Intro
Preface
Organization
Contents
Secure and Efficient Federated Learning by Combining Homomorphic Encryption and Gradient Pruning in Speech Emotion Recognition
1 Introduction
2 Background and Related Works
2.1 Speech Emotion Recognition Using Federated Learning
2.2 Privacy-Preserving Federated Learning
2.3 Communication and Computation-Efficient Federated Learning
3 Application of Secure and Efficient Federated Learning for Speech Emotion Recognition
3.1 Non-functional Requirements of Speech Emotion Recognition Application
3.2 Threat Model

3.3 Proposed Method: SEFL
4 Experimental Results
4.1 Use-Case Description and Simulation Setting
4.2 Privacy Considerations
4.3 Efficiency in Terms of Communication Traffic
4.4 Efficiency in Terms of Computation Time
4.5 Performance Metrics: Accuracy, F1-Score, Precision, and Loss
5 Conclusions and Future Work
References
FedLS: An Anti-poisoning Attack Mechanism for Federated Network Intrusion Detection Systems Using Autoencoder-Based Latent Space Representations
1 Introduction
2 Related Work
3 Methodology
3.1 Threat Model

3.2 Robust Federated Learning for NIDS
3.3 Autoencoder Pretraining Process
3.4 Workflow of FedLS
4 Experiments
4.1 Dataset and Data Preprocessing
4.2 Experimental Settings
4.3 Evaluation Result
5 Conclusion
References
Mitigating Sybil Attacks in Federated Learning
1 Introduction
2 Federated Learning: Defending Against Sybil Poisoning Attacks
2.1 FedSybil Design
3 Security Analysis
3.1 Threat Model
3.2 Attacks and Mitigations
4 Evaluation and Discussion
4.1 Experiment Setup
4.2 FedSybil Evaluation
4.3 FedSybil Under Non-IID Settings

4.4 Single Client Attacks
4.5 Coordinated Attacks
4.6 Scalability
5 Related Work
6 Conclusion
References
Privacy-Preserving Authentication Scheme for 5G Cloud-Fog Hybrid with Soft Biometrics
1 Introduction
2 Background and Related Works
2.1 Background
2.2 Related Works
3 Proposed Scheme
3.1 Registration Phase
3.2 Authentication Phase
3.3 Key Agreement
4 Analysis of Our Scheme
4.1 Performance Analysis
4.2 Security Analysis
5 Experiments and Results
5.1 Experiments and Results Based on Real Dataset

5.2 Experiments and Results Based on Public Datasets
6 Conclusion
References
Obfuscation Padding Schemes that Minimize Rényi Min-Entropy for Privacy
1 Introduction
1.1 Contributions
1.2 Related Work
2 Problem Formalization
2.1 Presentation in Terms of Privacy Leakage
2.2 Why Not Differential Privacy?
2.3 Simplification of the Output Set
3 Algorithms
3.1 Per-Object-Padding Scenario, PopRe
3.2 Per-Request-Padding Scenario, PrpRe
4 Experiments and Comparison
4.1 Brute-Force Tests for Correctness
4.2 Attacker Test for Illustration