Deep Learning Approaches to Cloud Security

Pramod Singh Rathore, PhD, is an assistant professor in the computer science and engineering department at the Aryabhatta Engineering College and Research Centre, Ajmer, Rajasthan, India and is also visiting faculty at the Government University, MDS Ajmer. He has over eight years of teaching experience and more than 45 publications in peer-reviewed journals, books, and conferences. He has also co-authored and edited numerous books with a variety of global publishers, such as the imprint, Wiley-Scrivener. Vishal Dutt, PhD, received his doctorate in computer science from the University of Madras, and he is an assistant professor in the computer science and engineering department at the Aryabhatta Engineering College in Ajmer, as well as visiting faculty at Maharshi Dayanand Saraswati University in Ajmer. He has four years of teaching experience and has more than 22 publications in peer-reviewed scientific and technical journals. He has also been working as a freelance writer for more than six years in the fields of data analytics, Java, Assembly Programmer, Desktop Designer, and Android Developer. Rashmi Agrawal, PhD, is a professor in the Department of Computer Applications at Manav Rachna International Institute of Research and Studies in Faridabad, India. She has over 18 years of experience in teaching and research and is a book series editor for a series on big data and machine learning. She has authored or coauthored numerous research papers in peer-reviewed scientific and technical journals and conferences and has also edited or authored books with a number of large book publishers, in imprints such as Wiley-Scrivener. She is also an active reviewer and editorial board member in various journals. Satya Murthy Sasubilli is a solutions architect with the Huntington National Bank, having received his masters in computer applications from the University of Madras, India. He has more than 15 years of experience in cloud-based technologies like big data solutions, cloud infrastructure, digital analytics delivery, data warehousing, and many others. He has worked with many Fortune 500 organizations, such as Infosys, Capgemini, and others and is an active reviewer for several scientific and technical journals. Srinivasa Rao Swarna is a program manager and senior data architect at Tata Consultancy Services in the USA. He received his BTech in chemical engineering from Jawaharlal Nehru Technological University, Hyderabad, India and completed his internship at Volkswagen AG, Wolfsburg, Germany in 2004. He has over 16 years of experience in this area, having worked with many Fortune 500 companies, and he is a frequent reviewer for several scientific and technical journals.

Foreword xv

Preface xvii

1 Biometric Identification Using Deep Learning for Advance Cloud Security 1
Navani Siroya and Manju Mandot

1.1 Introduction 2

1.2 Techniques of Biometric Identification 3

1.2.1 Fingerprint Identification 3

1.2.2 Iris Recognition 4

1.2.3 Facial Recognition 4

1.2.4 Voice Recognition 5

1.3 Approaches 6

1.3.1 Feature Selection 6

1.3.2 Feature Extraction 6

1.3.3 Face Marking 7

1.3.4 Nearest Neighbor Approach 8

1.4 Related Work, A Review 9

1.5 Proposed Work 10

1.6 Future Scope 12

1.7 Conclusion 12

References 12

2 Privacy in Multi-Tenancy Cloud Using Deep Learning 15
Shweta Solanki and Prafull Narooka

2.1 Introduction 15

2.2 Basic Structure 16

2.2.1 Basic Structure of Cloud Computing 17

2.2.2 Concept of Multi-Tenancy 18

2.2.3 Concept of Multi-Tenancy with Cloud Computing 19

2.3 Privacy in Cloud Environment Using Deep Learning 21

2.4 Privacy in Multi-Tenancy with Deep Learning Concept 22

2.5 Related Work 23

2.6 Conclusion 24

References 25

3 Emotional Classification Using EEG Signals and Facial Expression: A Survey 27
S J Savitha, Dr. M Paulraj and K Saranya

3.1 Introduction 27

3.2 Related Works 29

3.3 Methods 32

3.3.1 EEG Signal Pre-Processing 32

3.3.1.1 Discrete Fourier Transform (DFT) 32

3.3.1.2 Least Mean Square (LMS) Algorithm 32

3.3.1.3 Discrete Cosine Transform (DCT) 33

3.3.2 Feature Extraction Techniques 33

3.3.3 Classification Techniques 33

3.4 BCI Applications 34

3.4.1 Possible BCI Uses 36

3.4.2 Communication 36

3.4.3 Movement Control 36

3.4.4 Environment Control 37

3.4.5 Locomotion 38

3.5 Cloud-Based EEG Overview 38

3.5.1 Data Backup and Restoration 39

3.6 Conclusion 40

References 40

4 Effective and Efficient Wind Power Generation Using Bifarious Solar PV System 43
R. Amirtha Katesa Sai Raj, M. Arun Kumar, S. Dinesh, U. Harisudhan and Dr. R. Uthirasamy

4.1 Introduction 44

4.2 Study of Bi-Facial Solar Panel 45

4.3 Proposed System 46

4.3.1 Block Diagram 46

4.3.2 DC Motor Mechanism 47

4.3.3 Battery Bank 48

4.3.4 System Management Using IoT 48

4.3.5 Structure of Proposed System 50

4.3.6 Spoiler Design 51

4.3.7 Working Principle of Proposed System 52

4.3.8 Design and Analysis 53

4.4 Applications of IoT in Renewable Energy Resources 53

4.4.1 Wind Turbine Reliability Using IoT 54

4.4.2 Siting of Wind Resource Using IoT 55

4.4.3 Application of Renewable Energy in Medical Industries 56

4.4.4 Data Analysis Using Deep Learning 57

4.5 Conclusion 59

References 59

5 Background Mosaicing Model for Wide Area Surveillance System 63
Dr. E. Komagal

5.1 Introduction 64

5.2 Related Work 64

5.3 Methodology 65

5.3.1 Feature Extraction 66

5.3.2 Background Deep Learning Model Based on Mosaic 67

5.3.3 Foreground Segmentation 70

5.4 Results and Discussion 70

5.5 Conclusion 72

References 72

6 Prediction of CKD Stage 1 Using Three Different Classifiers 75
Thamizharasan, K., Yamini, P., Shimola, A. and Sudha, S.

6.1 Introduction 75

6.2 Materials and Methods 78

6.3 Results and Discussion 84

6.4 Conclusions and Future Scope 89

References 89

7 Classification of MRI Images to Aid in Diagnosis of Neurological Disorder Using SVM 93
Phavithra Selvaraj, Sruthi, M.S., Sridaran, M. and Dr. Jobin Christ M.C.

7.1 Introduction 93

7.2 Methodology 95

7.2.1 Data Acquisition 95

7.2.2 Image Preprocessing 96

7.2.3 Segmentation 97

7.2.4 Feature Extraction 98

7.2.5 Classification 99

7.3 Results and Discussions 100

7.3.1 Preprocessing 100

7.3.2 Classification 103

7.3.3 Validation 104

7.4 Conclusion 106

References 106

8 Convolutional Networks 109
Simran Kaur and Rashmi Agrawal

8.1 Introduction 110

8.2 Convolution Operation 110

8.3 CNN 110

8.4 Practical Applications 112

8.4.1 Audio Data 112

8.4.2 Image Data 112

8.4.3 Text Data 113

8.5 Challenges of Profound Models 113

8.6 Deep Learning In Object Detection 114

8.7 CNN Architectures 114

8.8 Challenges of Item Location 118

8.8.1 Scale Variation Problem 118

8.8.2 Occlusion Problem 119

8.8.3 Deformation Problem 120

References 121

9 Categorization of Cloud Computing & Deep Learning 123
Disha Shrmali

9.1 Introduction to Cloud Computing 123

9.1.1 Cloud Computing 123

9.1.2 Cloud Computing: History and Evolution 124

9.1.3 Working of Cloud 125

9.1.4 Characteristics of Cloud Computing 127

9.1.5 Different Types of Cloud Computing Service Models 128

9.1.5.1 Infrastructure as A Service (IAAS) 128

9.1.5.2 Platform as a Service (PAAS) 129

9.1.5.3 Software as a Service (SAAS) 129

9.1.6 Cloud Computing Advantages and Disadvantages 130

9.1.6.1 Advantages of Cloud Computing 130

9.1.6.2 Disadvantages of Cloud Computing 132

9.2 Introduction to Deep Learning 133

9.2.1 History and Revolution of Deep Learning 134

9.2.1.1 Development of Deep Learning Algorithms 134

9.2.1.2 The FORTRAN Code for Back Propagation 135

9.2.1.3 Deep Learning from the 2000s and Beyond 135

9.2.1.4 The Cat Experiment 136

9.2.2 Neural Networks 137

9.2.2.1 Artificial Neural Networks 137

9.2.2.2 Deep Neural Networks 138

9.2.3 Applications of Deep Learning 138

9.2.3.1 Automatic Speech Recognition 138

9.2.3.2 Electromyography (EMG) Recognition 139

9.2.3.3 Image Recognition 139

9.2.3.4 Visual Art Processing 140

9.2.3.5 Natural Language Processing 140

9.2.3.6 Drug Discovery and Toxicology 140

9.2.3.7 Customer Relationship Management 141

9.2.3.8 Recommendation Systems 141

9.2.3.9 Bioinformatics 141

9.2.3.10 Medical Image Analysis 141

9.2.3.11 Mobile Advertising 141

9.2.3.12 Image Restoration 142

9.2.3.13 Financial Fraud Detection 142

9.2.3.14 Military 142

9.3 Conclusion 142

References 143

10 Smart Load Balancing in Cloud Using Deep Learning 145
Astha Parihar and Shweta Sharma

10.1 Introduction 146

10.2 Load Balancing 147

10.2.1 Static Algorithm 148

10.2.2 Dynamic (Run-Time) Algorithms 148

10.3 Load Adjusting in Distributing Computing 149

10.3.1 Working of Load Balancing 151

10.4 Cloud Load Balancing Criteria (Measures) 152

10.5 Load Balancing Proposed for Cloud Computing 153

10.5.1 Calculation of Load Balancing in the Whole System 154

10.6 Load Balancing in Next Generation Cloud Computing 155

10.7 Dispersed AI Load Adjusting Methodology in Distributed Computing Administrations 157

10.7.1 Quantum Isochronous Parallel 158

10.7.2 Phase Isochronous Parallel 159

10.7.3 Dynamic Isochronous Coordinate Strategy 161

10.8 Adaptive-Dynamic Synchronous Coordinate Strategy 161

10.8.1 Adaptive Quick Reassignment (AdaptQR) 162

10.8.2 A-DIC (Adaptive-Dynamic Synchronous Parallel) 163

10.9 Conclusion 164

References 165

11 Biometric Identification for Advanced Cloud Security 167
Yojna khandelwal and Kapil Chauhan

11.1 Introduction 168

11.1.1 Biometric Identification 168

11.1.2 Biometric Characteristic 169

11.1.3 Types of Biometric Data 169

11.1.3.1 Face Recognition 169

11.1.3.2 Hand Vein 170

11.1.3.3 Signature Verification 170

11.1.3.4 Iris Recognition 170

11.1.3.5 Voice Recognition 170

11.1.3.6 Fingerprints 171

11.2 Literature Survey 172

11.3 Biometric Identification in Cloud Computing 174

11.3.1 How Biometric Authentication is Being Used on the Cloud Platform 176

11.4 Models and Design Goals 177

11.4.1 Models 177

11.4.1.1 System Model 177

11.4.1.2 Threat Model 177

11.4.2 Design Goals 178

11.5 Face Recognition Method as a Biometric Authentication 179

11.6 Deep Learning Techniques for Big Data in Biometrics 180

11.6.1 Issues and Challenges 181

11.6.2 Deep Learning Strategies For Biometric Identification 182

11.7 Conclusion 185

References 185

12 Application of Deep Learning in Cloud Security 189
Jaya Jain

12.1 Introduction 190

12.2 Literature Review 191

12.3 Deep Learning 192

12.4 The Uses of Fields in Deep Learning 195

12.5 Conclusion 202

References 203

13 Real Time Cloud Based Intrusion Detection 207
Ekta Bafna

13.1 Introduction 207

13.2 Literature Review 209

13.3 Incursion In Cloud 211

13.3.1 Denial of Service (DoS) Attack 212

13.3.2 Insider Attack 212

13.3.3 User To Root (U2R) Attack 213

13.3.4 Port Scanning 213

13.4 Intrusion Detection System 213

13.4.1 Signature-Based Intrusion Detection System (SIDS) 213

13.4.2 Anomaly-Based Intrusion Detection System (AIDS) 214

13.4.3 Intrusion Detection System Using Deep Learning 215

13.5 Types of IDS in Cloud 216

13.5.1 Host Intrusion Detection System 216

13.5.2 Network Based Intrusion Detection System 217

13.5.3 Distributed Based Intrusion Detection System 217

13.6 Model of Deep Learning 218

13.6.1 ConvNet Model 218

13.6.2 Recurrent Neural Network 219

13.6.3 Multi-Layer Perception Model 219

13.7 KDD Dataset 221

13.8 Evaluation 221

13.9 Conclusion 223

References 223

14 Applications of Deep Learning in Cloud Security 225
Disha Shrmali and Shweta Sharma

14.1 Introduction 226

14.1.1 Data Breaches 226

14.1.2 Accounts Hijacking 227

14.1.3 Insider Threat 227

14.1.3.1 Malware Injection 227

14.1.3.2 Abuse of Cloud Services 228

14.1.3.3 Insecure APIs 228

14.1.3.4 Denial of Service Attacks 228

14.1.3.5 Insufficient Due Diligence 229

14.1.3.6 Shared Vulnerabilities 229

14.1.3.7 Data Loss 229

14.2 Deep Learning Methods for Cloud Cyber Security 230

14.2.1 Deep Belief Networks 230

14.2.1.1 Deep Autoencoders 230

14.2.1.2 Restricted Boltzmann Machines 232

14.2.1.3 DBNs, RBMs, or Deep Autoencoders Coupled with Classification Layers 233

14.2.1.4 Recurrent Neural Networks 233

14.2.1.5 Convolutional Neural Networks 234

14.2.1.6 Generative Adversarial Networks 235

14.2.1.7 Recursive Neural Networks 236

14.2.2 Applications of Deep Learning in Cyber Security 237

14.2.2.1 Intrusion Detection and Prevention Systems (IDS/IPS) 237

14.2.2.2 Dealing with Malware 237

14.2.2.3 Spam and Social Engineering Detection 238

14.2.2.4 Network Traffic Analysis 238

14.2.2.5 User Behaviour Analytics 238

14.2.2.6 Insider Threat Detection 239

14.2.2.7 Border Gateway Protocol Anomaly Detection 239

14.2.2.8 Verification if Keystrokes were Typed by a Human 240

14.3 Framework to Improve Security in Cloud Computing 240

14.3.1 Introduction to Firewalls 241

14.3.2 Importance of Firewalls 242

14.3.2.1 Prevents the Passage of Unwanted Content 242

14.3.2.2 Prevents Unauthorized Remote Access 243

14.3.2.3 Restrict Indecent Content 243

14.3.2.4 Guarantees Security Based on Protocol and IP Address 244

14.3.2.5 Protects Seamless Operations in Enterprises 244

14.3.2.6 Protects Conversations and Coordination Contents 244

14.3.2.7 Restricts Online Videos and Games from Displaying Destructive Content 245

14.3.3 Types of Firewalls 245

14.3.3.1 Proxy-Based Firewalls 245

14.3.3.2 Stateful Firewalls 246

14.3.3.3 Next-Generation Firewalls (NGF) 247

14.3.3.4 Web Application Firewalls (WAF) 247

14.3.3.5 Working of WAF 248

14.3.3.6 How Web Application Firewalls (WAF) Work 248

14.3.3.7 Attacks that Web Application Firewalls Prevent 250

14.3.3.8 Cloud WAF 251

14.4 WAF Deployment 251

14.4.1 Web Application Firewall (WAF) Security Models 252

14.4.2 Firewall-as-a-Service (FWaaS) 252

14.4.3 Basic Difference Between a Cloud Firewall and a Next-Generation Firewall (NGFW) 253

14.4.4 Introduction and Effects of Firewall Network Parameters on Cloud Computing 253

14.5 Conclusion 254

References 254

About the Editors 257

Index 263