Ubiquitous Security. Second International Conference, UbiSec 2022 Zhangjiajie, China, December 28–31, 2022 Revised Selected Papers

Description

Preface Organization Contents Cyberspace Security Support Tool Selection in Digital Forensics Training 1 Introduction 2 Background 3 Related Work 4 Support Tool Selection in Digital Forensics Training 4.1 Layers of Abstraction 4.2 Typology 4.3 Decision-Making with AHP 5 Applying the Concept to a Use-Case 5.1 Use Case Description 5.2 Input for Abstraction Layer Concept 5.3 AHP Calculation 5.4 Results Interpretation of the AHP 5.5 Output of Abstraction Layer Concept 6 Discussion 7 Conclusion and Future Work References Listen to the Music: Evaluating the Use of Music in Audio Based Authentication 1 Introduction 2 Background and Literature Review 3 Source Audio Selection 4 Feature Extraction 5 Implementation 5.1 Source Audio Generation 5.2 Devices Under Test 5.3 Experiment Architecture 5.4 Environment 6 Classification and Evaluation 7 Evaluating the Effect of Noise 8 Results 8.1 Source Audio Comparison 8.2 Feature Comparison 8.3 Impact of Phones Registered 8.4 Impact of Training Samples 8.5 Impact of Noise 9 Conclusion References A Hybrid Secure Two-Party Protocol for Vertical Federated Learning 1 Introduction 2 Related Work 3 Preliminaries 3.1 Split Learning 3.2 Arithmetic Sharing 3.3 Garbled Circuit 4 Overview 4.1 System Model 4.2 Threat Model 4.3 Objective 5 The Proposed Scheme 5.1 Linear Protocol 5.2 Non-linear Protocol 6 Evaluation 7 Conclusion References Detecting Unknown Vulnerabilities in Smart Contracts with Multi-Label Classification Model Using CNN-BiLSTM 1 Introduction 2 Related Work 3 Implementation 4 Evaluation 5 Conclusion References CATS: A Serious Game in Industry Towards Stronger Cloud Security 1 Introduction 2 Related Work 3 Method 4 Design and Implementation 4.1 Overview of CATS 4.2 Design of CATS 4.3 Implementation of CATS 5 Design Evaluation 5.1 Game Dynamic Evaluation 5.2 Questionnaire and SSI Evaluation 5.3 Evaluation from Open Discussion and Open-Ended Questions in SSI 6 Discussion 7 Conclusion References Automated Vulnerability Detection in Source Code Using Quantum Natural Language Processing 1 Introduction 2 Literature Review 3 Methodology 3.1 Dataset Specification 3.2 Input Representation 3.3 Classification Models 3.4 Evaluation Metrics 4 Result and Discussion 5 Conclusion References System Call Processing Using Lightweight NLP for IoT Behavioral Malware Detection 1 Introduction 2 Previous Work 2.1 Behavioral Malware Detection 2.2 IoT Background 3 Malware 3.1 Advanced Persistent Threat 3.2 Denial of Service 4 Data Collection and Processing 4.1 Initial Data Processing 4.2 Data Transformation Using NLP 5 Experimental Results 5.1 Logistic Regression 5.2 Shallow Neural Network 6 Conclusions 6.1 Future Work References Vulnerability Detection with Representation Learning 1 Introduction 2 Related Works 2.1 Vulnerability Detection Based on Software Metrics 2.2 Machine Learning-Based Vulnerability Detection 2.3 Deep Learning-Based Vulnerability Detection 3 Methodology 3.1 Overview 3.2 Representation Learning 3.3 Ensemble Learning with Neural Representations 4 Empirical Evaluation 4.1 Data Introduction 4.2 Evaluation Metrics 4.3 Experiment Settings and Environment 5 Conclusion References Assessing Vulnerability from Its Description 1 Introduction 2 Background 2.1 Common Vulnerability Scoring System (CVSS) 2.2 Term Frequency Inverse Document Frequency and Support Vector Machine 2.3 Universal Sentence Encoder 2.4 Generative Pre-trained Transformer 3 3 Related Works 3.1 Machine Learning for CVSS Prediction 3.2 Conversion from CVSS V2.0 to CVSS V3.1 4 Methodology 4.1 Data 4.2 Tasks 4.3 Training 4.4 Evaluation 5 Empirical Evaluation Results 5.1 Predicting Common Vulnerability Scoring System 2.0 5.2 Predicting Common Vulnerability Scoring System 3.1 6 Discussion 7 Conclusion References Malware Traffic Classification Based on GAN and BP Neural Networks 1 Introduction 2 Related Work 2.1 Network Traffic Detection Technology 2.2 Generative Adversarial Network 3 Model 3.1 Combined Model 3.2 Design of the Generative Adversarial Network 4 Experiments 4.1 Datasets 4.2 Data Preprocessing 4.3 Settings 4.4 Results and Analysis 5 Summary References Source Code Vulnerability Detection Using Deep Learning Algorithms for Industrial Applications 1 Introduction 2 State-of-the-Art 2.1 Open Source Software Vulnerability Detection for C# 2.2 Deep Learning Software Vulnerability Detection Method 3 Proposed Solution 3.1 Vulnerability Detection Data Set 3.2 Deep Learning Vulnerability Detection 4 Experimental Results and Discussions 5 Conclusions and Future Work References Detecting Unknown Vulnerabilities in Smart Contracts with Binary Classification Model Using Machine Learning 1 Introduction 2 Related Work 3 System Model for Unknown Vulnerability Detection 3.1 Data Collection 3.2 Data Pre-processing 3.3 Model Training 3.4 Vulnerability Detection 4 Detailed Design 4.1 Scheme Process 4.2 N-gram Model 4.3 Vector Weight Penalty Mechanism 5 Experiment 5.1 Experimental Dataset 5.2 Evaluation Metrics 5.3 Experimental Results and Analysis 6 Conclusions References Prototyping the IDS Security Components in the Context of Industry 4.0 - A Textile and Clothing Industry Case Study 1 Introduction 2 Related Work 3 IDS Security Architecture and Components 4 Prototype Implementation 4.1 Requirements and Proposed Architecture 4.2 Components Configuration 4.3 Communication Testing 5 Discussion 6 Conclusions References An Aspect-Based Semi-supervised Generative Model for Online Review Spam Detection 1 Introduction 2 Related Work 3 Proposed Method 3.1 Problem Statement 3.2 The Solution 3.3 Aspect Level Review and Product Information Embedding 3.4 ACVAE: Spam Detection Based on CVAE at the Aspect Level 4 Experiment 4.1 Datasets and the Evaluation Metrics 4.2 Baselines 5 Result and Analysis 6 Case Study 7 Conclusion References Hierarchical Policies of Subgoals for Safe Deep Reinforcement Learning 1 Introduction 2 Related Work 3 Method 3.1 Reinforcement Learning 3.2 Subgoal Embedding 4 Experiments 4.1 Environment Set Up 4.2 Subgoal Embedding in Reinforcement Learning Algorithm 5 Discussion 6 Conclusion References Improved DeepLabV3+ based Railway Track Extraction to Enhance Railway Transportation Safety 1 Introduction 2 Methods in This Paper 2.1 Algorithm Flow 2.2 Mobilenetv3 Network 2.3 Channel Attention Mechanism 2.4 Network Structure and Algorithm of This Paper 3 Experimental Data and Evaluation Indicators 3.1 Experimental Data Set 3.2 Experimental Environment and Parameter Setting 3.3 Evaluating Indicator 4 Experimental Data and Evaluation Indicators 4.1 Analysis of Ablation Experiment 4.2 Visual Analysis of Loss Function 4.3 Analysis of Model Comparison Experiment 5 Conclusion References Analysis of Techniques for Detection and Removal of Zero-Day Attacks (ZDA) 1 Introduction 1.1 Zero-Day Attacks Process 1.2 Conventional Security 2 Literature Review 3 Research Methodology 4 Results and Discussion 4.1 Analysis 4.2 Discussion 5 Conclusion 6 Proposed Model References Android Malware Detection: A Literature Review 1 Introduction 2 Android Malware Detection 3 Android Malware Analysis Approaches 3.1 Static Approach 3.2 Dynamic Approach 3.3 Hybrid Approach 3.4 Other Approaches 4 Malware Detection: Challenges and Research Directions 5 Conclusion References Threat Modeling in Cloud Computing - A Literature Review 1 Introduction 2 Background 2.1 Cloud Computing 2.2 Threat Modeling 3 Threat Modeling: Approaches and Methods 3.1 Threat Modeling Language-Based Approaches 3.2 Other Threat Modeling Approaches 4 Discussion and Future Directions 5 Conclusion References A New Signal Packing Algorithm for CAN-FD with Security Consideration 1 Introduction 2 System Model and Security Model 2.1 System Model 2.2 Security Model 3 MILP-Based Packing Algorithm 3.1 MILP Formulation 3.2 Execution Steps of CSLLP 4 Simulated-Annealing-Based Algorithm 5 Experimental Results 6 Conclusion References An Adversarial Sample Defense Method Based on Saliency Information 1 Introduction 2 Related Work 2.1 Existing Attack Methods 2.2 Existing Defense Methods 2.3 Salient Feature Extraction Methods 3 Our Approach 3.1 Pipeline 3.2 Notation 3.3 Learning Objective 4 Experimental Results and Analysis 4.1 Details 4.2 Algorithm Structure 4.3 Qualitative Experiments 4.4 Comparative Experiments 4.5 Ablation Study 5 Conclusion References BlockLearning: A Modular Framework for Blockchain-Based Vertical Federated Learning 1 Introduction 2 Related Work 3 BlockLearning Framework's Design 3.1 Structure and Modules 4 BlockLearning Framework's Implementation 5 Experimental Setup and Evaluation 6 Results and Discussion 7 Conclusions and Future Directions References Cyberspace Privacy An Interactive Query Differential Privacy Protection Model Based on Big Data Analysis 1 Introduction 2 Related Work 3 Redundant Track Data Deletion Algorithm 4 Matrix Decomposition Method Based on Combinatorial Properties 5 Adaptive Noise Tracking Data Protection 6 Simulation Experiments 6.1 Time Complexity 6.2 Information Theft Rate 6.3 Information Loss Rate 7 Conclusion References Decentralized Collaborative Filtering Algorithm with Privacy Preserving for Recommendation in Mobile Edge Computing 1 Introduction 2 Related Work 3 Three Observations on Decentralized Recommendation in MEC 3.1 From Low Latency to Ultra-Low Latency 3.2 From Resource Wasting to High Utilization 3.3 From High Risk to Privacy Preserving 4 Decentralized Collaborative Filtering Algorithm 5 Conclusion References Encryption Proxies in a Confidential Computing Environment 1 Introduction 2 Background and Project Overview 2.1 Intel SGX 2.2 SCONE 2.3 Eperi Gateway 3 Implementation 3.1 Eperi with SGX 4 Experimental Setup 5 Results 5.1 Latency 5.2 Throughput 5.3 Security Requirements 6 Discussion 7 Conclusion 8 Future Work References FedTA: Locally-Differential Federated Learning with Top-k Mechanism and Adam Optimization 1 Introduction 2 Preliminary 2.1 Local Differential Privacy 2.2 Perturbation Mechanism 2.3 Different Optimizers 3 Top-k + Adam 3.1 Top-k Mechanism 3.2 Adam Mechanism 3.3 Top-k + Adam 4 Experiment 5 Conclusion References Differentially Private Clustering Algorithm for Mixed Data 1 Introduction 2 Related Work 3 Differentially Private Mixed Data Clustering Algorithm 3.1 Overview of the DPMC Algorithm 3.2 Differential Privacy Protection 3.3 Adaptive Privacy Budget Allocation 3.4 Optimization Based on Consistency Inference 4 Experiments 4.1 Data Set and Parameter Setting 4.2 Performance Evaluation 5 Conclusion References Impact of Reenactment Programs on Young Generation 1 Introduction 1.1 Objective of Study 2 Related Work 2.1 Research Gap 3 Research Design and Methodology 3.1 Problem Identification 3.2 Data Collection and Sampling 3.3 Measurement Instruments 3.4 Evaluation 4 Experimentation 4.1 Violence 4.2 Language 4.3 Seduction 4.4 Instigation 4.5 Fear 5 Results and Discussion 5.1 Discussion 6 Conclusion References Sensor Cloud Data Privacy Protection Model Based on Collaborative Deep Learning 1 Introduction 2 Algorithm 3 Experiment 4 Conclusion and Discussion References Cyberspace Anonymity An Improved Cuckoo Search Algorithm and Its Application in Function Optimization 1 Introduction 2 Introduction of the CS Algorithm 2.1 Algorithm Principle 2.2 Levy Flight 2.3 Random Migration 2.4 Advantages and Disadvantages of the Algorithm 3 Algorithm Optimization Strategies 3.1 Opposition-Based Learning Strategy 3.2 Dynamic Adjustment Strategy of Inertia Weight 3.3 Local Exploitation Strategy 4 Description and Analysis of ECSOW Algorithm 4.1 Algorithm Description 4.2 Time Complexity Analysis 5 Experiments 5.1 Experimental Environment and Test Functions 5.2 Experimental Results and Analysis 6 Conclusion References Empirical Likelihood for PLSIM with Missing Response Variables and Error-Prone Covariates 1 Introduction 2 Methodology and Result 2.1 Empirical Likelihood 2.2 Asymptotic Result 3 Numerical Examples 3.1 Simulation 3.2 A Real Data Example 4 Proofs 5 Discussion References High-Speed Anonymous Device Authentication Without Asymmetric Cryptography in the Internet-of-Things 1 Introduction 1.1 Our Contribution and Outline of the Paper 2 Related Work 3 Notation and Preliminaries 3.1 Notation 3.2 Cryptographic Primitives 4 High-Speed Authentication Protocols 4.1 One-Message Authentication with Anonymity 4.2 Authentication with Data Transmission 5 Implementation and Evaluation 5.1 Experiment Configuration 5.2 Result Summary 6 Conclusion and Future Work References A-VMD: Adaptive Variational Mode Decomposition Scheme for Noise Reduction in Sensor-Cloud 1 Introduction 2 System Model 2.1 Principle of VMD Algorithm 2.2 Sample Entropy 3 A-VMD Algorithm 3.1 Number of Decomposition Modes 3.2 Quadratic Penalty Factor 3.3 Component Selection and Processing of Reconstructed Signals 3.4 A-VMD Algorithm Flow 4 Noise Reduction Comparison 4.1 Simulation Data Analysis 4.2 Measured Data Analysis 5 Conclusion References A Thermal-Aware Scheduling Algorithm for Reducing Thermal Risks in DAG-Based Applications in Cyber-Physical Systems 1 Introduction 2 Related Work 3 System Model 3.1 Application Model 3.2 Power Model 3.3 Thermal Model 3.4 Threat Model 4 Proposed Algorithm 5 Experimental Setup 6 Result and Discussion 7 Conclusion References Short Papers Garbage Recognition Algorithm Based on Self-attention Mechanism and Deep Sorting 1 Introduction 2 Related Work 2.1 YOLOv5s 2.2 CA Self-attention Mechanism 2.3 DeepSort Algorithm 3 Improvement of YOLOv5s Network Structure 3.1 DeepSort and CA Algorithm Combined with YOLOv5s 4 Experimental Setup and Results Analysis 4.1 Dataset Settings 4.2 YOLOv5s Network Training 4.3 YOLOv5s Ablation and Comparative Test 4.4 Algorithm Test Results Analysis 5 Conclusion References Approaches for Zero Trust Adoption Based upon Organization Security Level 1 Introduction 2 Related Work 3 Background 4 Zero Trust Adoption and Deployment 4.1 Organizational Categories 4.2 Zero Trust Approaches 4.3 Mapping Zero Trust to Organization Categories 5 Zero Trust and the Cloud 6 Conclusion References Multi-Mobile Agent Security by Design Itinerary Planning Approach in Wireless Sensor Network 1 Introduction 2 Multi-mobile Agent Itinerary Planning in WSN 2.1 Dynamic Itinerary Planning 2.2 Data Security 2.3 Limitations of Prior Work 3 A Dynamic Multi-mobile Agent Itinerary Planning Approach in WSNs 3.1 Lightweight AES Implementation 3.2 Proposed Grouping Mechanism 3.3 A Secure Dynamic Itinerary Planning Approach 4 Conclusion and Future Work References On the Variability in the Application and Measurement of Supervised Machine Learning in Cyber Security 1 Introduction 2 Method 3 Literature Review 3.1 Supervised Machine Learning and Measurements 3.2 ML Applications in the Field of Cyber Security 4 Discussion 5 Conclusion Appendix References Author Index