Deep Learning for Security and Privacy Preservation in IoT

Contents
About the Editors
1 Metamorphosis of Industrial IoT using Deep Leaning
1.1 Introduction
1.2 Existing framework
1.3 Layered Architecture of IoT and IIoT Systems
1.4 Requirements of Access Control
1.4.1 Approaches of Different Access Control
1.4.2 Capability-Based Access Control
1.4.3 Role-Based Access Control
1.4.4 Usage-Based Access Control
1.4.5 Attribute-Based Access Control
1.5 Deep Learning Method
1.5.1 Why IIoT Needs Deep Learning
1.6 Detail Review on IIoT Systems
1.6.1 Access Control in Industrial IoT
1.6.2 Deep Learning in Industrial IoT Systems
1.7 An Analyzing Framework of IIoT
1.7.1 Existing Taxonomy Review
1.7.2 Devices Characteristics
1.7.3 Location
1.7.4 Technology
1.7.5 User
1.8 Limitations
1.9 Future Scope
1.10 Conclusion
References
2 Deep Learning Models and Their Architectures for Computer Vision Applications: A Review
2.1 Introduction
2.1.1 Motivation for Deep Learning
2.2 Preliminaries
2.3 Models of Deep Learning
2.3.1 Categorization of Deep Neural Networks
2.3.2 Taxonomy of Convolution Neural Network
2.3.3 ZF Net
2.3.4 Region Based CNN (RCNN)
2.3.5 Mask RCNN
2.3.6 Taxonomy of Sequential Neural Network
2.3.7 Taxonomy of Deep Generative Adversarial Network
2.3.8 Taxonomy of Graph Neural Networks (GNNs)
2.4 Applications in Multimedia and Internet of Things
2.5 Conclusion
References
3 IoT Data Security with Machine Learning Blckchain: Risks and Countermeasures
3.1 Introduction
3.2 Challenges in IoT
3.2.1 Security Threats
3.2.2 Privacy Threats
3.3 Existing Review Papers Using ML Algorithms as a Solution
3.3.1 Security Efforts
3.3.2 Privacy Efforts
3.4 Existing Review Papers Using Blockchain as a Solution
3.4.1 Security Efforts
3.4.2 Privacy Efforts
3.5 Solutions to IoT Threats
3.5.1 Existing Solutions Using Machine Learning (ML) Algorithms
3.5.2 Existing Solutions Using Blockchain (BC) Technology
3.5.3 Existing Solutions Using Machine Learning (ML) and Blockchain (BC)
3.6 Deep Learning Approaches-Based IoT Data Privacy and Security
3.6.1 Deep Neural Network
3.6.2 Feedforward Deep Neural Network
3.6.3 Recurrent Neural Network
3.6.4 Convolution Neural Network
3.6.5 Restricted Boltzmann Machine
3.6.6 Deep Belief Network
3.6.7 Deep Auto-Encoder
3.6.8 Deep Migration Learning
3.6.9 Self-taught Learning
3.7 Discussion of Taxonomy of Machine Learning Algorithms
3.8 Research Challenges
3.8.1 Challenge to Machine Learning Algorithms in IoT
3.8.2 Challenges to Blockchain in IoT
3.8.3 Challenges to Machine Learning (ML) and Blockchain (BC) in IoT
3.9 Conclusion and Future Work
References
4 A Review on Cyber Crimes on the Internet of Things
4.1 Introduction
4.2 Literature Review
4.3 Types of Cyber Attacks on IoT Devices
4.3.1 Physical Attacks
4.3.2 Encryption Attacks
4.3.3 DoS (Denial of Service) Attack
4.3.4 Firmware Hijacking
4.3.5 Botnet Attack
4.3.6 Man-In-The-Middle Attack
4.3.7 Ransomware Attack
4.3.8 Eavesdropping Attack
4.4 Smart Home and It’s Subsystems
4.5 Challenges for the Current Approaches
4.5.1 Testing Drawbacks
4.5.2 Default Passwords
4.5.3 IoT Ransomware
4.5.4 IoT AI and Automation:
4.5.5 Botnet Attacks
4.6 Future Directions
4.7 Recommendations for Cyber Hygiene of IoT Devices
4.8 Conclusion
References
5 Deep Learning Framework for Anomaly Detection in Iot Enabled Systems
5.1 Introduction
5.2 Related Works
5.3 Materials and Methods
5.3.1 Data Set Description
5.3.2 Deep Learning Framework for Securing IoT Based Network System
5.4 Results and Discussions
5.5 Conclusion
References
6 Anomaly Detection Using Unsupervised Machine Learning Algorithms
6.1 Introduction
6.2 Literature Survey
6.3 Preliminaries and Proposed Architecture
6.3.1 Autoencoder Neural Network
6.3.2 PCA
6.4 Results and Discussion
6.5 Future Scope
6.6 Conclusion
References
7 Game Theory Based Privacy Preserving Approach for Collaborative Deep Learning in IoT
7.1 Introduction
7.2 Background and Related Work
7.2.1 Information Leakage on Deep Learning Models in IoT
7.2.2 Privacy Preserving Deep Learning
7.2.3 Game Theory
7.3 Deep Learning in IoT
7.3.1 Convolutional Neural Network (CNN)
7.3.2 Recurrent Neural Network (RNN)
7.3.3 Generative Adversarial Networks (GAN)
7.3.4 Federated Deep Learning (FDL)
7.4 System Model
7.4.1 Collaborative Deep Learning Model in IoT
7.4.2 Training Cost on Edge Gateways
7.4.3 Rationality Assumption
7.5 The Collaborative Deep Learning Game
7.5.1 Game Theoretic Model
7.5.2 Game Analysis
7.5.3 Fair Collaboration Strategy
7.6 Implementation and Analysis
7.6.1 Data Collection
7.6.2 Data Analysis
7.6.3 Experimental Results
7.7 Conclusion and Future Work
References
8 Deep Learning Based Security Preservation of IoT: An Industrial Machine Health Monitoring Scenario
8.1 Introduction
8.1.1 Deep Learning and IIoT
8.1.2 Security and Risks
8.1.3 Security and Integrity Issues of Rotating Machinery: A Background
8.1.4 Structural Rotor Faults for Case Analysis
8.2 Framework Description
8.2.1 Data Acquisition Module
8.2.2 Feature Processing
8.2.3 Decision-Making Module
8.2.4 Performance Checking Module
8.2.5 Maintenance Planning and Corrective Decision Module
8.3 IIoT Security Issues and Attacks
8.3.1 The Way DL Deal with IIoT Security
8.3.2 Challenges of IIoT Security Implementation
8.3.3 Industrial IoT Security Solutions
8.3.4 SRF Case Study for IIoT Security with DL
8.4 Conclusion
References
9 Deep Learning Models: An Understandable Interpretable Approach
9.1 Introduction
9.1.1 Transparency of Model Used
9.1.2 Functionality of Model Used
9.2 Work Done so Far
9.3 Deep Learning Methods
9.3.1 Convolutional Neural Networks (CNN)
9.3.2 Recurrent Neural Networks (RNN)
9.3.3 DE Noising Auto-Encoder (DAE)
9.3.4 Deep Belief Network (DBN)
9.3.5 Long Short Term Memory (LSTM)
9.4 Importance of Deep Learning
9.5 The Future of Deep Learning
9.5.1 Hybrid Learning
9.5.2 Composite Learning
9.5.3 Reduced Learning
9.6 Machine Learning Versus Deep Learning
9.7 Conclusion
References