Preface
Key Features
About the Book
Contents
About the Editors
Part I: Architecture, Systems, and Services
Artificial Intelligence-based Internet of Things for Industry 5.0
1 Introduction
2 Industry 5.0 Paradigm
3 Elements of IoT
4 IoT Architecture
4.1 Perception Layer
4.2 Connectivity Layer
4.3 Edge or Fog Computing Layer
4.4 Processing Layer
4.5 Application Layer
4.6 Security Layer
5 Enabling Technologies
5.1 Radiofrequency Identification (RFID)
5.2 Power-Line Communication (PLC)
5.3 Electronic Product Code (EPC)
5.4 Wireless Sensor Network
5.5 Near-Field Communication
5.6 Actuator
5.7 Machine to Machine (M2M)
5.8 ZigBee
5.9 Wireless Fidelity (Wi-Fi)
5.10 IEEE 802.15.4
5.11 Z-Wave
5.12 Bluetooth LE
6 Artificial Intelligence (AI) in the Internet of Things (IoT)
6.1 Artificial Intelligence for Intelligent Sensing
6.2 Decision Tree in IoT
6.3 Random Forest in IoT
6.4 Clustering
6.5 One-Class Support Vector Machine (OC-SVM)
6.6 Ensemble Learning Models in IoT
6.7 Neural Networks
6.8 Support Vector Machine (SVM) in UIoT
6.9 Internet of Intelligent Things (IoIT) for Social Networks
6.10 Principal Component Analysis
6.11 Bagging
6.12 Artificial Intelligence in Analytical Skills (IoT)
6.13 Deep Learning for Analytics (IoT)
6.14 Edge Computing in IoT
6.15 Federated Learning
7 AI-Based Trustworthiness in IoT Systems
8 AI Tools for IoT
9 Applications of the Internet of Things
9.1 Agriculture
9.2 Augmented Reality
9.3 Virtual Reality
9.4 Mixed Reality
9.5 Smart Locks
9.6 Smart Factories
9.7 Intelligent Road Toll and Traffic Monitoring
9.8 Smart Intelligent Grid
9.9 Intelligent Robotics
9.10 Waste Management
9.11 Near-Field Communication (NFC) Payment
9.12 AI-Enabled Internet of Underwater Things
9.13 Intelligent UAV
9.14 IoT-Based Forensic Applications
9.15 Intelligent Healthcare Systems Using IoT Systems
9.16 Intelligent Disaster Management
9.17 Music
10 Consumer Electronic Products for IoT
11 Open Research Challenges for AI-Based IoT Systems
12 Conclusions
References
IoT Ecosystem: Functioning Framework, Hierarchy of Knowledge, and Intelligence
1 Introduction
2 Related Works and Motivation
3 IoT Architecture
3.1 Sensing or Perception Layer
3.2 Network Layer
3.3 Service Layer
3.4 Application Layer
4 Taxonomy of IoT
5 Core Elements
5.1 IoT Devices
5.2 IoT Gateways
5.3 Computation and OS
5.4 IoT Communication
5.5 IoT Middleware
6 IoT Knowledge Hierarchy
7 Paradigms of Intelligent IoT
7.1 Generalized Fog-Edge-Cloud-Enabled IoT
7.2 Machine Learning-Enabled IoT Intelligence
8 Applications of IoT Ecosystems
9 Conclusion
References
Artificial Neural Networks and Support Vector Machine for IoT
1 Introduction
2 Preliminaries
2.1 History of Neural Networks
2.2 Role of ANNs in Wireless Sensor Networks
2.3 ML Models in IoT
2.4 Artificial Neural Network Preliminaries and Types of ANN
2.5 Modifications of ANN
3 Applications of ANN in WSNs and IoT
3.1 UAVs for Wireless Networking Communications with ANN
3.2 Wireless Virtual Reality Over Wireless Networks with ANN
3.3 Coexistence of Manifold Radio Access Machineries with ANN
3.4 Internet of Things with ANN
4 Support Vector Machine for WSNs and IoTs
4.1 Support Vector Machine in Radio Access Frequency
4.2 SVM for Energy Harvesting
5 Smart Transportation with SVM, ANN, and DL
5.1 Artificial Neural Networks
5.2 Deep Learning
6 Anomaly Detection Analysis and Prediction
7 Challenges and Future Directions
8 Conclusion
References
The Role of Machine Learning Techniques in Internet of Things-Based Cloud Applications
1 Introduction
1.1 Machine Learning
1.1.1 Importance of Machine Learning in Present Business Scenario
1.1.2 Applications of Machine Learning
1.2 Internet of Things (IoTs) Technology and Infrastructure
1.3 Cloud Technology and Infrastructure
2 Evolution of Machine Learning Techniques
3 Motivation
4 Internet of Things and Cloud Applications
4.1 IoT Applications
4.2 Cloud Applications
5 Comparison of Scope of Machine Learning Techniques Pre-COVID-19 and Post-COVID-19 Era
6 Machine Learning Techniques for IoT-Based Cloud Applications
7 Internet of Things and Cloud Computing in Post-COVID-19 Era: Applications and Challenges
7.1 Challenges in Internet of Things (IoT)-Based Cloud Applications
7.2 Internet of Things Challenges and Solutions
7.3 Cloud Computing Challenges and Solutions
8 Future Research Directions for Machine Learning Towards IoT-Based Cloud Applications
8.1 Advanced Analytics with Machine Leaning: A Way to Forward
9 Conclusion
References
Further Reading
Deep Learning Frameworks for Internet of Things
1 Introduction
2 Architecture for Deep Neural Network
2.1 Convolution Neural Networks
2.2 Recurrent Neural Network
2.3 Autoencoders (AEs)
2.4 Generative Adversarial Networks (GANs)
3 Framework for Deep Neural Network
4 Deep Reinforcement Learning Approaches
5 Applications of Deep Learning in IoT Scenarios
6 Challenges and Future Research Directions
6.1 Challenges
6.2 Future Research Directions
7 Conclusion
References
Fog-Cloud Enabled Internet of Things Using Extended Classifier System (XCS)
1 Introduction
2 Literature Review
3 Architecture of LCS
4 Driving Mechanisms of LCS
5 Accuracy-Based Extended Learning Classifier System (XCS)
6 Case Study
7 Conclusion and Future Enhancements
References
Convolutional Neural Network (CNN)-Based Signature Verification via Cloud-Enabled Raspberry Pi System
1 Introduction
2 Literature Review
2.1 Background of Signature Verification
2.2 Convolutional Neural Network (CNN)
2.3 Standard Components of CNN
2.4 Activation Function
2.5 Rectified Linear Units (ReLU) Activation Function
2.6 CNN Against Traditional Image Classifier
2.7 CNN-Based Signature Verification Process
2.8 Writer-Dependent and Writer-Independent Signature
2.9 IoT and Machine Learning
2.10 Single-Board Computer (SBC)
2.11 Deep Learning on Raspberry Pi
2.12 Raspberry Pi-Enabled Cloud System
3 Methodology
3.1 Hardware and Environment Setup
3.2 Cloud Storage Setup
3.3 Pushbullet
3.4 Application Design
3.5 Data Handling
3.6 Signature Sample Acquisition
3.7 Image Augmentation
3.8 Convolutional Neural Network (CNN)
3.9 Training Result
4 Problems and Limitations
5 Conclusion
6 Future Research
References
Machine to Machine (M2M), Radio-frequency Identification (RFID), and Software-Defined Networking (SDN): Facilitators of the Internet of Things
1 Introduction
2 Facilitators of the IoT
3 Machine to Machine (M2M)
3.1 Autonomous Device Management System via M2M
3.2 Collaborative Infrastructure Formulated by M2M
3.2.1 M2M Device Domain
3.2.2 M2M Network Area Domain
3.2.3 M2M User or Admin Domain
3.3 Significance of M2M in the IoT
4 Radio-frequency Identification (RFID)
4.1 Framework of RFID
4.2 Significance of RFID in the IoT
5 Software-Defined Networking (SDN)
5.1 Three-Layered SDN Architecture
5.2 SDN Application Plane
5.3 SDN Control Plane
5.4 SDN Data Plane
5.5 Significance of SDN in the IoT
5.6 Central and Standardized Policy
5.7 Single Plane of Glass
5.8 Native Integration
5.9 Visibility and Analytics
5.10 Data Integration
5.11 Rapid Cloud Adoption
6 Issues and Challenges of the IoT
6.1 Connectivity
6.2 Security
6.3 Power Management
6.4 Complexity
6.5 Rapid Evolution
7 Conclusion
References
Architecture, Generative Model, and Deep Reinforcement Learning for IoT Applications: Deep Learning Perspective
1 Introduction
2 Evolution of Deep Learning Techniques
3 Motivation
4 Scope of Machine Learning and Deep Learning During COVID-19 and Post COVID-19 Era
4.1 Use of Machine Learning and Deep Learning to Fight Coronavirus
4.2 Role of Artificial Intelligence with the Internet of Things During COVID Situation
4.3 Artificial Intelligence with IoT Post COVID Outbreak
5 Architecture of Deep Learning Frameworks for Various Applications (e.g., Healthcare, Transportation, etc.)
5.1 Convolutional Neural Network (CNN)
5.1.1 CNN-Based Architectures
5.2 Pretrained Unsupervised Networks
5.3 Recurrent Neural Networks (RNNS)
6 Generative Models
6.1 Generative Adversarial Networks (GANs)
7 Deep Learning Applications in the Internet of Things
7.1 Foundation Service of the Internet of Things That Is Used by Deep Learning
8 Possibilities with Deep Learning in IoT-Based Applications
8.1 Possibilities with Deep Learning
8.1.1 Others
9 Deep Learning for IoT-Based Applications: Opportunities and Challenges
10 Future Research Gaps in Deep Learning-Based IoT Environment for Twenty-First-Century Generation
11 Conclusion
Bibliography
Enabling Inference and Training of Deep Learning Models for AI Applications on IoT Edge Devices
1 Introduction
2 Inference at the Edge
2.1 Model Pruning
2.2 Model Quantization
2.3 Knowledge Distillation
3 Training Models at the Edge
3.1 Federated Learning
3.2 Adaptive Retraining
4 Conclusion
References
Nonvolatile Memory-Based Internet of Things: A Survey
1 Introduction
2 Memory Overview
3 State of the Art Techniques
3.1 Flash Memory
3.2 MRAM
3.3 ReRAM
3.4 FeRAM
3.5 SCM
3.6 STTRAM and TCAM
3.7 Hybrid
3.8 Others
4 Conclusion and Future Outlook
References
Integration of AI and IoT Approaches for Evaluating Cybersecurity Risk on Smart City
1 Introduction
2 Background
2.1 Smart City
2.2 Smart City and Cybersecurity
3 Risk Analysis
3.1 Systematic Risk
3.2 Bayesian Network for Risk Analysis
3.3 Digital Twin
4 Modeling Smart City
4.1 Experiment
5 Conclusions
References
Cognitive Internet of Things: Challenges and Solutions
1 Introduction
2 Internet of Things, Artificial Intelligence, and Cognitive Systems: Definitions, Combinations, and Challenges
2.1 Internet of Things
2.2 Artificial Intelligence
2.3 Cognitive Systems
2.4 Artificial Intelligence and Cognitive Systems
2.5 Artificial Intelligence and the Internet of Things
2.6 Cognitive Systems and Internet of Things
2.6.1 Priority of Cognitive Internet of Things
3 Challenges and Solutions in Designing Cognitive Internet of Things
3.1 Problem Identification and Formulation
3.2 Energy Consumption and Environmental Pollution
3.3 Explainable Systems
3.4 Layered and Flat Cooperation Methods
3.5 Architecture Design
3.6 Safety
3.7 Morality, Privacy, and Responsibility
3.8 Inspired Models
3.9 Predictability and Controllability
3.10 Security and Trust
3.11 Semantic and Communication
3.12 Machine Selection
4 Conclusions
References
Part II: Applications
An AI Approach to Rebalance Bike-Sharing Systems with Adaptive User Incentive
1 Introduction
2 Problem Statement
2.1 Overview
2.2 Incentive Scheme Model
2.3 Environment Model
2.4 An Existing Pricing Scheme for Source Incentive
2.5 Problem Formulation
3 Hybrid Incentive Scheme
3.1 Benefits of Destination Incentive
3.2 A Hybrid Incentive Scheme
3.3 Adaptively Adjust Source and Destination Incentive
3.4 Properties
4 Hybrid Incentive in Docked BSS
5 Experiment
5.1 Dataset
5.2 Experiment Setup
5.3 Results
6 Related Work
7 Conclusion
References
IoT-Driven Bayesian Learning: A Case Study of Reducing Road Accidents of Commercial Vehicles on Highways
1 Introduction
2 IoT Strategies and Implementations
2.1 Machine Learning (ML)
2.2 Transportation and Sensor Technologies
2.3 Traffic Models
2.3.1 Collision Avoidance Model (CAM)
2.3.2 The Action Point Model (APM)
2.3.3 Intelligent Driver (ID) Model
2.3.4 Latent Class (LC) Model
2.4 Review of Relevant Developments
3 Case Study
3.1 Materials
3.2 Description of the Case Study
3.3 Design Approach
3.4 Data Preprocessing and Feature Selection
3.5 Bayesian Model Building and Testing
4 Results
4.1 Bayesian Model
4.2 The Remote Server and Its Operation
5 Conclusion
References
On the Integration of AI and IoT Systems: A Case Study of Airport Smart Parking
1 Introduction
2 Related Work
3 Airport Parking Problem and Its Requirement Analysis
4 System Design
4.1 Design Considerations
4.1.1 Energy Consumption
4.1.2 Network Technology
4.1.3 Data Volume, Processing Time and Location
4.1.4 Weather and Environment
4.2 Design Architectures
4.2.1 Design Architecture for Generic IoT Applications
4.2.2 Proposed Architecture for the Airport Smart Parking
4.3 Vehicle Identification Framework
5 Experiment Results and Discussion
5.1 Energy Consumption
5.2 Detection Accuracy
5.3 Processing Time
5.4 Discussion
6 Conclusion and Future Avenue
References
Vision-Based End-to-End Deep Learning for Autonomous Driving in Next-Generation IoT Systems
1 Introduction
2 Ethics of Autonomous Driving in IoAT
3 Autonomous Driving Paradigms in IoAT
3.1 Modular Pipeline Paradigm
3.2 End-to-End Learning Paradigm
4 Proposed Model for Autonomous Driving in IoAT
5 Analysis of Deep Learning Models for Autonomous Driving
5.1 The Autonomous Driving Paradigm
5.2 The Perception Source
5.3 The Training Method
5.4 Controllable Routing
5.5 Exploring Temporal Information
5.6 Image Perception Network
5.7 Flexibility
6 Conclusion and Future Research Directions
References
A Study on the Application of Bayesian Learning and Decision Trees IoT-Enabled System in Postharvest Storage
1 Introduction
2 Technology and Postharvest Management of Vegetables and Fruits
2.1 Environmental (Abiotic) Factors Affecting Fruit and Vegetable Preservation
2.1.1 Temperature
2.1.2 Relative Humidity
2.1.3 pH
2.1.4 Nitrogen Oxide (NO)
2.1.5 Chlorine Dioxide (ClO2)
2.1.6 Carbon (IV) Oxide (CO2)
2.1.7 Oxygen (O2)
2.1.8 Arachidonic Acid
2.1.9 Y-Aminobutyric Acid (GABA)
2.2 IoT as a Machinery for Relaying Data from Postharvest Storage Sites
2.3 IoT Requirements for Postharvest Vegetable and Fruit Storage Instrumentation
2.4 Supervised Learning
2.5 Control of Storage Processes Using Classification Models
3 Methodology for the Design of an Intelligent IoT-Driven Postharvest Fruit Storage Facility
3.1 Microcontroller Interfacing and Specification
3.2 Data Specification and Model Development
4 Results
4.1 DT and NBC Modeling Results
4.2 Proposed Implementation
5 Conclusion
References
Index