Parallel and Distributed Computing, Applications and Technologies. Proceedings of PDCAT 2023

✍ Scribed by Ji Su Park, Hiroyuki Takizawa, Hong Shen, James J. Park

• Publisher: Springer
• Year: 2024
• Tongue: English
• Leaves: 325
• Series: Lecture Notes in Electrical Engineering, 1112
• Category: Library

✦ Table of Contents

Message from the PDCAT 2023 General Chair
Organization
Contents
A Blockchain System for Fake News Detection
1 Introduction
2 The Structure of the Blockchain
3 The Problem of the Byzantine Generals
3.1 A Simplified Solution with Signed Messages
3.2 Selection of the Consensus Algorithm
3.3 Network Effectiveness Testing
4 Conclusion
References
Formalization and Verification of the Zab Protocol Using CSP
1 Introduction
2 Background
2.1 Workflows of the Zab Protocol
2.2 CSP
3 Modeling
3.1 Sets, Messages and Channels
3.2 Overall Modeling
3.3 Discovery Phase
3.4 Synchronization Phase
3.5 Broadcast Phase
4 Verification
4.1 Properties
4.2 Verification and Results
5 Conclusion and Future Work
References
Using MPIs Non-Blocking Allreduce for Health Checks in Dynamic Simulations
1 Introduction and Methods
2 Results
3 Conclusion and Future Work
References
Parallelizable Loop Detection using Pre-trained Transformer Models for Code Understanding
1 Introduction
2 Related Work
2.1 Automatic Parallelization
2.2 Natural Language Processing Deep Learning Model
3 Methodology
3.1 Overview
3.2 Data Collection
3.3 Data Labeling
3.4 Fine-Tuning the CodeT5 Model
4 Evaluation Setup
4.1 Hardware and Software
4.2 Fine-Tuning
4.3 Evaluation Benchmarks
4.4 Baseline Models
4.5 Performance Metrics
5 Evaluation Result
5.1 GitHub Projects
5.2 NAS Parallel Benchmarks
6 Discussion
6.1 Why Can Our Model Predict Correctly?
6.2 Why Does Our Model Mispredict?
7 Conclusion and Future Work
References
SAHF-LightPoseResNet: Spatially-Aware Attention-Based Hierarchical Features Enabled Lightweight PoseResNet for 2D Human Pose Estimation
1 Introduction
2 Related Works
3 Our Proposed SAHF-LightPoseNet
3.1 Enhancing Backbone Model with Modified ResNet and Deconvolution Module
3.2 Amplifying Model Performance with Global Context Blocks
3.3 SAHF Module
3.4 Heatmap Joint Prediction
4 Experimental Setup
4.1 Dataset
4.2 Implementation Details
4.3 Evaluation Metrics
5 Experimental Results and Discussion
6 Conclusion and Future Work
References
List-Based Workflow Scheduling Utilizing Deep Reinforcement Learning
1 Introduction
2 Related Work
3 Deep Reinforcement Learning for Workflow Scheduling
4 Performance Evaluation
5 Conclusions and Future Work
References
Federated Learning for Skin Cancer Classification
1 Introduction
1.1 Background
1.2 Motivations
1.3 Goal
2 Research Method
2.1 Problem Description
2.2 Dataset
2.3 Model
3 Experiments and Results Discussion
3.1 Experimental Design
3.2 The Effects of Federated Learning
3.3 The Effects of Data Augmentation
3.4 The Effects of Data Balancing
4 Conclusion
References
A Task Offloading and Content Caching Strategy for the Internet of Vehicles in Cloud-Edge Environment
1 Introduction
2 Related Work
3 System Model and Problem Formulation
3.1 System Framework
3.2 Execution Time and Energy Consumption Model
3.3 Edge Data Caching Model
3.4 Problem Definition
4 Joint Optimization for Offloading and Content Caching with Traffic Stream Prediction
4.1 FOST Enabled Traffic Stream Prediction
4.2 MOEA/D-Based on Joint Optimization
4.3 Description of the Algorithm
5 Experimental Evaluation
5.1 Experimental Setup
5.2 Comparative Experiments
6 Conclusion
References
Privacy-Preserving Retrieval Scheme Over Encrypted Medical Records with Relevance Ranking
1 Introduction
2 Related Work
3 Spelling Correction on Electronic Medical Records
3.1 Electronic Medical Records Templates
3.2 Spelling Correction Evaluation of EMRs
4 System Construction and Preliminaries
4.1 System Model
4.2 Notations
4.3 Cryptographic Preliminaries
5 Encrypted EMR Searching with Privacy Preserving
5.1 EMR Index Building
5.2 Encrypted EMR Retrieval
6 Security Analysis
7 Performance Test
7.1 Index Building Efficiency
7.2 Trapdoor Generation Time
7.3 Search Efficiency
8 Conclusion
References
A Data-Centric Approach for Efficient and Scalable CFD Implementation on Multi-GPUs Clusters
1 Introduction
2 Numerical Foundations
3 Data-Centric MPI-CUDA CFD Implementation
3.1 Minimum-Cost MPI Transfer Strategy
3.2 Code Refactoring Strategy
3.3 Dual-Mode API for CFD Solver
4 Results and Discussion
5 Conclusion
References
Research on Psychological Testing Methods of Criminal Suspects Based on Multi-features of EEG
1 Introduction
2 Related Work
3 Psychological Testing for Criminal Suspects
4 Data Pre-processing Methods
4.1 ICA De-artifacting
4.2 Wavelet Transform Filtering and Denoising
5 Feature Extraction Methods
5.1 Peak-To-Peak Time Domain Feature Extraction
5.2 Welch Power Spectrum Estimation for Extracting Frequency Domain Features
5.3 Mallat Extracts Time-Frequency Domain Features
6 EEG Signal Processing Process
6.1 Pre-processing
6.2 Multi-feature Extraction
6.3 SVM-Based Classification
7 Summary
References
Insider Trading Detection Algorithm in Industrial Chain Based on Logistics Time Interval Characteristics
1 Introduction
2 Related Work
3 Problem Formulation and Analysis
4 Algorithm Design
4.1 Interval Logistics Index Construction Algorithm
4.2 Interval Anomaly Detection Algorithm for Logistics Characteristics
4.3 Abnormal Logistics Feature Detection Based on Isolation Forest
5 Experiments
5.1 Experimental Environment
5.2 Performance Indices and Benchmark Algorithms
5.3 Experimental Result
6 Conclusions
References
Link Attributes Based Multi-service Routing for Software-Defined Satellite Networks
1 Introduction
2 SDN-Enabled LEO/GEO Satellite Network Model
3 Design of LAODR
3.1 Description of the LEO Satellite Network
3.2 Quantification of Dynamic Link Attributes
3.3 Link-Attributes-Based Multi-service On-Demand Routing
4 Evaluation
4.1 Experimental Setup
4.2 Performance Evaluation Under Different Traffic Loads
4.3 Performance Evaluation of Different Services
5 Conclusion
References
A Fuzzy Logical RAT Selection Scheme in SDN-Enabled 5G HetNets
1 Introduction
2 Related Works
3 Handover Approach in 5G HetNets
3.1 The Integrated 5G MIH-SDN Controller Architecture
3.2 A Proposed Multi-criteria RAT Selection Scheme
4 Performance Evaluation
5 Conclusion
References
SSR-MGTI: Self-attention Sequential Recommendation Algorithm Based on Movie Genre Time Interval
1 Introduction
2 Related Work
3 Problem Description of Movie Genre Time Interval
4 Multi-head Self-attention Mechanism Based on Movie Genre Time Interval
4.1 Embedding Layer
4.2 Multi-head Self-attention Mechanism Layer
4.3 Convolutional Neural Network Layer
5 Model Prediction
5.1 Prediction Layer
5.2 Loss Function
6 Experimental Evaluation
6.1 Experimental Configuration
6.2 Results and Analysis
7 Conclusion
References
Fine Time Granularity Allocation Optimization of Multiple Networks Industrial Chains in Task Processing Systems
1 Introduction
2 Related Work
3 Problem Formulation and Analyses
3.1 Fine Time Granularity Online Task Scheduling Model
3.2 Task Execution Product Agents’ Model
3.3 Problem Formulation
4 Algorithm Design
4.1 Online Scheduling Framework
4.2 Passive Scheduling Strategy
4.3 Online Task Allocation Analysis Module
5 Experiments
5.1 Experimental Settings
5.2 Benchmark Algorithms
5.3 Optimization Objective
5.4 Experimental Results of Passive Scheduling
6 Conclusions
References
ε-Maximum Critic Deep Deterministic Policy Gradient for Multi-agent Reinforcement Learning
1 Introduction
2 Analysis of MADDPG Algorithms
3 Algorithms
4 Experiments
4.1 Environments
4.2 Results Analysis
5 Conclusions
References
Effective Density-Based Concept Drift Detection for Evolving Data Streams
1 Introduction
2 Related Work
3 Preliminaries
3.1 Density-Based Clustering
3.2 Definitions
3.3 Concept Drifting Detection
4 The Proposed Algorithm
4.1 Overall Framework
4.2 Time-Weighted Concept Drift Detection
4.3 Sliding Window Size Calculation
4.4 Prediction on Concept-Feature Classification
4.5 Algorithm Description
5 Experimental Results
5.1 Performance Evaluation
5.2 Result Comparison
6 Conclusion
References
An End-to-End Multiple Hyper-parameters Prediction Method for Distributed Constraint Optimization Problem
1 Introduction
2 Background
2.1 Definition of Optimal Set of Parameters
2.2 Comparison with Different Sets of Parameters
3 Multiply Hyper-parameter Prediction Model
3.1 Preprocessing Module
3.2 Graph Feature Extraction Module
3.3 Hyper-parameter Prediction Module
4 Experimental Results and Analysis
4.1 Experimental Data
4.2 Experimental Results and Analysis
5 Summary
References
Dynamic Priority Coflow Scheduling in Optical Circuit Switched Networks
1 Introduction
2 Motivation
2.1 Drawbacks of diameter-based scheduling
2.2 Drawbacks of Head-of-line blocking
3 System Model
3.1 Switch model
3.2 Coflow Model
4 Problem Formulation
5 Algorithm Design
5.1 Coflow Pre-processing
5.2 Coflow Scheduling on Dynamic Priority
6 Experimental Evaluations
6.1 Simulation Settings
6.2 Simulation Results
7 Conclusion
References
Deep Reinforcement Learning Based Multi-WiFi Offloading of UAV Traffic
1 Introduction
2 Related Work
3 Network model
4 Problem Formulation
5 NAPPO Algorithm
5.1 MDP Model Design
5.2 Algorithm Design
6 Simulation Experiments
6.1 Simulation Environment Setting
6.2 Performance Comparison
7 Conclusion
References
Triple-Path RNN Network: A Time-and-Frequency Joint Domain Speech Separation Model
1 Introduction
2 Triple-path Recurrent Neural Network
2.1 Encoder
2.2 Separator
2.3 Signals Reconstruction
3 Experimental Setup
3.1 Datasets
3.2 Experimental Setup and Training Details
4 Results And Discussions
4.1 Results on WSJ0-2MIX
4.2 Results on WSJ0-3MIX
5 Conclusion
References
Design of Query Based Gallery Selector and Mask-Aware Loss for Person Search
1 Introduction
2 Related Work
2.1 Pedestrian Detection
2.2 IoU design
3 Methods
3.1 Base Model
3.2 Query Based Gallery Selector
3.3 Mask Aware Design
3.4 Loss Function
4 Experimental Comparisons
4.1 Datasets and Evaluation
4.2 Experimental result
5 Conclusions and Future Work
References
A Privacy-Preserving Blockchain Scheme for the Reliable Exchange of IoT Data
1 Introduction
2 Preliminaries
2.1 IoT Systems
2.2 Blockchain Technology
3 Related Work
4 Proposed Scheme
4.1 The Proposed Scheme Phases
4.2 The Proposed Scheme Computation Time
5 Discussion and Analysis
6 Conclusion and Future Work
References
R-RPT-A Reliable Routing Protocol for Industrial Wireless Sensor Networks
1 Introduction
2 Related Work
3 RPL Objective and Functions
3.1 Objective Function
3.2 Minimum Rank with Hysteresis Objective Function (MRHOF)
3.3 Objective Function Zero (OF0)
4 Research Methodology
5 Performance Evaluation
6 Conclusion
References
Action Segmentation Based on Encoder-Decoder and Global Timing Information
1 Introduction
2 Related Works
3 Encoder-Decoder and Global Timing Information Model
3.1 Encoder
3.2 Decoder
3.3 LSTM
3.4 Probability Ensemble
4 Experiment
4.1 Dataset and Implementation Details
4.2 Evaluation Metrics
4.3 Analysis of Results
4.4 Analysis of results
5 Conclusion
References
Security Challenges and Lightweight Cryptography in IoT: Comparative Study and Testing Method for PRESENT-32bit Cipher
1 Introduction
2 Countermeasures and Security Solutions for IoT Systems
2.1 A. Type of Countermeasures to Attacks and Security Solutions in IoT systems
2.2 Requirements and Criteria for Security Solutions in IoT Systems
3 Cryptography Deployments in the IoT
3.1 Protocol Testing Method
3.2 Testing Result
4 Conclusion
References
The Prediction Model of Water Level in Front of the Check Gate of the LSTM Neural Network Based on AIW-CLPSO
1 Introduction
2 LSTM Neural Network Based on Adaptive Inertia Weight Comprehensive Learning Particle Swarm Optimization
2.1 Comprehensive Learning Particle Swarm Optimization
2.2 Adaptive Updating Inertia Weight Strategy
2.3 Adaptive Inertia Weight Comprehensive Learning Particle Swarm Optimization
2.4 LSTM Neural Network Model Based on AIW-CLPSO
3 An Example of Water Level Prediction in front of Chaohu Lake Check Gate
4 Conclusions
References
Author Index