Front Cover
Artificial Intelligence and Machine Learning for EDGE Computing
Copyright
Contents
Contributors
Preface
Part I: AI and machine learning
Chapter 1: Supervised learning
1. Introduction
2. Perceptron
3. Linear regression
3.1. Training a linear regression
3.2. Steepest descent
3.3. Conjugate gradient
4. Logistic regression
4.1. Softmax classifier
5. Multilayer perceptron
5.1. Structure and notation
5.2. Initialization
5.2.1. Input means and standard deviations
5.2.2. Randomizing the input weights
5.3. First-order learning algorithms
5.3.1. Backpropagation algorithm
5.3.2. Training lemmas
Implications
Implications
Implications
5.4. Second-order learning algorithms
5.4.1. Newtons method
5.4.2. LM algorithm
6. KL divergence
7. Generalized linear models
8. Kernel method
9. Nonlinear SVM classifier
10. Tree ensembles
10.1. Decision trees
10.2. Random forest
10.3. Boosting
10.3.1. AdaBoosting
10.3.2. Gradient boosting
References
Chapter 2: Supervised learning: From theory to applications
1. Introduction
1.1. Supervised learning
1.2. Unsupervised learning
2. What are regression and classification problems?
3. Learning algorithms
3.1. Linear regression
3.2. Logistic regression
3.3. Decision tree
4. Evaluation metrics
4.1. Mean square error
4.2. Root mean square error
4.3. Confusion matrix
4.4. Accuracy
4.5. Recall
4.6. Precision
4.7. F1 score
5. Supervised learning to detect fraudulent credit card transactions
5.1. Data exploration
5.2. Data preprocessing
5.3. Fitting and evaluation
6. Supervised learning for hand writing recognition
7. Conclusion
References
Chapter 3: Unsupervised learning
1. Introduction
2. k-means clustering
3. k-means++ clustering
4. Sequential leader clustering
5. EM algorithm
6. Gaussian mixture model
7. Autoencoders
7.1. AEs: Structure, notations, and training
7.2. Variants of AEs
7.2.1. SAE
7.2.2. DAE
8. Principal component analysis
8.1. Generic PCA derivation
8.2. Advantages of PCA
8.3. Assumptions behind PCA
8.4. Comments
9. Linear discriminant analysis
9.1. Algorithm
9.2. Derivation of LDA algorithm
9.3. PCA vs. LDA
9.4. Comments and programmers perspective
10. Independent component analysis
10.1. Limitations of ICA
10.2. Assumptions
References
Chapter 4: Regression analysis
1. Introduction
2. Linear regression
3. Cost functions
3.1. MSE
3.2. MSLE
3.3. RMSE
3.4. MAE
4. Gradient descent
Example
5. Polynomial regression
6. Regularization
6.1. Ridge regression
6.2. Lasso regression
6.3. Dropout
6.4. Early stopping
7. Evaluating a machine learning model
7.1. Bias-variance trade-off
Bias error
Variance error
7.2. R-squared
7.3. Adjusted R-squared
References
Chapter 5: The integrity of machine learning algorithms against software defect prediction
1. Introduction
2. Related works
3. Proposed method
3.1. Overview
3.2. KMFOS
3.2.1. K-means clustering
3.2.2. Oversampling
3.2.3. Noise filtering
3.3. Dataset
4. Experiment
4.1. Design
4.2. Evaluation metrics
5. Results
5.1. Hyperparameters
5.2. Individual algorithms
5.3. PC4 dataset
5.4. PC3 dataset
5.5. KC1 dataset
6. Threats to validity
6.1. Threats to internal validity
6.2. Threats to external validity
7. Conclusions
References
Chapter 6: Learning in sequential decision-making under uncertainty
1. Introduction
2. Multiarmed bandit problem
2.1. Applications
2.2. Algorithms for multiarmed bandit problem
2.2.1. -Greedy algorithms
2.2.2. Upper confidence bound algorithm
2.2.3. Thompson sampling
2.3. Nonstationary environment
3. Markov decision process planning problem
3.1. Multiarmed bandits and MDP planning problem
4. Reinforcement learning
4.1. RL and MDP planning problem
4.2. Model-free RL algorithms
4.2.1. Q-learning
4.2.2. SARSA
4.3. Model-based RL algorithms
4.4. RL in nonstationary environment
4.4.1. Change point detection algorithms
4.4.2. Repeated update Q-learning (RUQL)
4.4.3. Context Q-learning
4.4.4. Other approaches
5. Summary
References
Chapter 7: Geospatial crime analysis and forecasting with machine learning techniques
1. Introduction
2. Related work
2.1. Motivation and objective of the research
2.2. Literature-based problem identification
2.3. List of crime keywords considered
3. Methodology
3.1. Implementation of the process
3.2. Proposed analytic approach
3.2.1. Kernel density estimation
4. Results and discussion
4.1. India: Crime visualization using nave Bayes and K-means algorithms
4.2. Geo-space-crime visualization (hotspot detection)-Bangalore using nave Bayes and K-means algorithms
4.3. Analysis of geospatial crime density using the KDE algorithm-India and Bangalore
4.4. Time series analysis using ARIMA model
4.4.1. One day forecasting analysis-India
4.4.2. Forecasting analysis-1day-Bangalore
4.4.3. Validation of news feed crime statistics with RTI
4.4.4. Validation findings
4.4.5. Validation hotspot detection
4.4.6. Validation of proposed model forecasting with ARIMA model
5. Conclusions
References
Chapter 8: Trust discovery and information retrieval using artificial intelligence tools from multiple conflicting source ...
1. Introduction
1.1. Trustworthiness of online or web information
2. Trusted computing
2.1. Computational trust
2.2. Trust process
3. Problem identification
4. Truth content discovery algorithm
5. Trustworthy and scalable service providers algorithm
5.1. TSSP system architecture
5.2. Graphical representation
5.3. Flow diagram of TSSP
6. Efficient feature extraction and classification (EFEC) algorithm
6.1. Graphical representation of the EFEC algorithm
6.2. Data flow diagram of EFEC algorithm
6.2.1. Feature extraction
6.2.2. Feature gathering
7. QUERY retrieval time (QRT)
7.1. Programming environment
7.2. Comparison with state-of-the-art methods
8. Trust content discovery and trustworthy and scalable service providers algorithm
8.1. Simulation result
8.2. System execution time (SET)
8.3. Communication cost (CC)
8.4. Trust score (TS)
9. Efficient feature xtraction and classification (EFEC) algorithm and customer review datasets
9.1. Performance evaluation matrix
9.2. Accuracy and F-measure
10. Summary
11. Conclusions
12. Future enhancements
References
Chapter 9: Reliable diabetes mellitus forecasting using artificial neural network multilayer perceptron
1. Introduction
2. Related works
3. Methodology
3.1. Challenges in applying machine learning algorithms
4. Building the diabetic diagnostic criteria
5. Evaluating the diabetes outcomes using classification algorithms
5.1. Improving the accuracy of SVM and RF algorithms
5.2. ANN: Multilayer perceptron
6. Conclusions
References
Chapter 10: A study of deep learning approach for the classification of electroencephalogram (EEG) brain signals
1. Introduction
2. Methods
3. Results
3.1. EEG dataset
3.2. Implementation domain
3.3. EEG recordings and data augmentations
3.4. Data preprocessing techniques
3.5. Deep learning architectures
3.5.1. Model training
3.5.2. Regularization on models
3.5.3. Optimization
3.6. Performance evaluations
3.7. Comparative analysis
4. Discussion
4.1. Rationale
4.1.1. Choosing a proper framework for analysis
4.1.2. Collecting and storing the EEG signals
4.1.3. Extracting end-to-end features effectively
4.1.4. Validation of implemented model
4.1.5. Measurement of results
4.2. Proposed architecture to overcome the challenges
4.2.1. Dataset
4.2.2. Methodology
4.2.3. Model validation
4.2.4. Evaluation
5. Conclusions
References
Chapter 11: Integrating AI in e-procurement of hospitality industry in the UAE
1. Introduction
2. Problem statement
3. Authors contributions
4. Significance of the study
5. Theoretical framework
6. Research aims and objectives
7. Literature review
7.1. Big data business analytics in the hospitality industry
7.2. BDBA itself has two dimensions: Big data (BD) and business analytics (BA)
7.3. Deep learning and machine learning techniques in the hospitality industry
7.4. Ecosystem in hospitality
7.5. Predictive analysis in the hospitality industry
7.5.1. The initial framework for predictive analytics
7.5.2. Initial AI framework Integrating AI
7.6. Agent-based technology (ABT) in the hospitality industry
8. Major findings
8.1. Statistics of the trend in publishing
8.2. Major areas of research
8.3. Content analysis in the selected publications
8.4. New proposed conceptual framework for the hospitality industry
8.5. Conceptual model for e-procurement in the hospitality industry
8.6. Comparing various studies
8.7. Case study
8.8. Interview and survey with subject matter expert(s)
8.9. Interview and survey validation
9. Discussions
10. Major gaps in the study
11. Conclusions
References
Chapter 12: Application of artificial intelligence and machine learning in blockchain technology
1. Introduction
1.1. Blockchain characteristics
1.2. Advantages of blockchain
1.3. Blockchain trust builder attributes
2. Applications of artificial intelligence, machine learning, and blockchain technology
2.1. Flag bearers of blockchain technology
2.2. Flag bearers of artificial intelligence and machine learning
2.3. Flag bearers of blockchain and artificial intelligence
2.4. Blockchain initiatives by the government of India
2.5. Current application areas of blockchain with artificial intelligence
3. It takes two to tango: Future of artificial intelligence and machine learning in blockchain technology
3.1. Sustainability
3.2. Scalability
3.3. Security
3.4. Privacy
3.5. Adaptability
3.6. Efficiency
3.7. Transaction speed
3.8. Performance
3.9. Validation of various elements
3.10. Maintainability
3.11. Scope estimation
3.12. Cost estimation
3.13. Effort estimation
3.14. Project-specific modus operandi
4. Edge computing: A potential use case of blockchain
4.1. Edge computing architectures
4.2. Flag bearers of edge computing
4.3. Applications of blockchain technology in edge computing
5. Conclusions
References
Part II: Data science and predictive analysis
Chapter 13: Implementing convolutional neural network model for prediction in medical imaging
1. Introduction
1.1. Deep learning against machine learning
1.2. Deep learning algorithms
2. Convolutional neural networks
2.1. Computer image recognition
2.2. Image classification
2.3. Why convolutional neural networks?
2.4. Functional description of a CNN
2.4.1. Convolution layer
Multiply the corresponding pixel values
2.4.2. ReLU layer
2.4.3. Pooling layer
2.4.4. Fully connected layer
3. Implementing CNN for biomedical imaging and analysis
3.1. Importing the essential python libraries and Keras library
3.2. Printing the folder name by using the librarys list directory function
3.3. Image generation for evaluation
3.4. Implementing CNN through high-level library Keras
3.5. Making CNN model
3.6. Analysis of accuracy and result
3.7. Plotting accuracy and loss graph for each epoch process
4. Architecture models for different image type
4.1. VGG 16
4.2. VGG on chest X-ray dataset
5. Conclusion
6. Future scope
References
Chapter 14: Fuzzy-machine learning models for the prediction of fire outbreaks: A comparative analysis
1. Introduction
2. Related literature
3. Research methodology
3.1. Data acquisition
3.2. Data label estimation using interval type-2 fuzzy logic
3.3. Data normalization
3.4. Feature selection and dimensionality reductions
4. Machine learning algorithms for fire outbreak prediction
4.1. Support vector machine
4.2. K-nearest neighbor
4.3. Random forest
4.4. Linear discriminant analysis
4.5. Classification and regression tree
4.6. K-fold cross-validation
5. Result and discussion
5.1. Graphical representation of both features and the normalized datasets
5.2. Principal component analysis result
5.3. The decision boundaries of the SVM, K-NN, RF, LDA, and CART training results
5.4. Performance evaluation of training results of the five models using ROC, specificity, and sensitivity
5.5. K-fold cross-validation results
5.6. Testing results of the SVM, K-NN, RF, LDA, and CART models for fire outbreak prediction
6. Conclusions
References
Chapter 15: Vehicle telematics: An Internet of Things and Big Data approach
1. Introduction
2. Big Data
2.1. Definition and characterization
2.2. Challenges with Big Data analytics
2.3. Big Data architecture
3. Big Data with cloud computing
3.1. Cloud computing
3.2. Cloud computing with Big Data
4. Internet of Things (IoT)
5. Vehicle telematics
5.1. Definition and overview
5.2. How a telematics system works
5.3. Architecture of a telematics system
5.4. Issues with the telematics system
5.5. Vehicle telematics and Big Data use cases
5.6. Vehicle telematics data description
6. Case study-Vehicle reaction time prediction
6.1. Dataset
6.2. Data preprocessing
6.3. Sequence formation
6.4. Feature selection
6.5. Prediction
6.6. Training the model
6.7. Evaluating the model
6.8. End notes
7. Conclusions
References
Chapter 16: Evaluate learner level assessment in intelligent e-learning systems using probabilistic network model
1. Introduction
2. Related work
3. Contribution of intelligent e-learning system using BN model
3.1. Outline of intelligent tutoring systems
3.1.1. Existing intelligent tutoring systems
3.1.2. Proposed intelligent tutoring systems
3.2. Methods of handling uncertainty
3.3. Bayesian network (BN)
3.3.1. Example
4. Learner assessment model
4.1. Design model
5. Results and discussions
6. Conclusions and future work
References
Chapter 17: Ensemble method for multiclassification of COVID-19 virus using spatial and frequency domain features over X- ...
1. Introduction
1.1. Contribution and organization of paper
2. Literature review
3. Proposed methodology
3.1. Dataset description
3.2. Preprocessing
3.3. Feature extraction
3.3.1. Wavelet transform (WT)
3.3.2. FFT
3.3.3. GLCM
3.3.4. GLDM
3.4. Supervised classifiers
4. Result analysis
4.1. Feature extraction methods analysis for the multiclassification
5. Discussion and conclusions
5.1. Discussion
5.2. Conclusions
References
Chapter 18: Chronological text similarity with pretrained embedding and edit distance
1. Introduction
2. Literature review
3. Theoretical background
3.1. Edit distance
3.2. Embeddings
4. Modeling
4.1. Modified Levenstien distance (edit score)
4.2. Embedding cosine score
4.3. Ensemble model
5. Experimental settings
5.1. Datasets
5.2. Evaluation tasks
5.3. Evaluation metrics
6. Results and discussion
6.1. Evaluation of baseline models
6.2. Evaluation of ensemble models
7. Conclusions
References
Chapter 19: Neural hybrid recommendation based on GMF and hybrid MLP
1. Introduction
2. Theoretical background and related works
2.1. Recommender systems
2.2. Machine learning- and deep learning-based recommendation
3. Neural hybrid recommendation (NHybF)
3.1. Description of the model layers
3.2. Training
4. Experiments
4.1. Implementation of the recommender system
4.2. Evaluation metrics
4.3. Datasets
4.4. Evaluation results
4.4.1. Preliminary assessments
4.4.2. Evaluation of NHybF
4.4.3. Scrambling of the dataset
4.4.4. Evaluation of the Top K of item recommendation
4.4.5. Evaluation with unknown users
4.5. Discussion
5. Conclusions
References
Chapter 20: A real-time performance monitoring model for processing of IoT and big data using machine learning
1. Introduction
1.1. Monitoring system using IoT-based sensors
1.2. Big data processing
1.3. Involvement of machine learning in manufacturing industries
2. Experimental study
2.1. System modeling
2.2. System implementation
2.3. Fault detection prediction model
3. Major findings
3.1. Monitoring system
3.2. IoT-based sensor performance
3.3. Big data processing performance
3.4. Fault detection prediction model
4. Conclusions
References
Chapter 21: COVID-19 prediction from chest X-ray images using deep convolutional neural network
1. Introduction
1.1. Contributions of this study
1.2. Literature review
2. Methodology
2.1. Dataset development
2.2. Data augmentation
2.3. Proposed architecture
2.4. Model development
3. Results and discussions
4. Conclusions
References
Further reading
Chapter 22: Hybrid deep learning neuro-fuzzy networks for industrial parameters estimation
1. Introduction
1.1. Literature survey
1.2. Research gaps
1.3. Objectives of this work
2. Preliminaries
2.1. Deep learning neural network (DNN) controller
2.2. Fuzzy logic controller
2.2.1. Normalization
2.2.2. Fuzzification
2.2.3. Knowledge base
2.2.4. Database
2.2.5. Choice of membership function
2.2.6. Choice of scaling factor
2.2.7. Rule base
2.2.8. Inference engine
2.2.9. Defuzzification and denormalization
2.3. Hybrid deep learning neuro-fuzzy logic controller (HDNFLC)
2.3.1. Structure of a deep learning neuro-fuzzy controller
3. Methodology
3.1. Development of the deep learning neural network (DNN) model
3.2. Development of the fuzzy logic (FLC) model
3.3. Hybrid deep learning neuro-fuzzy system
3.3.1. Altering scale factors
4. Results and discussion
4.1. Development of the deep learning neural network model
4.1.1. Generation of input and output data
4.1.2. Identification of optimal architecture
4.2. Development of fuzzy controller for AHS
4.3. Development of the hybrid deep learning neuro-fuzzy model
4.3.1. Convolutional neural network (CNN)
5. Validation of model
6. Discussions on performance evaluation
7. Conclusions
8. Future scope
References
Chapter 23: An intelligent framework to assess core competency using the level prediction model (LPM)
1. Introduction
2. Related work
2.1. Summary of limitations
2.2. Limitations that are considered for design
3. Existing applications
3.1. JAGRAN JOSH computer GK quiz
3.2. EDU ZIP the knowledge hub
3.3. TREE KNOX computer quiz
4. Proposed system
4.1. Architecture of the system
4.2. Experimental setup
4.2.1. Categorization of questions
4.2.2. Database creation and insertion
4.2.3. Decision tree algorithm: Model creation
5. Experimental
5.1. Classical methods of conducting tests
5.2. Exam conducted through the level prediction model
5.3. Comparison of the classic exam method and the level prediction model
6. Conclusions
References
Part III: Edge computing
Chapter 24: Edge computing: A soul to Internet of things (IoT) data
1. Introduction
2. Edge computing characteristics
2.1. Dense geographical distributions
2.2. Mobility support
2.3. Location awareness
2.4. Proximity
2.5. Low latency
2.6. Heterogeneity
3. New challenges in Internet of technology (IoT): Edge computing
3.1. Data aggregation amount and rate of IoT devices
3.2. Latency
3.3. Network bandwidth constraints
3.4. Resource constrained devices
3.5. Uninterrupted services with intermittent connectivity to the cloud
3.6. Security challenges
3.6.1. Managing security credentials and software updates
3.6.2. Assessing the security status of large distributed systems in a trustworthy manner
3.6.3. Responding to security compromises without causing intolerable disruptions
3.7. Scalability
3.8. Privacy
3.9. Domination of few stakeholders (monopoly vs. open IoT competition)
4. Edge computing support to IoT functionality
4.1. Device management
4.2. Security
4.3. Priority messaging
4.4. Data aggregation
4.5. Data replication
4.6. Cloud enablement
4.7. IoT image and audio processing
5. IoT applications: Cloud or edge computing?
6. Benefits and potential of edge computing for IoT
6.1. Low latency
6.2. Less power consumption by IoT devices
6.3. Simpler, cheaper devices
6.4. Bandwidth availability and efficient data management
6.5. Network connectivity
6.6. Network security
6.7. Autonomy
6.8. Data privacy
6.9. Data filtering/prioritization
6.10. Support to 5G technology
7. Use case: Edge computing in IoT
7.1. Autonomous vehicles
7.2. Smart cities
7.3. Smart grid
7.4. Industrial manufacturing
7.5. Health care
7.6. Cloud gaming
7.7. Augmented reality devices
8. Pertinent open issues which require additional investigations for edge computing
8.1. Privacy and security
8.2. Convergence and consistency
8.3. Managing edge resources
8.4. Software and hardware updates
8.5. Service delivery and mobility
8.6. Cost
8.7. Collaborations between heterogeneous edge computing systems
9. Conclusions
References
Chapter 25: 5G: The next-generation technology for edge communication
1. Introduction
2. History
2.1. 1G: That is where it all started
2.2. 2G: Cultural revolution
2.3. 3G: ``Pack-switching´´ version
2.4. 4G: Broadcast time
2.5. 5G: Internet of things age
3. 5G technology
3.1. 1G: Radio access network
3.2. Core network
4. 5G cellular network
5. Components used in 5G technology/network
5.1. 3GPP on 5G
5.2. Spectrum for 5G and frequency
5.3. MEC (multiaccess edge computing)
5.4. NFV (network function virtualization) and 5G
5.5. 5G RAN architecture
5.6. eCPRI
5.7. Network slicing
5.8. Beamforming
6. Differences from 4G architecture
6.1. Worldwide adoption of 5G
7. Security of 5G architecture
8. 5G time period
9. Case study on 5G technology
9.1. 5G use cases and services
9.2. The 5G project use cases
9.2.1. E-health
9.2.2. Smart grid
9.2.3. IoT and sensor
9.2.4. Video surveillance
9.2.5. Agriculture 2.0
9.2.6. Security issues
9.3. Smart mobility
10. 5G advancement
10.1. Superspeed
10.2. Increased bandwidth
10.3. Global wide coverage
10.4. Our own world will be a Wi-Fi zone
10.5. Improved battery life
11. Advantage and disadvantage of 5G technology
11.1. Important benefits
11.2. Other benefits of common people
11.3. Disadvantages
12. Challenges
12.1. Technological challenges
12.2. Common challenges
13. Future scope
14. Conclusions
References
Chapter 26: Challenges and opportunities in edge computing architecture using machine learning approaches
1. Introduction
2. Overview of edge computing
2.1. Architecture of edge computing
2.2. Use cases of edge computing
2.2.1. Smart grid
2.2.2. Cloud Offloading
2.2.3. Video analysis
2.2.4. Autonomous driving
2.3. Advantages of edge computing
2.3.1. Latency
2.3.2. High energy efficiency
2.3.3. Reduction in bandwidth
3. Security and privacy in edge computing
4. Intersection of machine learning and edge using enabling technologies
4.1. Defining AI, ML, DL
4.2. Enabling technologies for machine learning and edge computing
4.2.1. Tensor processing unit (TPU)
4.2.2. CPU and GPUs
4.2.3. Field programmable gate array (FPGA)
5. Machine learning and edge bringing AI to IoT
6. OpenVINO toolkit
6.1. Example of edge computing architecture for malaria detection
6.1.1. Setup and dataset
6.1.2. Pretrained model
6.1.3. Inferences
6.2. Edge computing architecture developed by industry pioneers
6.2.1. Edge TPU by Google
6.2.2. AmazonΒ΄s AWS for edge
7. Challenges in machine learning and edge computing integration
7.1. Different data distribution
7.2. Discovering edge node
7.3. Secure usage of edge nodes
7.4. Heterogeneity in data
7.5. Energy consumption of edge devices
8. Conclusions
References
Chapter 27: State of the art for edge security in software-defined networks
1. Introduction
2. Hybrid software-defined networks
3. Security challenges in hybrid software-defined networks
4. Solutions for hybrid software-defined networks
4.1. QoS (quality of service)
4.2. DDoS (distributed denial-of-service) attack
4.3. MITM (man In the middle) attack
4.4. Programmable network solution
4.5. ARP poisoning
4.6. DoS (denial-of-service) attack
4.7. Botnet attacks
4.8. Platforms for hybrid software-defined networks
5. Learning techniques for hybrid software-defined networks
5.1. Machine-learning techniques
5.2. Supervised learning
5.3. Unsupervised learning
5.4. Deep learning
6. Discussion and implementation
7. Conclusions
References
Further reading
Chapter 28: Moving to the cloud, fog, and edge computing paradigms: Convergences and future research direction
1. Introduction
2. Features and differences between cloud, fog, and edge computing
2.1. Cloud computing
2.1.1. Cloud computing
2.2. Edge computing (EC)
2.2.1. Edge computing technical difficulties and challenges
2.2.2. Edge computing perspective
2.3. Fog computing
3. Framework and programming models: Architecture of fog computing
3.1. Framework as well as programming models: Data modeling within fog computing
4. Moving cloud to edge computing
4.1. The necessity for edge computing
4.2. Challenges in industries that are different
4.2.1. 5G brought edge/edge brought 5G
4.2.2. Information caching at the edge data center
4.2.3. Digital era transformation in the manufacturing industry
5. Case study: Edge computing for intelligent aquaculture
5.1. Technology considerations
5.2. Guide architectures
5.3. Logically centralized control plane
5.4. Architectural considerations that are shaping future edge computing
6. Conclusions
References
Chapter 29: A comparative study on IoT-aided smart grids using blockchain platform
1. Introduction to smart grid, IoT role, and challenges of smart grid implementations
1.1. Introduction
1.1.1. Motivation
1.1.2. Background
1.1.3. Smart grid
Detailed smart grid insight
Smart grid, a grid modernization
Traditional electricity grids vs smart grid
Smart grid components
1.1.4. Characteristics of smart grid
Functionalities of smart grid
Vulnerabilities associated with smart grids
1.1.5. Internet of things
Importance of IoT
Technologies associated with IoT
Industrial IoT
Business insight of IoT
IoT applications
IoT in energy management
Applications and programs for IoT in smart grid
1.1.6. Working principle of IoT-aided smart grid
Smart grid architecture
Conceptual model
Communication network
Home area network (HAN)
Neighborhood area network (NAN)
Field area network (FAN)
Wide area network (WAN)
Wireless sensor networks (WSNs)
1.1.7. IoT role in smart grid
1.1.8. Summary of optimized smart grid IoT architectures
1.1.9. Technological requirements to implement IoT enabled smart grid
1.1.10. Challenges associated with IoT enabled smart grid
2. Secure smart grid using blockchain technology
2.1. Blockchains opportunities and emerging solutions in energy sector
2.2. Blockchain-based smart grid
2.2.1. A potential blockchain framework
Structure of blocks
2.2.2. Procedure for a smart grid based on blockchain
2.3. Proposed blockchain-enabled smart grid framework
2.3.1. Environment setup
2.3.2. Description of the proposed framework
2.3.3. Proposed environment design
2.3.4. Blockchain application development
2.3.5. Blockchain network creation
2.3.6. Proposed smart contract
2.4. Result discussion
2.4.1. Proof of work and nonce
2.4.2. Dynamic block difficulty
2.4.3. Wallet and chain utility key generation
2.4.4. Testing and signature of transaction
2.5. Proposed work benefit
2.6. Comparative analysis
3. Conclusions
References
Chapter 30: AI cardiologist at the edge: A use case of a dew computing heart monitoring solution
1. Introduction
2. Related work
2.1. Internet of medical things
2.2. Health-care edge computing IoT solutions
2.3. ML and DL with edge computing
3. Architectural approach
3.1. Postcloud architectures
3.2. Dew computing solution
3.3. Autonomous AI-based solution
4. ECGalert use case
4.1. Functional description
4.2. AI solution at the edge
5. Discussion
5.1. Challenges
5.2. Benefits and disadvantages
6. Conclusions
References
Index
Back Cover