerging Trends
		12.6.1 Roadmap for Quantum IoT in Smart Healthcare
		12.6.2 Potential Impact on the Healthcare Industry
		12.6.3 Opportunities for Further Research and Development
	12.7 Conclusion
	References
13 Enhancing Security in Intelligent Transport Systems: A Blockchain-Based Approach for IoT Data Management
	13.1 Introduction
		13.1.1 Problem
		13.1.2 Motivation
		13.1.3 Outline
	13.2 Background
		13.2.1 Intelligent Transport System (ITS)
		13.2.2 Edge, Fog, and Cloud Computing
		13.2.3 Blockchain
		13.2.4 Hyperledger Fabric (HLF)
		13.2.5 Corda
		13.2.6 Hyperledger Fabric vs Corda
		13.2.7 Why Hyperledger Fabric?
	13.3 E2C-Block in ITS Usecase
		13.3.1 Intelligent Transport System (ITS)
		13.3.2 Fog Blockchain Network
		13.3.3 Cloud Blockchain Network
		13.3.4 Offshore Data Store
	13.4 Implementation of E2C-Block in ITS
		13.4.1 Registration and Authentication in ITS
		13.4.2 Fog Blockchain Network
		13.4.3 Cloud Blockchain Network
		13.4.4 Offshore Data Repository
		13.4.5 How Is Stored Data Queried?
		13.4.6 E2C-Block Deployment
	13.5 Experiments
		13.5.1 Experiment Setup
			13.5.1.1 Benchmarking Tool
			13.5.1.2 Network Configuration
			13.5.1.3 Workload Generation
			13.5.1.4 Hardware and Software Specification
		13.5.2 Performance Metrics
		13.5.3 Impact of Block Size
		13.5.4 Impact of Transaction Rates
		13.5.5 Impact of Number of Participating Peers
	13.6 Conclusion
	References
Index