Smart Cities Performability, Cognition, & Security (eBook)

Preface 7
Contents 8
About the Editor 10
1 An Effective Design for Polar Codes over Multipath Fading Channels 11
1.1 Introduction 11
1.2 System Model and Preliminaries 14
1.3 Proposed Transceiver Design 17
1.4 Simulation Results 19
1.5 Conclusion 23
References 24
2 LearningCity: Knowledge Generation for Smart Cities 26
2.1 Introduction 26
2.2 Previous Work 29
2.3 Data Annotation in Smart Cities: Challenges 31
2.3.1 Use Cases 33
2.4 Architecture 34
2.5 Implementation 35
2.5.1 Communication and Frameworks Used 36
2.5.2 Machine-Learning Frameworks 36
2.5.3 Knowledge Warehouse 37
2.5.4 End-User Interfaces and Integration with Existing Tools 38
2.6 Results and Discussion 41
2.7 Conclusions and Future Work 47
References 48
3 Deep Reinforcement Learning Paradigm for Dense Wireless Networks in Smart Cities 51
3.1 Introduction 51
3.1.1 Motivation 51
3.1.2 Scope of the Chapter 52
3.1.3 Contributions of the Chapter 53
3.2 Preliminaries 53
3.2.1 IEEE 802.11ax High Efficiency WLAN (HEW) 54
3.2.2 MAC Layer Resource Allocation in IEEE 802.11 Wireless Networks 55
3.2.2.1 MAC Layer Coordination Functions 56
3.2.3 Problem Statement 58
3.3 Deep Reinforcement Learning Paradigm 60
3.3.1 Deep Reinforcement Learning 60
3.3.2 Q-Learning as a MAC-RA Paradigm 62
3.3.2.1 Q-Learning Algorithm 62
3.3.2.2 Scope and Limitations of QL 64
3.4 Intelligent Q-Learning-Based Resource Allocation (iQRA) 64
3.4.1 Channel Observation-Based Scaled Backoff (COSB) Mechanism 65
3.4.2 *10pt 67
3.5 Performance Evaluation 70
3.5.1 Simulation Scenarios and Parameters 70
3.5.2 Throughput 72
3.5.3 Average Channel Access Delay 72
3.5.4 Fairness 73
3.5.5 Network Dynamicity 74
3.5.6 Distance-Based Rate Adaptation Models 75
3.6 Conclusion 76
Appendix: SCI/SCIE Journal Publications Related to the Chapter 77
References 77
4 Energy Demand Forecasting Using Deep Learning 79
Introduction to Machine Learning 79
Artificial Neural Network 83
Learning Process in ANN 86
Deep ANN 90
Recurrent Neural Network 90
Deep LSTM RNN 94
Modeling Time Series Events 95
Time Series Decomposition Procedure 96
Energy Demand Analysis by Time Series Decomposition 97
Energy Demand Forecasting Using Decomposed Series 100
LSTM Deep Learning Model for Energy Demand Prediction 101
Autoencoder Deep Neural Network 104
Training Process of LSTM Deep Learning Model 104
Evaluation Process of the LSTM Deep Learning Model 106
Testing Process of the LSTM Deep Learning Model 107
Deep Learning Model as a Cloud Solution for Smart Cities 108
References 110
5 RETRACTED CHAPTER: Context-Aware Location Recommendations for Smart Cities 113
Abbreviations 113
Introduction 113
Existing Frameworks 115
Proposed Work 116
Discussion 120
Conclusion 120
References 121
6 Fractional Derivatives for Edge Detection: Application to Road Obstacles 123
Introduction 123
Overview of the Fractional Calculus 125
Edge Detection Techniques 126
Conventional Methods 127
Fractional Methods Implementation 129
Road Obstacle Detection 136
Results Discussion 139
Applications 141
Conclusion and Future Work 142
References 142
7 Machine Learning Parameter Estimation in a Smart-City Paradigm for the Medical Field 146
Introduction 146
Methodology 148
Gaussian Mixture Model 149
Maximum Likelihood Parameter Estimation 149
Expectation Maximization Algorithm 150
Support Vector Machine 152
Results and Discussion 152
Classifier Performance 152
Contingency Table 153
Conclusion 156
References 156
8 Open Source Tools for Machine Learning with Big Data in Smart Cities 159
Introduction 159
Big Data 160
Machine Learning in Big Data 161
Supervised Learning Algorithms 162
Unsupervised Learning Algorithms 162
Semi-supervised Learning Algorithms 162
Data Availability 163
Batch Learning 163
Online Learning 163
Open Source Tools for Big Data 164
Hadoop Ecosystem 164
Storage Layer 165
Processing Layer 166
Management Layer 169
Machine Learning Toolkits 169
Mahout 170
MLLib 170
H2O 170
Samoa 170
Data Movement and Integration Tools 171
Kafka 171
Flume 171
Sqoop 171
Hive 171
Open Research Issues 172
Conclusion 173
References 173
9 Identity Verification Using Biometrics in Smart-Cities 175
Acronyms 175
Introduction 175
Highlights of the Current Approach 177
Potential Benefits of Utilizing Periocular Region as a Useful Biometric Trait 177
Potential Periocular Sub-Region 178
Computationally Efficient Variation of LBP 180
Bit-Plane Representation of the Original Image 180
Feature Extraction Using Dominant Bit-Plane LBP 181
Segmentation of LCPR 181
Construction of Dominant Bit-Plane 184
Bit-Plane 5 Selection Justification Using Structural Similarity Index 186
Bit-Plane 5 Selection Justification Using Texture Information 187
Dominant Bit-Plane LBP Feature Extraction Using Radial Filters 189
Determination of Dominant Bit-Plane LBP Feature Vectors 192
Experiments 195
Experimental Datasets 195
UBIRISv2 Dataset 196
High Resolution Images 196
Low Resolution Images 197
Results and Discussion 198
Experimental Validation of Dominant Bit-Plane 198
Authentication Accuracies 199
Comparison of Entire Face, Periocular Region, and LCPR Using DB-LBP 201
Conclusion 203
References 203
10 Network Analysis of Dark Web Traffic Through the Geo-Location of South African IP Address Space 206
Introduction 206
Related Research 207
Step 1: Finding the Anonymity Set 208
Anonymous Network Communication Systems 209
The Deep and Dark Web Defined 210
Accessing and Navigating the Dark Web 211
What Are the Uses of the Dark Web? 212
The Impact of Anonymous Communication Networks on Cyber Security in Smart Cities 212
Research Methodology 213
Research Question 213
Research Taxonomy 213
Experiment Design 214
Configuration of a Private TOR Network 216
Network Layout 216
TOR Configuration 216
Data Collection Methodology 217
Results 217
TOR Usage 217
Final Observation 218
Conclusion 222
References 223
11 LBCLCT: Location Based Cross Language Cipher Technique 225
Nomenclature 225
Introduction 226
Infrastructure as a Service 226
Platform as a Service 226
Software as a Service 226
Deployment Models for Cloud Architecture Solution 227
Private Cloud 227
Community Cloud 227
Public Cloud 227
Hybrid Cloud 227
Cryptography 228
Literature Survey 228
Methodology Adopted 230
Encryption Using Affine Cipher 230
Encryption Using Rail Fence Cipher 230
Translation Mapping 231
Proposed Algorithms/Pseudocode 231
Simulation and Result 234
Conclusion and Future Scope 236
References 237
Retraction Note to: Context-Aware Location Recommendations for Smart Cities 239
Index 240