Artificial Intelligence in Industry 4.0 and 5G Technology E-Book
115,99 €
Mehr erfahren.
- Herausgeber: John Wiley & Sons
- Kategorie: Wissenschaft und neue Technologien
- Sprache: Englisch
Artificial Intelligence in Industry 4.0 and 5G Technology Explores innovative and value-added solutions for application problems in the commercial, business, and industry sectors As the pace of Artificial Intelligence (AI) technology innovation continues to accelerate, identifying the appropriate AI capabilities to embed in key decision processes has never been more critical to establishing competitive advantage. New and emerging analytics tools and technologies can be configured to optimize business value, change how an organization gains insights, and significantly improve the decision-making process across the enterprise. Artificial Intelligence in Industry 4.0 and 5G Technology helps readers solve real-world technological engineering optimization problems using evolutionary and swarm intelligence, mathematical programming, multi-objective optimization, and other cutting-edge intelligent optimization methods. Contributions from leading experts in the field present original research on both the theoretical and practical aspects of implementing new AI techniques in a variety of sectors, including Big Data analytics, smart manufacturing, renewable energy, smart cities, robotics, and the Internet of Things (IoT). * Presents detailed information on meta-heuristic applications with a focus on technology and engineering sectors such as smart manufacturing, smart production, innovative cities, and 5G networks. * Offers insights into the use of metaheuristic strategies to solve optimization problems in business, economics, finance, and industry where uncertainty is a factor. * Provides guidance on implementing metaheuristics in different applications and hybrid technological systems. * Describes various AI approaches utilizing hybrid meta-heuristics optimization algorithms, including meta-search engines for innovative research and hyper-heuristics algorithms for performance measurement. Artificial Intelligence in Industry 4.0 and 5G Technology is a valuable resource for IT specialists, industry professionals, managers and executives, researchers, scientists, engineers, and advanced students an up-to-date reference to innovative computing, uncertainty management, and optimization approaches.
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 569
Veröffentlichungsjahr: 2022
Ähnliche
Table of Contents
Cover
Title Page
Copyright
List of Contributors
Preface
Profile of Editors
Acknowledgments
1 Dynamic Key‐based Biometric End‐User Authentication Proposal for IoT in Industry 4.0
1.1 Introduction
1.2 Literature Review
1.3 Proposed Framework
1.4 Comparative Analysis
1.5 Conclusion
References
2 Decision Support Methodology for Scheduling Orders in Additive Manufacturing
2.1 Introduction
2.2 The Additive Manufacturing Process
2.3 Some Background
2.4 Proposed Approach
2.5 Results
2.6 Conclusions
References
3 Significance of Consuming 5G‐Built Artificial Intelligence in Smart Cities
3.1 Introduction
3.2 Background and Related Work
3.3 Challenges in Smart Cities
3.4 Need for AI and Data Analytics
3.5 Applications of AI in Smart Cities
3.6 AI‐based Modeling for Smart Cities
3.7 Conclusion
References
4 Neural Network Approach to Segmentation of Economic Infrastructure Objects on High‐Resolution Satellite Images
4.1 Introduction
4.2 Methodology for Constructing a Digital Terrain Model
4.3 Image Segmentation Problem
4.4 Segmentation Quality Assessment
4.5 Existing Segmentation Methods and Algorithms
4.6 Classical Methods
4.7 Neural Network Methods
4.8 Segmentation with Neural Networks
4.9 Convolutional Neural Networks
4.10 Batch Normalization
4.11 Residual Blocks
4.12 Training of Neural Networks
4.13 Loss Functions
4.14 Optimization
4.15 Numerical Experiments
4.16 Description of the Training Set
4.17 Class Analysis
4.18 Augmentation
4.19 NN Architecture
4.20 Training and Results
4.21 Conclusion
Acknowledgments
References
5 The Impact of Data Security on the Internet of Things
5.1 Introduction
5.2 Background of the Study
5.3 Problem Statement
5.4 Research Questions
5.5 Literature Review
5.6 Research Methodology
5.7 Chapter Results and Discussions
5.8 Answers to the Chapter Questions
5.9 Chapter Recommendations
5.10 Conclusion
References
6 Sustainable Renewable Energy and Waste Management on Weathering Corporate Pollution
6.1 Introduction
6.2 Literature Review
6.3 Conceptual Framework
6.4 Conclusion
Acknowledgment
References
7 Adam Adaptive Optimization Method for Neural Network Models Regression in Image Recognition Tasks
7.1 Introduction
7.2 Problem Statement
7.3 Modifications of the Adam Optimization Method for Training a Regression Model
7.4 Computational Experiments
7.5 Conclusion
Acknowledgments
References
8 Application of Integer Programming in Allocating Energy Resources in Rural Africa
8.1 Introduction
8.2 Quadratic Assignment Problem Formulation
8.3 Current Linearization Technique
8.4 Algorithm
8.5 Conclusions
References
9 Feasibility of Drones as the Next Step in Innovative Solution for Emerging Society
9.1 Introduction
9.2 An Overview of Drone Technology and Its Future Prospects in Indian Market
9.3 Literature Review
9.4 Methodology
9.5 Discussion
9.6 Conclusions
References
Notes
10 Designing a Distribution Network for a Soda Company: Formulation and Efficient Solution Procedure
10.1 Introduction
10.2 New Distribution System
10.3 The Mathematical Model to Design the Distribution Network
10.4 Solution Technique
10.5 Heuristic Algorithm to Restore Feasibility
10.6 Numerical Analysis
10.7 Conclusions
References
11 Machine Learning and MCDM Approach to Characterize Student Attrition in Higher Education
11.1 Introduction
11.2 Proposed Approach
11.3 Case Study
11.4 Results
11.5 Conclusion
References
12 A Concise Review on Recent Optimization and Deep Learning Applications in Blockchain Technology
12.1 Background
12.2 Computational Optimization Frameworks
12.3 Internet of Things (IoT) Systems
12.4 Smart Grids Data Systems
12.5 Supply Chain Management
12.6 Healthcare Data Management Systems
12.7 Outlook
References
13 Inventory Routing Problem with Fuzzy Demand and Deliveries with Priority
13.1 Introduction
13.2 Problem Description
13.3 Mathematical Formulation
13.4 Computational Experiments
13.5 Conclusions and Future Work
References
14 Comparison of Defuzzification Methods for Project Selection
14.1 Introduction
14.2 Problem Description
14.3 Mathematical Model
14.4 Constraints
14.5 Methods of Defuzzification and Solution Algorithm
14.6 Results
14.7 Conclusions
References
15 Re‐Identification‐Based Models for Multiple Object Tracking
15.1 Introduction
15.2 Multiple Object Tracking Problem
15.3 Decomposition of Tracking into Filtering and Assignment Tasks
15.4 Cost Matrix Adjustment in Assignment Problem Based on Re‐Identification with Pre‐Filtering of Descriptors by Quality
15.5 Computational Experiments
15.6 Conclusion
Acknowledgments
References
Index
End User License Agreement
List of Tables
Chapter 1
Table 1.1 Key terms of IoT security.
Table 1.2 Security threats to IoT architecture.
Table 1.3 Notation and description.
Table 1.4 Comparative Analysis of the Proposed Framework with Existing Auth...
Chapter 2
Table 2.1 Parameters of orders.
Table 2.2 Details of orders.
Table 2.3 Pareto solutions.
Table 2.4 Solution 1 schedule.
Table 2.5 Solution 2 schedule.
Table 2.6 Solution 1 execution time by printer.
Table 2.7 Solution 2 execution time by printer.
Table 2.8 Solution 1 orders completion times.
Table 2.9 Solution 2 orders completion times.
Chapter 4
Table 4.1 The initial ratio of the classes in the dataset.
Table 4.2 The ratio of classes in the dataset after augmentation.
Table 4.3 Comparison of ROC AUC values of different segmentation architectu...
Table 4.4 Comparison of
F
1‐score values of different segmentation architectu...
Table 4.5 Comparison of IoU values of different segmentation architectures....
Chapter 5
Table 5.1 Reliability Table.
Table 5.2 Remote recording.
Table 5.3 Service delivery and customer satisfaction.
Table 5.4 Internet security software.
Table 5.5 Descriptive analysis (Mean 1).
Table 5.6 Security threats that are affecting IoT devices (Mean 2).
Table 5.7 Security threats effects on IoT (Mean 3).
Table 5.8 Descriptive analysis of section (Mean 4).
Chapter 6
Table 6.1 A review of definitions of corporate social responsibility (CSR)....
Chapter 7
Table 7.1 Accuracy of trained regression models of the eye images blurring ...
Table 7.2 The accuracy of a model trained by various optimization methods f...
Chapter 10
Table 10.1 Production capacity per plant.
Table 10.2 Fixed costs and storage capacity in the cross‐docking.
Table 10.3 Plant distribution costs to cross‐docking.
Table 10.4 Distribution costs in Mexican pesos ($) per pallet.
Table 10.5 Demand‐Supply of each plant for each distribution center.
Table 10.6 Supply of product from each plant.
Table 10.7 Distribution costs in the scenario corresponding to the year 201...
Table 10.8 Results obtained with the relaxations proposed for the year 2019...
Table 10.9 Demand‐supply of each plant for each distribution center.
Table 10.10 Results corresponding to scenario 2020.
Table 10.11 Quantity of products shipped from plant
k
to cross‐docking ware...
Table 10.12 Demand‐Supply of each plant for each distribution center.
Table 10.13 Results corresponding to scenario 2021.
Table 10.14 Quantity of products shipped from plant
k
to cross‐docking ware...
Table 10.15 Demand‐Supply of each plant for each distribution center.
Table 10.16 Results corresponding to scenario 2022.
Table 10.17 Quantity of products shipped from plant
k
to the cross‐docking ...
Table 10.18 Demand‐Supply of each plant for each distribution center.
Table 10.19 Results corresponding to scenario 2023.
Table 10.20 Quantity of products shipped from plant k to the cross‐docking
Chapter 11
Table 11.1 Dataset variables and calculated variables.
Table 11.2 Accuracy of classification methods.
Table 11.3 Accuracy for the simplified model.
Chapter 12
Table 12.1 Summary of computational optimization framework in blockchain te...
Table 12.2 Summary of optimization efforts in blockchain‐enabled IoT applic...
Table 12.3 Summary of recent optimization works in blockchain‐enhanced smar...
Table 12.4 Summary of recent optimization works in blockchain‐integrated su...
Table 12.5 Overview of recent optimization works in blockchain‐integrated h...
Chapter 13
Table 13.1 Data of the customers.
Table 13.2 Driving distance between locations of customers in kilometers.
Table 13.3 Solution times and total costs.
Chapter 14
Table 14.1 Average number of points in the pareto‐front and average runtime...
Table 14.2 Portfolios of the pareto‐front for the instance P32T16A8‐1 with ...
Table 14.3 Portfolios of the pareto‐front for instance P32T16A8‐1 with 75% ...
Table 14.4 Average number of points in the pareto‐front and average runtime...
Table 14.5 Portfolios of the pareto‐front for instance P32T16A8‐1 with 100%...
Table 14.6 Portfolios of the pareto‐front for instance P32T16A8‐1 with 60% ...
Chapter 15
Table 15.1 Datasets description.
Table 15.2 Results of the MOT20‐01 dataset.
Table 15.3 Results of the MOT20‐02 dataset.
List of Illustrations
Chapter 1
Figure 1.1 IoT authentication dimensions.
Figure 1.2 Proposed framework.
Figure 1.3 Enrolment phase of a new user.
Figure 1.4 High‐level diagram for pre‐processing.
Figure 1.5 Neighborhood of the central pixel
P
in 3 × 3 window.
Figure 1.6 Input byte array.
Figure 1.7 2D‐State array.
Chapter 2
Figure 2.1 Additive manufacturing stages.
Figure 2.2 Order processing flow.
Figure 2.3 Proposed approach.
Chapter 3
Figure 3.1 5G‐built AI domains.
Figure 3.2 5G‐built AI‐enabled smart cities.
Figure 3.3 AI‐based smart city applications.
Figure 3.4 AI‐based expert system.
Figure 3.5 Smart cities deployment model.
Figure 3.6 5G‐built AI‐enabled model for smart cities.
Figure 3.7 Filter methods vs wrapper methods.
Chapter 4
Figure 4.1 The scheme of the first stage of the method.
Figure 4.2 The scheme of the second and third stages of the method.
Figure 4.3 Possible graph structures for the method of conditional random fi...
Figure 4.4 U‐net [19] architecture. After each step of increasing spatial re...
Figure 4.5 Dense‐block with four convolutional layers [22].
c
is a concatena...
Figure 4.6 Examples of activation functions: (a) is a rectified linear funct...
Figure 4.7 Schematic representation of a simple fully connected neural netwo...
Figure 4.8 Schematic image of the convolutional layer.
Figure 4.9 VGG 16 architecture [34].
Figure 4.10 Schematic representation of the segmentation neural network arch...
Figure 4.11 An illustration of how the max pooling and max unpooling functio...
Figure 4.12 Residual block [39].
Figure 4.13 Regions of the dataset on the map. The pushpin icon marks the ce...
Figure 4.14 Scheme of the main augmentation steps.
Figure 4.15 (a) is the scheme of the FCDenseNet architecture family; (b) is ...
Figure 4.16 The structure of dense layer (a), and transition layers (b).
Figure 4.17 ROC curves for classes “Roofs,” “Walls of buildings,” “Shadows o...
Figure 4.18 ROC curves for classes “Railroad infrastructure,” “Railways,” “C...
Figure 4.19 Some examples of how algorithms work. (a) The original image; (b...
Chapter 5
Figure 5.1 Transmission of data on an IoT device.
Chapter 6
Figure 6.1 Social responsibility in the process of management; a comparison ...
Figure 6.2 Proposed conceptual framework.
Chapter 7
Figure 7.1 Scheme of neural network model training to assess the blurring le...
Figure 7.2 Sample images from databases: (a) BATH; (b) CASIA.
Figure 7.3 Dependence of the average model error on the epoch number when te...
Figure 7.4 Dependence of the average error of the model on the number of the...
Figure 7.5 Distribution of the number of the example by the deviation of the...
Figure 7.6 Examples of images from the DeepGlint database.
Chapter 9
Figure 9.1 Factors for drone's technovation.
Figure 9.2 Interdependence of three modules related to drone.
Chapter 10
Figure 10.1 Current distribution system.
Chapter 11
Figure 11.1 RapidMiner auto model feature.
Figure 11.2 The tree main reasons for dropping out.
Chapter 12
Figure 12.1 Complexities in the field of optimization and machine learning....
Chapter 13
Figure 13.1 The north region of Mexico for distribution.
Figure 13.2 Current decision process.
Figure 13.3 Example of the route both with and without priority.
Figure 13.4 Solving time for instances with 5, 10, and 15 customers and diff...
Figure 13.5 Locations of the consumers.
Chapter 14
Figure 14.1 Pareto‐fronts for instance P32T16A8‐1 with the method of
k
‐prefe...
Figure 14.2 Pareto‐fronts for instance P32T16A8‐1 with integral value method...
Chapter 15
Figure 15.1 Dependence of the ratio of the IDS method to the IDS of the basi...
Guide
Cover
Table of Contents
Title Page
Copyright
List of Contributors
Preface
Profile of Editors
Acknowledgments
Begin Reading
Index
End User License Agreement
Pages
iii
iv
xv
xvi
xvii
xviii
xix
xx
xxi
xxii
xxiii
xxiv
xxv
xxvii
xxviii
xxix
xxx
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
165
166
167
168
169
170
171
172
173
174
175
176
177
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
319
320
321
322
Artificial Intelligence in Industry 4.0 and 5G Technology
Edited by
Pandian VasantMERLIN Research Centre, Ton Duc Thang UniversityVietnam
Elias MunapoDepartment of Business Statistics and Operations ResearchNorth West University, MahikengMmabatho, SA
J. Joshua ThomasUOW Malaysia, KDU Penang University CollegePenang, Malaysia
Gerhard-Wilhelm WeberDepartment of Marketing and Economic EngineeringPoznan University of TechnologyPoznan, PL
This edition first published 2022© 2022 John Wiley & Sons, Inc.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions.
The right of Pandian Vasant, Elias Munapo, J. Joshua Thomas, and Gerhard‐Wilhelm Weber to be identified as the authors of the editorial material in this work has been asserted in accordance with law.
Registered OfficeJohn Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA
Editorial Office111 River Street, Hoboken, NJ 07030, USA
For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.
Wiley also publishes its books in a variety of electronic formats and by print‐on‐demand. Some content that appears in standard print versions of this book may not be available in other formats.
Limit of Liability/Disclaimer of WarrantyThe contents of this work are intended to further general scientific research, understanding, and discussion only and are not intended and should not be relied upon as recommending or promoting scientific method, diagnosis, or treatment by physicians for any particular patient. In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of medicines, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each medicine, equipment, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.
Library of Congress Cataloging‐in‐Publication Data Applied forISBN: 9781119798767
Cover Design: WileyCover Image: © Fit Ztudio/Shutterstock
List of Contributors
Haidar Abbas
Department of Business Administration, Salalah College of Applied Sciences
University of Technology and Applied Sciences
Sultanate of Oman
Sadia S. Ali
Department of Industrial Engineering
College of Engineering
Jeddah 21589
Saudi Arabia
Nancy M. Arratia‐Martinez
Universidad de las Américas Puebla
Department of Business Administration, Ex‐Hacienda Santa Catarina Mártir S/N
Puebla, 72810
México
Arrieta‐M Luisa F
Simon Bolivar University
Barranquilla
Atlántico
Colombia
Paulina A. Avila‐Torres
Universidad de las Américas Puebla
Ex‐Hacienda Santa Catarina Mártir
Department of Business Administration
San Andrés Cholula Puebla, C.P. 72810
México
Swapnoj Banerjee
Maulana Abul Kalam Azad University of Technology
Meghnad Saha Institute of Technology
Department of Computer Science and Engineering
Kolkata
West Bengal 700150
India
Choo K. Chin
University of Northern Iowa
Bachelor of Arts in Computer Information System
Joshua E. Chukwuere
North‐West University
Department of Information Systems
South Africa
Irraivan Elamvazuthi
Persiaran UTP
Seri Iskandar
Perak
Timothy Ganesan
Member of American Mathematical Society
University Drive NW
University of Calgary
Alberta
Canada
Alexander N. Gneushev
Moscow Institute of Physics and Technology
Department of Control and Applied Mathematics
Moscow Region, 141701
Russia
and
Federal Research Center “Computer Science and Control” of Russian Academy of Sciences
Dorodnicyn Computing Center
Moscow
Russia
Alexey D. Grigorev
Moscow Institute of Physics and Technology
Department of Control and Applied Mathematics
Moscow Region, 141701
Russia
Soumodipto Halder
Maulana Abul Kalam Azad University of Technology
Meghnad Saha Institute of Technology
Department of Computer Science and Engineering
Kolkata
West Bengal 700150
India
Y. Bevish Jinila
Department of Information Technology
Sathyabama Institute of Science and Technology
Chennai
India
Cinthia Joy
Sydney International School of Technology and Commerce
Sydney
Australia
Rajbir Kaur
Government Girls College
Panchkula
Haryana 134001
India
Vladimir A. Kozub
Moscow Institute of Physics and Technology
Department of Control and Applied Mathematics
Moscow Region, 141701
Russia
and
AEROCOSMOS Research Institute for Aerospace Monitoring
Moscow, 105064
Russia
Igor S. Litvinchev
Federal Research Center "Computer Science and Control" of Russian Academy of Sciences
Dorodnicyn Computing Center
Moscow, 119333
Russia
Lopez‐I Fernando
Autonomous University of Nuevo León
Department of Mechanical and Electronic Engineering
San Nicolás de los Garza
Nuevo León
México
Elias Manopo
Department of Statistics and Operations Research, School of Economic Sciences
North West University
Mahikeng
South Africa
Ivan A. Matveev
Federal Research Center "Computer Science and Control" of Russian Academy of Sciences
Dorodnicyn Computing Center
Moscow, 119333
Russia
Boitumelo Molefe
North‐West University
Department of Information Systems
South Africa
Subhash Mondal
Maulana Abul Kalam Azad University of Technology
Meghnad Saha Institute of Technology
Department of Computer Science and Engineering
Kolkata, West Bengal 700150
India
Alexander B. Murynin
AEROCOSMOS Research Institute for Aerospace Monitoring
Moscow, 105064
Russia
and
Federal Research Center “Computer Science and Control” of Russian Academy of Sciences
Dorodnicyn Computing Center
Moscow, 119333
Russia
Denis Y. Nartsev
Moscow Institute of Physics and Technology
Department of Control and Applied Mathematics
Moscow Region, 141701
Russia
Juan Jesús Tello Rodríguez
Universidad Autonoma de Nuevo León
Facultad de Ingeniería Mecánica y Eléctrica (FIME), Posgrado en Ingeniería en Sistemas (PISIS)
San Nicolás de los Garza
Nuevo León 66451
México
Diganta Sengupta
Maulana Abul Kalam Azad University of Technology
Meghnad Saha Institute of Technology
Department of Computer Science and Engineering
Kolkata
West Bengal 700150
India
S. Prayla Shyry
Sathyabama Institute of Science and Technology
Chennai
India
Isidro Soria‐Arguello
Departamento de Ingeniería Química, Industrial y de Alimentos
Universidad Iberoamericana
Ciudad de Mexico 01219
Mexico
J. Joshua Thomas
Department of Computing
UOW Malaysia
KDU Penang University College
Penang
Malaysia
Rafael Torres‐Esobar
Facultad de Ingeniería
Universidad Anáhuac México
Huixquilucan 52786
Mexico
Pandian Vasant
Ton Duc Thang University
Modeling Evolutionary Algorithms Simulation & Artificial Intelligence (MERLIN), Faculty of Electrical & Electronic Engineering
Ho Chi Minh City 700000
Vietnam
Deng H. Xiang
He Ying Metal Industries Sdn Bhd.
Malaysia
Preface
Regarded from our editorial views, this monograph has been wished to become a most valuable reference work of 2021 within the domains of modern information and decision aiding, urban and international, local and regional, ecological and spatiotemporal, industrial and natural operational research, artificial intelligence and creative arts and sciences. Proceeding diligent work from the sides of the authors and of us all, the endeavor of this work succeeded by stimulating, gathering, and compiling the newest research on the present state and inventions of electric and electronic, informational and energetic, green and recoverable, creative and re‐creative means and their smart and dynamic employment, with care and responsibility. Our thus given monograph and handbook about scientific research on concerns about natural and human resources and their supply is a remarkable scholarly resource. It analyzes and discusses the efficient utilization of those resources that have a supportive impact on sustainable development and relations of, within and between us humans and our communities, cities and rural countryside, and finally migrations, social peace, and peace among our countries.
For this book's international directions, it has advanced toward a very special resource outlining the remarkable progress obtained worldwide, related to artificial intelligence, operational research, electronic, information and mobility devices and methods, renewable energy, natural resources, etc. The book is on the way to becoming respected on a global level for its broad analytic and practical contents.
“Artificial Intelligence in Industry 4.0 and 5G Technology” provides details on cutting‐edge methodologies utilized in business and industrial sectors. It gives a holistic background on innovative optimization applications, focusing on main technology sectors such as 5G networks, Industry 4.0, and robotics. It discusses topics such as hyper‐heuristics algorithmic enhancements and performance measurement approaches and provides keen insights into the implementation of meta‐heuristic strategies to many‐objectives optimization real‐life problems in business, economics, and finance. With this book, the esteemed readers can learn to solve real‐world sustainable optimization problems effectively using the appropriate techniques from emerging fields including artificial intelligence, hybrid evolutionary and swarm intelligence, hyper‐heuristics programming, and many‐objectives optimization.
“Artificial Intelligence in Industry 4.0 and 5G Technology” is a well‐chosen collection of creative research about the methodologies and utilization of deep learning approaches in business, economics and finance, science and engineering, neuroscience, and medicine. While highlighting topics including intelligent optimization and computational modeling, data hybridization, and artificial intelligence, this work is ideally shaped and elaborated for high‐tech experts and engineers, IT specialists and big‐data analysts, data scientists and engineers, researchers and academicians, philanthropes, and political decision makers who look for contemporary studies on deep learning and its fruitful utilization in upcoming smart and green industries.
Deep Learning is a method that is transforming the world of data and analytics. Optimization of this new approach is still unclear; however, there is a need for research on various applications and techniques of deep learning in the areas of artificial intelligence and machine learning. Modern methodologies and tools from neuro‐imaging, brain imaging, etc., today generate high‐quality neurophysiological data with a resolution quality never reached before. These accelerating dynamics generate promising pathways to improve our comprehension of the nervous system and eventually of deeper learning. Computational issues occurred because of the high complexity of neuronal systems and the big number of constituents with still unknown connections between them. Highly innovative considerations and methods of computational neuroscience lead to more realistic biophysical representations which provide amazing chances for conditional behavior and links among brain regions in economic, professional, and our daily decision‐making spheres.
This handbook encompasses three areas called “units:”
Unit 1 Industry 4.0:
Advanced Techniques and Technologies for Energy Saving, Artificial Intelligence in Smart Agriculture and Agroengineering and their AI‐optimized Hardware, Biometrics, Big‐Data, Cloud Computing, Cybersecurity, Embedded Systems, Fractional Differential Approach on Machine Learning, Graph‐based Data Analysis, Grid Computing, Internet of Things (IoT), Intelligent Spindle Frameworks, Knowledge Representation and Reasoning, Smart Manufacturing, Spindle Frameworks, Manufacturing Intelligence and Informatics, Unmanned Arial Vehicles (drones technology).
Unit 2 Artificial Intelligence:
Augmented AI, Adaptive Systems, Bioinformatics, Data Mining, Deep Learning, Evolutionary Computations, Fuzzy Logic, Hybrid and Nonlinear Systems, Knowledge Representation and Reasoning, Machine Learning, Meta Heuristics, Mathematical Modelling in Artificial Intelligence, Natural Language Processing, Natured‐Inspired Algorithms, Robotics Automation, Swarm Intelligence.
Unit 3 5G Technology:
Innovative Smart Cities Design and Applications, Mathematical Optimization in Engineering and Business Applications, Multi‐task Learning, Radio Communications Technologies, Remote Access and Control, Robotic Process Automation, Sequential and Image Processing, Smart Communications Systems, Smart Grid, Speech Recognition, Sensors, Virtual Machines, Vehicular Networks, Wireless Sensor Networks, Wearable Technologies.
Subsequently, we provide a short introduction to the fifteen chapters of this work.
In the first chapter “Dynamic Key‐based Biometric End User Authentication Proposal for IoT in Industry 4.0,” the coauthors Subhash Mondal, Swapnoj Banerjee, Soumodipto Halder, and Diganta Sengupta recalls research and growth in IoT, both in the prospects of architecture as well as a voluminous increase in inter‐networked devices. This chapter concentrates on securing the data acquisition at the end‐user node level by means of biometric authentication. Extended AES algorithm incorporates an S‐box, which defines the edge security through a nonlinear behavior. This seems to be a first attempt to combine Minutiae Extraction algorithm, Key Generation algorithm, AES, and fingerprint authentication to generate a single Edge‐Framework for fingerprint authentication for the end users.
By intelligence algorithms, the second chapter “Decision Support Methodology for Scheduling Orders in Addictive Manufacturing” co‐authored by Juan Jesús Tello Rodriguez and Fernando Lopez Irarragorri studies additive manufacturing, a family of manufacturing technologies where a 3‐dimensional solid is manufactured, shaped from a computerized model by depositing thin layers of material. It is considered one of the most important emerging technologies because of the multiple benefits it brings for businesses, and because it is harmoniously coupled with Industry 4.0 and the digitization of manufacturing. This work addresses a variant of the job‐shop re‐scheduling problem in additive manufacturing. The new approach is applied to a real‐world problem.
In the third chapter called “Significance in consuming 5G built Artificial Intelligence in smart cities,” Y.Bevish Jinila, Cinthia Joy Godly, Joshua Thomas and S.Prayla Shyry recall that smart cities have a greater potential toward convenient, comfortable, and automated applications that could help humans make things easier. The traditional systems do not have sufficient models. AI has experienced a great “boom” in smart transportation. The advent of 5G has created a greater impact in the field of telecommunication, where the data transfer speed enormously increased. In this chapter, the significance of 5G‐built AI in smart cities is presented. The 5G‐built AI model for smart cities is also modeled. The proposed model highlights the importance of applying 5G with AI to improve the performance of the system.
The fourth chapter, titled “Neural network approach to segmentation of economic infrastructure objects on high‐resolution satellite images,” authored by V.A. Kozub, A.B. Murynin, I.S. Litvinchev, I.A. Matveev, and P. Vasant, addresses the problem of semantic segmentation of infrastructure objects in high‐resolution satellite images, considered as an integral part of the method for constructing digital terrain models. Semantic segmentation involves classes such as buildings, roads, and railways. A set of labeled satellite images is collected, and neural network architecture is selected and trained. In order to reduce the imbalance of classes in the training sample, a probabilistic method of augmentation is developed and applied.
The fifth chapter “The impact of data security on the Internet of Things” written by coauthors Joshua Ebere Chukwuere and Boitumelo Molefe reminds us that the IoT is making lives easier and more productive; security on the IoT has also increased. The main purpose of this chapter is to determine the impact of data security on IoT. A quantitative methodology was used where data were collected using questionnaires and analyzed using Statistical Package for Social Sciences (SPSS) software. The results show that security threats negatively impact the IoT which affects IoT devices adversely. Lastly, users of IoT should be made aware of these security threats and how they can keep their devices safe.
In the sixth chapter “Sustainable renewable energy and waste management on weathering corporate pollution,” the coauthors Choo Kwong Chin and Deng Huo Xiang investigate the fourth Industrial Revolution, the man, and the earth. The majority of the environmental problems can be traced to industrialization, particularly since the “great acceleration” in global economic activity. As the fourth Industrial Revolution is pacing fast, innovations are becoming faster, more efficient, and widely accessible than ever before. Emerging technologies, including the IoT, virtual reality, AI, Block‐Chain technology, etc. This study in this chapter shows that enterprises responsibly consider the environmental impacts of their activities and undertake actions aimed at preserving the environment and its resources.
The seventh chapter “Adam adaptive optimization method for neural network models regression in image recognition tasks,” coauthored by D. Yu. Nartsev, A. N. Gneushev, and I. A. Matveev proposes to apply adaptive optimization in neural network regression tasks like estimating image quality or aligning objects in the frame. Two sample tasks are presented. The first one is the evaluation of the degree of blurring in the iris recognition system. Eye images are taken from BATH and CASIA databases and Gaussian blurring generates samples. The second sample task is the alignment of the face in an image. Both tasks are solved by the direct estimation of parameters with neural networks. The resulting accuracy of parameter estimation is acceptable for practical use. The Adam algorithm and its modifications, such as AdamW and Radam, are applied.
In the eighth Chapter “Application of integer programming in allocating energy resources in rural Africa”, the author, Elias Manopo, presents a solution method for the quadratic assignment problem. In this approach, the quadratic assignment problem is first linearized into a linear binary problem. The linear binary problem is then converted into a convex quadratic problem, which is then solved efficiently by interior point algorithms. In addition to allocating resources, the quadratic assignment problem can also be used in numerical analyses and dartboard constructions, archaeology, and statistical analysis, chemical reactions and processes, economic problem modeling and decision frameworks, hospital layouts, and backboard wiring problems, campus planning models, and many more.
In the ninth chapter “The Feasibility of Drones as the Next Step in Innovative Solution for Emerging Society,” the coauthors Sadia Samar Ali, Rajbir Kaur, and Haidar Abbas state that organizations are looking for technologies capable of accomplishing multiple tasks, providing economic benefits and an edge over their competitors. Technology‐related innovations are taking businesses toward a broad spectrum, related to philanthropy, social welfare, and well‐being of all the related stakeholders. Drone technology has arrived to advance human efficiencies. This study examines various aspects of drone technology viz. utilities, complexities involved, and its prospects, especially in the Indian Context. Furthermore, it aims to identify and establish the various factors impacting “Technovation.”
The 10th Chapter called “Designing a distribution network for a Soda Company: Formulation and efficient solution procedure” by coauthors Isidro Soria Arguello, Rafael Torres Escobar, Hugo Alexer Pérez‐Vicente and Pandian Vasant aims at professionals and university students interested in the redesign and optimization of primary distribution networks related to consumer products, in a three‐tier distribution system. A MILP model is presented to design a distribution network for a soda company. To solve the model efficiently, Lagrangian relaxation technique is used. Four scenarios are analyzed. The solutions obtained are feasible in the original problem. It is guaranteed that the demand of each distribution center will be satisfied by only one macro‐center.
In the 11th Chapter named “Machine learning and MCDM approach to characterize student attrition in higher education,” the coauthors Luisa Fernanda Arrieta and Fernando Lopez‐Irarragorri discuss school dropout as a global problem. The authors present a methodology based on multicriteria decision making, and machine learning addresses assess factors related to student attrition in universities. As a real case, data from the “Universidad Simon Bolivar” are developed to demonstrate the application of the proposed methodology. Some interesting findings of factors related to student dropout are discussed.
In the 12th Chapter “A concise review on recent optimization and deep learning applications in blockchain technology,” the coauthors Timothy Ganesan, Irraivan Elamvazuthi, Pandian Vasant, and J. Joshua Thomas investigate blockchain technology as one of the most influential disruptive technologies today. The objective is to provide a concise and recent review of the implementation of optimization and deep learning techniques on blockchain‐based systems. Application areas considered are computational optimization frameworks, the IoT, smart grids, supply chain management, and healthcare data systems. Implementation of optimization and deep learning techniques are presented and discussed.
In the 13th Chapter, “Inventory routing problem with fuzzy demand and deliveries with priority,” the coauthors, Paulina A. Avila‐Torres and Nancy Maribel Arratia Martínez, consider the inventory routing problem based on a producer and distributor company of gases with three main products. The main customers are industry and hospitals. The inventory level is monitored to establish the amount of product to deliver. The problem studied aims to guarantee a minimum inventory level of the customer prioritizing a group of customers in each route and taking into account demand as fuzzy. A mathematical model based on the vehicle routing problem is presented with the main objective is to minimize the distribution cost. The model is tested with a case of study.
In the 14th Chapter, “A comparison of defuzzification methods for selecting projects,” the coauthors Nancy Maribel Arratia Martínez, Paulina A. Avila‐Torres, and Fernando López state that study the problem of project portfolio selection when the resources are subject to budget availability. They model the uncertainty in the total budget to carry out projects and the uncertainty in restricted budgets to different areas using triangular fuzzy numbers. Naturally, different methods result in different project portfolios. For this reason, the authors discuss the results of the application of different methods to compare fuzzy numbers, using SAUGMECON as the solution method.
The 15th Chapter, “Re‐identification‐based models for multiple object tracking” by the coauthors A.D. Grigorev, A.N. Gneushev, and I.S. Litvinchev addresses the multiple objects tracking problem. By factorization of the posterior distribution of objects' parameters, it is proven that the original problem is equivalent to the procedure containing two steps. Given that track measurements are not equal in terms of their usefulness for re‐identification, the technique of track descriptor pre‐filtering is introduced. Known quality assessment methods and an alternative detector‐based approach are taken into account. Computational experiments are conducted. The results showed computational efficiency and increased stability.
We, the editors, hope that the chosen fields and picked themes within this work will reflect a key selection of worldwide research coping with emerging and complex, sometimes long‐lasting problems of Artificial Intelligence in Industry 4.0 and 5G Technology and their domains in everyone's life, in professional and daily lives, in natural sciences and development, in city and country planning, in high technology and the arts, in economics and finance, in healthcare and medicine, by approaches, methods and results of operational research and artificial intelligence. We are very thankful to the publishing house of Wiley for the distinction of hosting our compendium as a pioneering scientific endeavor. Special gratitude is conveyed to the editors of publishing house Wiley, to editorial managers and staff for their continuous advice and help in every respect. We convey our thanks to all the authors for their hard work and willingness to share their novel insights and amazing inventions with our worldwide community. We sincerely hope that our authors' studies will crystalize and initialize collaboration and progress at a global and premium stage, and as a service of humility and excellence to humankind and the entire creation.
Profile of Editors
Pandian Vasant is an Editor‐in‐Chief of the International Journal of Energy Optimization and Engineering (IJEOE) and a Research Associate at MERLIN Research Centre of Ton Duc Thang University. He holds PhD in Computational Intelligence (UNEM, Costa Rica), MSc (University Malaysia Sabah, Malaysia, Engineering Mathematics), and BSc (Hons, Second Class Upper) in Mathematics (University of Malaya, Malaysia). His research interests include Soft Computing, Hybrid Optimization, Innovative Computing and Applications. He has co‐authored numerous research articles in journals, conference proceedings, presentations, special issues guest editor, book chapters (300 publications indexed in Google Scholar) and General Chair of the EAI International Conference on Computer Science and Engineering in Penang, Malaysia (2016) and Bangkok, Thailand (2018). In the years 2009 and 2015, Dr. Pandian Vasant was awarded top reviewer and outstanding reviewer for the journal Applied Soft Computing (Elsevier), respectively. He has 31 years of working experience at various universities. Currently, he is an Editor‐in‐Chief of International Journal of Energy Optimization & Engineering, Member of AMS (USA), NAVY Research Group (TUO, Czech Republic), MERLIN Research Centre (TDTU, Vietnam), and General Chair of the International Conference on Intelligent Computing and Optimization (https://www.icico.info/). H‐Index Google Scholar = 34; i‐10‐index = 144.
http://www.igi-global.com/ijeoe
E‐mail: [email protected]
Elias Munapo is a Professor of Operations Research and he holds a BSc. (Hons) in Applied Mathematics (1997), MSc. in Operations Research (2002), and a Ph.D. in Applied Mathematics (2010). All these qualifications are from the National University of Science and Technology (N.U.S.T.) in Zimbabwe. In addition, he has a certificate in outcomes‐based assessment in Higher Education and Open distance learning, from the University of South Africa (UNISA) and another certificate in University Education Induction Programme from the University of KwaZulu‐Natal (UKZN). Elias Munapo is a Professional Natural Scientist certified by the South African Council for Natural Scientific Professions (SACNASP) in 2012. Elias Munapo has vast experience in university education and has worked for five (5) institutions of higher learning. The institutions are Zimbabwe Open University (ZOU), Chinhoyi University of Technology (CUT), University of South Africa (UNISA), University of KwaZulu‐Natal (UKZN) and North‐West University (NWU). Elias Munapo has successfully supervised/co‐supervised 10 doctoral students and over 20 master's students to completion. Professor Munapo has published over 100 research articles. Of these publications, one is a book, several are book chapters and conference proceedings and the majority are journal articles. In addition, he has been awarded the North West University Institutional Research Excellence Award (IREA) thrice and is an editor of a couple of journals, has edited several books and is a reviewer of several journals.
Email: [email protected]
Elias Munapo ‐ Business Statistics and Operations Research, Northwest University, South Africa
E‐mail: [email protected]
https://commerce.nwu.ac.za/business-statistics-and-operations-research/elias-munapo, https://scholar.google.co.za/citations?user=4Npmpr0AAAAJ&hl=en
J. Joshua Thomas is an Associate Professor in Computer Science at UOW Malaysia KDU Penang University College. He obtained his PhD (Intelligent Systems Techniques) in 2015 from University Sains Malaysia, Penang and Master's degree in 1999 from Madurai Kamaraj University, India. He served as Head and deputy head of the Department of Computing between the years 2012 to 2017. From July to September 2005, he worked as a research assistant at the Artificial Intelligence Lab in University Sains, Malaysia. From March 2008 to March 2010, he worked as a research associate at the same University. His work involves intelligent systems techniques in which he adopts computational algorithm implementation in inter‐discipline field areas. His expertise is evident in working with International collaborators in publications, visiting research fellows to share knowledge. Recently, work in Deep Learning, data analytics, especially targeting Graph Convolutional Neural Networks (GCNN) and Graph Recurrent Neural Networks (GRNN), Hyper‐Graph Attentions, End‐to‐end steering learning systems, design algorithms in drug discovery and Quantum machine learning. He is a principal investigator, co‐investigator in various grants funding, including internal, national and international levels. He is an editorial board member for the Journal of Energy Optimization and Engineering (IJEOE), Book author, guest editor for Applied Sciences, Computations (MDPI), Mathematics Biosciences and Engineering (MBE), and Computer Modelling in Engineering & Sciences (CMES). He has authored and edited several books. He has published more than 50 research papers in leading international conference proceedings and peer‐reviewed journals He has been a Regular Invitee, Plenary, Keynote speaker, and Workshop Presenter in IAIM2019, LCQAI2021, ICRITCC'21, and IAIM2022. He is a “Visiting Research Fellow” to the Sathyabama Institute of Science and Technology, Chennai, India.
Weblink:https://www.uowmkdu.edu.my/research/our-people/dr-joshua-thomas
Email: [email protected]
J. Joshua Thomas, UOW Malaysia; KDU Penang University College, Malaysia
E‐mail: [email protected]
https://www.uowmkdu.edu.my/research/our-people/dr-joshua-thomas
Gerhard‐Wilhelm Weber is a Professor at Poznan University of Technology, Poznan, Poland, at Faculty of Engineering Management. His research is on OR, Financial Mathematics, Optimization and Control, neuro‐ and bio‐sciences, data mining, education, and development. He is involved in the organization of scientific life internationally. He received his Diploma and Doctorate in mathematics, and economics/business administration, at RWTH Aachen, and his Habilitation at TU Darmstadt. He held Professorships by proxy at the University of Cologne, and TU Chemnitz, Germany. At IAM, Middle East Technical University (METU), Ankara, Turkey, Prof. Weber served as a Professor in the programs of Financial Mathematics and Scientific Computing, and Assistant to the Director, and he has been a member of further graduate schools, institutes and departments of METU. Further, he has affiliations at the Universities of Siegen, Ballarat, Aveiro, North Sumatra, and the Malaysia University of Technology. Furthermore, he is an “Advisor to EURO Conferences” and “IFORS Fellow”.
Weblink: https://www.researchgate.net/profile/Gerhard_Wilhelm_Weber
Email: gerhard‐[email protected]
Gerhard‐Wilhelm Weber‐ Poznań University of Technology, Poland
E‐mail: gerhard‐[email protected] l
https://www.researchgate.net/profile/Gerhard_Wilhelm_Weber
Acknowledgments
The editors would like to acknowledge the contributions of everyone involved in the development of this book and, in particular, would like to thank all authors for their contributions. Our sincere thanks go to the chapter authors who contributed their valuable time and special thanks go to the people who, in addition to the preparations, actively participated in the evaluation process.
My thanks go to the members of the Editorial Board who have done so much in making the book of high quality, especially to the critical positions of the members of the Editorial Committee, Professor Gerhard Wilhelm Weber, Professor Pandian Vasant, Professor Ugo Fiore for their excellent progress and valuable help.
The editors are confident that readers will find this book extremely useful because, among other things, it gives them an opportunity to learn about the results of the published research. It would be an indication of the enormous role Wiley has played in publishing this information.
Finally, thanks to the editors who would like to thank all the editorial staff Team Wiley, especially Lemore Sarah, for helping the editors and authors prepare the manuscript of this book in a professional and extremely kind manner, and they went a long way to make it this book on the exceptional quality edition of the book.
I would like to thank our Lord and Savior Jesus Christ who gave life and strength to the who worked throughout the book preparation process to complete the book during the time of Covid‐19 pandemic.
1Dynamic Key‐based Biometric End‐User Authentication Proposal for IoT in Industry 4.0
Subhash Mondal, Swapnoj Banerjee, Soumodipto Halder, and Diganta Sengupta
Maulana Abul Kalam Azad University of Technology, Meghnad Saha Institute of Technology, Department of Computer Science and Engineering, Behind Urbana Complex Near, Ruby General Hospital, Nazirabad Rd, Uchhepota, Kolkata, West Bengal, 700150, India
1.1 Introduction
Internet of Things (IoT) has emerged as one of the top research and subsequent application domains lately, accounting for a rise from 4.7 billion connected devices in 2016 [1] to 11.7 billion in 2020 [2], and having a projected upper ceiling expansion of 30 billion in 2025 [2]. The phenomenal rise in IoT has also led to Distributed Denial of Service (DDoS) attacks by cybercriminals [1], which is expected to gain further momentum by 2025 owing to the explosive rise in IoT devices, approximately four devices per person [1]. Hence, the security of the IoT architecture in both the middleware and the edges has attracted huge global research. This chapter concentrates on securing the edge framework of the IoT architecture by generating a dynamic (cipher) key based on fingerprint impressions (images) of the end user. IoT devices are typically grouped into different clusters [3] having gateways (cluster heads). These gateways are further inter‐connected via the cloud platform [4,5]. The devices within the cluster communicate with each other and with the internet through the gateways, which have varied topological placements. Communication among the different devices within the cluster and the gateways within diverse topologies is generally managed through the most popular communication protocols such as the MQ Telemetry Transport (MQTT) [6] and Constrained Application Protocol (CoAP) [7]. The difference between these two is given in ([8]). Furthermore, wireless communication forms the backbone of IoT infrastructure (gateway) wherein the terminal nodes send the message packets to the framework (platform), which subsequently transmits the packets wirelessly [9]. Choice of these communication protocols [6,7] depends upon the application, but the protocols all experience certain security vulnerabilities, which are generally addressed by authentication. Moreover, authentication schemes such as Elliptic Curve Cryptography [10], Self‐Certified Keys Cryptosystem [11], and Hash Functions [12] are employed at the wireless communication end. We propose an authentication system that acts at the node level itself de‐voiding the requirement at the middleware much before the wireless communication. The proposed authentication system is primitively a template‐based architecture that uses a modified Advanced Encryption Standard (AES) algorithm for data (image) pre‐processing and feature (minutiae) extraction. Biometrics used for authentication varies from fingerprints, irises, and facial images.
As discussed earlier, IoT has witnessed a huge rise in connected devices over the past few years and is poised to increase even further [1]. This increase demands proper data management in terms of storage and processing. Several Big Data architectures have been employed for such data management leading to the use of the cloud infrastructure [13]. In terms of data management, cloud services also aid in high scalability and flexibility. Moreover, cloud‐driven IoT aids in real‐time processing at minimal cost [14]. The bottleneck in cloud‐driven IoT is authentication as the data are available across different topologies and can be accessed by any third party. Latency accounts for a further issue, which occurs while analyzing and retrieving data from the cloud. Secured cloud access with proper authentication is managed by Edge Computing [15], which provides solutions not only in the current IoT landscapes [16,17] but also in 5G. Our proposal works primarily at the authentication of the edge level and also through the cloud infrastructure. End users provide their fingerprint images, which are stored as biometric templates in the cloud database. The end users have to initially authenticate themselves to connect their devices to the IoT network. The second stage calls for encrypting the biometric template using the AES algorithm followed by template matching at the cloud level.
The rest of the paper is arranged as follows. Section 1.2 provides the related work, while Section 1.3 details the proposed approach. Comparative analysis is given in Section 1.4 and the conclusion in Section 1.5.
1.2 Literature Review
IoT and Wireless Sensor Networks (WSN) generally operate in a hand‐shaking mode, thereby raising major authentication issues. Primary investigation reveals four dimensions of authentication (Figure 1.1). Authentication factors such as ownership, knowledge, and biometrics [18] play a key role in authentication mechanisms. Knowledge and biometric factors constitute passwords and fingerprints, respectively. The four authentication techniques have been further elaborated.
Figure 1.1 IoT authentication dimensions.
Physical Unclonable Functions (PUF), defined as physical objects which are physically defined digital fingerprint responses, serve as unique identifiers for a particular input and challenge. A lightweight system for the attestation and authentication of things (AAoT) has been proposed [19], which is based on PUFs and proposes a tamper‐proof authentication feature for smart embedded devices. Memory resources are attenuated via random memory filling based on PUF technology. Mutual authentication is achieved by block optimization of AAoT. A theoretical PUF‐based authentication protocol has been experimentally validated in [20] resulting in a huge reduction in computational and storage burden. In (Wallrabenstein, 2016), PUF‐based algorithms were used for the generation of digital signature, device authentication purposes followed by enrolment and decryption [20].
Blockchain‐based authentication proposals have been implemented wherein Li et al. [21] confirm that such proposals can be made secure under the condition that the cooperating group of attacker nodes lower hash rates than the collective control of the honest nodes. Prior to the use of blockchain technology for IoT architecture authentication, it is mandatory that the complete network architecture be decentralized. Decentralization helps in efficient management and authentication of smaller blocks as proposed in [22], and the mechanism is termed as “Bubbles of Trust.” The bubbles were deployed as smart contracts, which have identification and trust capabilities among themselves. The process generates a transaction that includes two identifiers – one belonging to the device and the other to the group. A valid transaction results in a bubble creation at the device end, which generated the transaction. Blockchain technology validates the uniqueness of the identifiers. Although the scheme in [22] ensures complete anonymity, the proposal lacks identification assurance. This bottleneck can be addressed by deploying private blockchain architectures, though they suffer from new device or service insertion. Li et al. [21] have provided each device a unique identification number, which is further grafted into the blockchain, thereby enabling the device authentication and violating central authority.
The third proposal in Figure 1.1 is the key‐based authentication system, which provides an improvement in the brute force authentication techniques. Adriano Witkovski et al. [23] have proposed an authentication scheme where they provide single‐sign‐on in IoT architecture. The proposal generates integration via interaction between the key‐based internet scheme and the key‐based IoT scheme. The proposal limits the device count in the IoT network to 50 and the message sizes to 4096 bytes, which to some extent limits the deployment capabilities and feasibilities. The proposal in [24] contains an agreement scheme between a symmetric key‐based authentication and the session key. The authentication proposal scheme in [25] uses the symmetric key for wireless sensor networks claiming a 52.63% efficiency. The proposal has leverage over the Denial‐of‐Service (DoS) attacks and user traceability but attracts high computation costs. Shah and Venkatesan [26] proposed a secure vault system, which basically contains equal‐sized keys. This vault aids in a multi‐key‐based mutual authentication process between the IoT device and the IoT server. Every successful communication changes the contents of the vault whose initial content had been shared between the two stakeholders – IoT device and the server. Mohammad Wazid et al. presented a proposal for resource‐constrained smart devices [27], which focus on a one‐way hash function, symmetric encryption and decryption processes, and bitwise exclusive or (XOR) operation. The scheme has been supported through the popular real‐or‐random (ROR) model.
The final dimension shown in Figure 1.1 is the biometric‐based authentication system. This paper also focuses on this branch of authentication techniques. In this system, primarily the client nodes access the IoT application server located at different geographical locations. The client nodes have certain vulnerabilities as follows:
user impersonation attack
offline password guessing attack
user‐specific key theft attack
insider attack.
All the above attacks can be efficiently addressed via biometric authentication, which has already proven to be reliable in contrast to the traditional password‐dependent authentication schemes. IoT security has been addressed by “lightweight biometric multi‐factor remote user authentication” in [28], which employed a gateway node for initial user registration, subsequently connecting the sensor node via the smart device. Other such proposals working in a similar architecture include those in [29,30] but generating high message exchange cost and insufficient security advancement. Huang et al. have combined biometric, smart card, and password techniques for designing the authentication framework but lack a threat identification network [31]. Authentication systems for wireless sensor networks have been proposed in [32] and for cloud‐based biometric identification via Cloud‐Id in [33] but at the cost of high computational complexity similar to the process proposed in [34]. Other notable proposals include [35] where Rivest–Shamir–Adleman (RSA) and Blowfish algorithms are employed to generate an algorithm, which takes care of secure communication of the message packets over the internet. Punithavathi et al. [36,37] have proposed a template bases enrolment and authentication framework, which addresses the issues of availability, integrity, and confidentiality via data integrity, data encryption, and session key agreement. Although the proposal in [36,37] has been accepted to be highly advanced in contrast to traditional biometric techniques, the major challenge exists in the deployment of such a system in the cloud environment.
Securing the IoT devices and corresponding networks is termed IoT Security. Enabling security features at the device end still calls for extensive research and applications/algorithms in this domain are heavily appreciated. In addition to the security features, malware installation on the devices forms another issue of huge concern as it has network infection capabilities. Therefore, IoT security is an area of great importance in the advanced technology domain. Security requirements in IoT infrastructure are alphabetically defined in Table 1.1 and security threats are shown in Table 1.2.
1.3 Proposed Framework
The authentication process is based on a key‐based biometric encryption technique in our proposal. The fingerprint image generated by the biometric is taken as the input from the user. These fingerprint images are fed into a cloud‐based database during the enrolment process. These fingerprint images generate the dynamic cipher key for the AES encryption process. During the authentication phase, users provide their biometric fingerprint images. This image is pre‐processed to enhance the image quality for feature (minutiae) extraction. The extracted minutiae points are used to generate the cipher key for the encryption process. The input fingerprint image and a stored fingerprint image of an enrolled user are encrypted using this cipher key and the resultant cipher texts are matched. Results of the matching determine the authentication of a user. The framework for our proposal is presented in Figure 1.2.
1.3.1 Enrolment Phase
The process flow for the Enrolment phase is presented in Figure 1.3. Algorithm 1.1 presents the algorithm for this phase. During this phase, the biometric sensor acquires the fingerprint image of the user. The fingerprint image captured using the biometric sensor is used as the input data and stored in the cloud‐based database as discussed earlier.
Table 1.1 Key terms of IoT security.
Term
Responsibilities
Anonymity
The attacker cannot access the authentic information
Authentication
It helps in the verification of the IoT device communication
Authorization
It is responsible for ensuring the authorization of the various resource devices of the communicating parties
Availability
The IoT devices are available to the users on the requirement
Confidentiality
It helps in preventing unauthorized access and data protection
Data freshness
This adheres to the fact that the dataset transmitted by the IoT device is unique with no repetition
Forward secrecy
If a message is compromised, then this adheres that random session‐dependent public keys are generated to avoid compromising the other messages
Integrity
The data acquired from the IoT sensor remain unaltered during transmission and also un‐compromised at the warehouse
Non‐repudiation
