Parallel and Distributed Computing, Applications and Technologies: 22nd International Conference, PDCAT 2021, Guangzhou, China, December 17–19, 2021, ... Computer Science and General Issues)

✍ Scribed by Hong Shen (editor), Yingpeng Sang (editor), Yong Zhang (editor), Nong Xiao (editor), Hamid R. Arabnia (editor), Geoffrey Fox (editor), Ajay Gupta (editor), Manu Malek (editor)

• Publisher: Springer
• Year: 2022
• Tongue: English
• Leaves: 643
• Category: Library

✦ Synopsis

This book constitutes the proceedings of the 22nd International Conference on Parallel and Distributed Computing, Applications, and Technologies, PDCAT 2021, which took place in Guangzhou, China, during December 17-19, 2021.

The 24 full papers and 34 short papers included in this volume were carefully reviewed and selected from 97 submissions. The papers are categorized into the following topical sub-headings: networking and architectures, software systems and technologies, algorithms and applications, and security and privacy.

✦ Table of Contents

Preface
Organization
Contents
Networking and Architectures
Accelerating GPU-Based Out-of-Core Stencil Computation with On-the-Fly Compression
1 Introduction
2 Previous Work
3 Out-of-Core Stencil Computation
4 On-the-Fly Compression
5 Proposed Method
5.1 Separate Compression
5.2 Pipelining cuZFP
6 Experimental Results
6.1 Evaluation of Performance Benefits
6.2 Detailed Analysis of Achieved Performance Improvement
6.3 Evaluation of Precision Loss
7 Conclusions and Future Work
References
Routing with Ant Colony Optimization in Wireless Mesh Networks
1 Introduction
2 Related Work
3 Routing Algorithm
4 Performance Evaluation
5 Conclusion
References
A Light-Weight Scheme for Detecting Component Structure of Network Traffic
1 Background
2 Methodology
2.1 Overview
2.2 Preprocessing
2.3 Traffic Analysis
2.4 Proportion Analysis
3 Evaluation
3.1 Data Set
3.2 Experiment Setting
3.3 Experiment
4 Conclusion
References
Evaluating the Performance and Conformance of a SYCL Implementation for SX-Aurora TSUBASA
1 Introduction
2 NEC SX-Aurora TSUBASA
3 Overview of neoSYCL
4 Evaluation and Discussions
4.1 Conformance Test Cases
4.2 VEO-SYCL-Bench
5 Conclusions
References
Bayesian Optimization-Based Task Scheduling Algorithm on Heterogeneous System
1 Introduction
2 Related Work
3 Problem Description
3.1 Scheduling Model
3.2 Constraints
4 Algorithm
4.1 Scheduling Strategy
4.2 First phase: Bayesian optimization training
4.3 Second Phase: Task Scheduling Simulation
4.4 Theoretical Analysis
5 Experimental Results
5.1 Research Questions
5.2 Baselines
5.3 Experiments
6 Conclusion
References
Optimizing Uplink Bandwidth Utilization for Crowdsourced Livecast
1 Introduction
2 Related Work
3 System Model
3.1 System Overview
3.2 QoE and Bandwidth Cost
3.3 Problem Formulation
4 Upload Bitrate Allocation with Bandwidth Constraints
4.1 Problem Relaxation
4.2 UBA Algorithm Design
4.3 Bubal: UBA Algorithm with Optimal Bandwidth Budget
5 Performance Evaluation
5.1 Experimental Settings
5.2 Performance Comparison
6 Conclusion
References
A Batched Jacobi SVD Algorithm on GPUs and Its Application to Quantum Lattice Systems
1 Introduction
2 Algorithmic Background
2.1 Jacobi Algorithms
2.2 Parallel Blocked Jacobi Algorithms
3 Design Details
3.1 Overall Design
3.2 Kernel Optimization
3.3 Convergence Criterion
4 Experimental Results and Analysis
4.1 Experimental Setup
4.2 Performance Analysis
5 Application to Quantum Lattice Systems
6 Conclusion
References
A Molecular Dynamics Based Multi-scale Platelet Aggregation Model and Its High-Throughput Simulation
1 Introduction
2 A Rigid Platelet Multi-scale Model
2.1 DPD Model for Blood Flow
2.2 CGMD-DPD Model for Fluid-Platelet Interaction
2.3 The Morse Potential for Platelet-Platelet Interaction
3 Parameters Calibration
3.1 Contact Area and Detaching Force
3.2 Parameter Calibration with High-Throughput Simulation
4 Conclusion
References
Approximation and Polynomial Algorithms for Multi-depot Capacitated Arc Routing Problems
1 Introduction
2 Notations
3 The Nonfixed Destination MCARP
4 The (Fixed Destination) MCARP
5 The Multi-depot Rural Postman Problem
6 Multi-depot CARP on a Line
References
Zero-Shot Face Swapping with De-identification Adversarial Learning
1 Introduction
2 Method
2.1 Overall Framework
2.2 De-identification Module
2.3 Attentive Spatial Style Modulation Module
2.4 Mix-Batch Training Strategy
3 Experiments
3.1 Experiment Setup
3.2 Ablation Study
3.3 Comparison
3.4 Facial Expression Translation
4 Conclusion
References
An User-Driven Active Way to Push ACL in Software-Defined Networking
1 Introduction
2 Clack Design
2.1 Overview
2.2 Accessing Pair (AP) Manager
2.3 Access Control List (ACL) Manager
3 Evaluation
4 Conclusion and Future Work
References
Photonic Computing and Communication for Neural Network Accelerators
1 Introduction
2 Optical Implementations for Computing
2.1 Optical Resonance Based Neural Network Accelerators
2.2 Optical Diffraction Based Neural Network Accelerators
2.3 Optical Interference Based Neural Network Accelerators
3 Optical Implementations for Communication
3.1 Off-Chip Communication for Neural Network Accelerators
3.2 On-Chip Communication for Neural Network Accelerators
4 Challenges and Opportunities
5 Conclusion
References
Performance Comparison of Multi-layer Perceptron Training on Electrical and Optical Network-on-Chips
1 Introduction
2 Background
2.1 Training of MLP
2.2 Optical and Electrical On-Chip Interconnects
3 Methodology on Parallel MLP Training on ONoC and ENoC System
3.1 Parallel MLP Training
3.2 Performance Model
3.3 Energy Model
4 Comparison of MLP Training on ENoC and ONoC
4.1 Simulation Setup
4.2 Performance Comparison
4.3 Comparison of Energy Consumption
5 Conclusion
References
The Design and Implementation of Reconfigurable Quaternary Logic Processor
1 Introduction
2 Reconfigurable Quaternary Logic Processor
2.1 General Structure
2.2 Circuit of RQLP's Processor Bit
3 Experiments
4 Conclusions and Future Work
References
A 3D Dubins Curve Constructing Method Based on Particle Swarm Optimization
1 Introduction
2 Problem Formulation
3 Proposed Methods for Path Planning with Dubins Curve
3.1 Smooth Path with Dubins Curve
3.2 Dynamic Programming Algorithm with Turning Radius Constraint
3.3 Improved Partical Swarm Optimization
4 Experiment
4.1 Experimental Set Up
4.2 Experimental Results on Simulated Data
5 Conclusion
References
Software Systems and Technologies
Towards Conflict-Aware Workload Co-execution on SX-Aurora TSUBASA
1 Introduction
2 Resource Conflicts on an SX-Aurora TSUBASA System
3 Performance Interference by Workload Co-execution
3.1 Evaluation Setup
3.2 Interference Evaluation Results
3.3 Avoidance of Performance Interferences
4 Related Work
5 Concluding Remarks
References
A Learning-Based Scheduler for High Volume Processing in Data Warehouse Using Graph Neural Networks
1 Introduction
2 Problem Statement
3 Proposed Approach: Design and Analysis
3.1 Preliminaries
3.2 State Representation
3.3 Graph Embedding
4 Performance Evaluation Results
5 Conclusion
References
Adaptive Updates for Erasure-Coded Storage Systems Based on Data Delta and Logging
1 Introduction
2 Preliminary
3 Proposed Approach DETOG
3.1 File Classification
3.2 File Updates
4 Implementation
5 Experiments
5.1 Trace Evaluations
5.2 Storage Overhead
6 Conclusion
References
Matching Program Implementations and Heterogeneous Computing Systems
1 Introduction
2 Context and Related Work
3 Platform Descriptors
3.1 Programming Models
3.2 Abstraction Levels
3.3 Programming Support
4 Case Study: Improving System Utilization
5 Model Performance Evaluation
6 Conclusion
References
FastDCF: A Partial Index Based Distributed and Scalable Near-Miss Code Clone Detection Approach for Very Large Code Repositories
1 Introduction
2 Related Work
3 Design
3.1 Preliminary Concepts and Definition
3.2 Efficiency and Scalability Limitation of Existing Techniques: Experiments and Analysis
3.3 Design Goals and Our Approaches
4 FastDCF Implementation
5 Evaluation
5.1 Execution Time and Scalability
5.2 Distributed Parallelization
5.3 Precision and Recall
5.4 Multi-granularity Detection
6 Conclusion
References
Towards Optimal Fast Matrix Multiplication on CPU-GPU Platforms
1 Introduction
2 Related Work
3 Method
3.1 Overall Framework
3.2 Winograd
4 Regression Model Predictor
4.1 Depth First'' andBreadth First''
4.2 Strategy Sequence
4.3 Details of the Implementation
5 Experiment
5.1 Experimental Setup
5.2 Performance Evaluation
5.3 Heterogeneous Implementation
5.4 Extension Strategy
5.5 Regression Model Predictor
6 Conclusion
References
Temperature Matrix-Based Data Placement Using Improved Hungarian Algorithm in Edge Computing Environments
1 Introduction
2 Related Work
3 System Model and Problem Definition
3.1 System Framework
3.2 Data Temperature and Calculation
3.3 Network Latency
3.4 System Cost
3.5 Load Balancing
3.6 Problem Definition
4 Algorithm Design
5 Experimental Evaluation
5.1 Experiment Setup
5.2 Experimental Results and Analysis
6 Conclusion and Future Work
References
Realtime Physics Simulation of Large Virtual Space with Docker Containers
1 Introduction
2 Related Work
3 Prototype System
3.1 Overview
3.2 Partitioning into Subspaces and Assigning to Clients
3.3 Distributed Simulation of the Virtual Space
4 Evaluation
4.1 Setup
4.2 Effect of Decentralization for Reducing the Simulation Time
4.3 Overhead of Decentralization
5 Concluding Remarks
References
A Deep Reinforcement Learning-Based Approach to the Scheduling of Multiple Workflows on Non-dedicated Edge Servers
1 Introduction
2 Related Work
3 Problem Modeling
3.1 System Model
3.2 Workflow Model
3.3 Scheduling Model
4 The Prioritized Replay Double DQN Algorithm
4.1 Algorithm Theory
4.2 Algorithm Framework
5 Experiments, Results and Discussion
5.1 Experiments Setup
5.2 Results and Discussion
6 Conclusion
References
A MVCC Approach to Parallelizing Interoperability of Consortium Blockchain
1 Introduction
2 Background
2.1 Hyperledger Fabric
2.2 Related Work
3 Problem Statement
4 Multi-version Merkle Directed Acyclic Graph
5 Algorithms
5.1 Operations on States
5.2 Interoperable Transaction Operation Interfaces
6 Evaluation
7 Conclusion
References
An Effective and Reliable Cross-Blockchain Data Migration Approach
1 Introduction
2 Related Work
3 The Proposed Migration Approach
3.1 Preparing for the Migration
3.2 Cross-Chain Data Migration Process
4 Experimental Results
4.1 Comparison Study of solo and aggr
4.2 Configuration Parameter Study for Aggregation
5 Conclusion
References
Algorithm for the Facility Location Problem with Origin and Destination
1 Introduction
1.1 Related Work
1.2 Our Contribution
1.3 Organization
2 Uncapacitated Facility Location Problem with Origin And Destination
2.1 Problem Statement
2.2 Algorithm
2.3 Analysis
3 Conclusion
References
Reinforcement Learning-Based Auto-scaling Algorithm for Elastic Cloud Workflow Service
1 Introduction
2 Problem Description
3 RL-Based Auto-scaling Algorithm for Elastic Cloud Workflow Service
3.1 System Design
3.2 Algorithm Design
4 Experiment
4.1 Experiment Design
4.2 Experiment Result
5 Conclusion
References
Optimal Energy Efficiency Strategy of mm Wave Cooperative Communication Small Cell Based on SWITP
1 Introduction
2 System Model
2.1 Network Model
2.2 System Model
3 Problem Description and Solution Strategy
4 Experimental Results and Analysis
4.1 Experimental Environment and Parameter Setting
4.2 Performance Analysis and Comparison of Algorithms
5 Conclusions
References
Low Latency Execution Guarantee Under Uncertainty in Serverless Platforms
1 Introduction
2 Problem Statement
2.1 FaaS Platform and Application Structure
2.2 Quality of Service Semantic
3 Design Approach
3.1 Main Components
3.2 Monitoring and System Model
3.3 Optimizer
3.4 Cost-Benefit Analysis (CBA)
4 Performance Evaluation
4.1 Result Summary
4.2 Computational Overhead
5 Conclusion
References
High Resolution Patient-Specific Blood Flow Simulation in a Full-Size Aneurysmal Aorta Based on a Parallel Two-Level Method
1 Introduction
2 Methodology
2.1 Image Segmentation and Mesh Generation
2.2 Governing Equation and Boundary Conditions
2.3 Newton-Krylov-Schwarz Method with a Two-Level Preconditioner
3 Results and Discussion
3.1 Simulation Results and Discussion
3.2 Robustness and Scalability
4 Conclusion
References
Optimizing Data Locality by Executor Allocation in Reduce Stage for Spark Framework
1 Introduction
2 Related Work
3 Executor Allocation Algorithm
3.1 Optimal Executor Allocation Problem
3.2 Approximation Algorithm
4 Experimental Evaluation
4.1 Performance
5 Conclusion
References
TEFRED: A Temperature and Energy Cognizant Fault-Tolerant Real-Time Scheduler Based on Deadline Partitioning for Heterogeneous Platforms
1 Introduction
2 Specifications
3 Proposed Scheduling Scheme
4 Experimental Set Up and Results
5 Conclusion
References
Algorithms and Applications
Social Recommendation via Graph Attentive Aggregation
1 Introduction
2 Related Work
2.1 Graph-Based CF
2.2 Social Recommendation
2.3 Attention
3 Methodology
3.1 Pre-trained Embedding Layer
3.2 Graph Attentive Aggregation Layer
3.3 Preference Prediction
4 Experiment
4.1 Dataset
4.2 Baselines
4.3 Evaluation Metrics
4.4 Parameter Settings
4.5 Experiment Results
4.6 Parallel Efficiency Evaluation
5 Conclusion and Future Work
References
MACSQ: Massively Accelerated DeepQ Learning on GPUs Using On-the-fly State Construction
1 Introduction
2 Related Work
3 MACSQ
4 Implementation Details
5 Evaluation
6 Conclusion
References
Model-Based Multi-agent Policy Optimization with Dynamic Dependence Modeling
1 Introduction
2 Related Work
3 Preliminaries
3.1 MARL
3.2 Model-Based RL
4 The POD2M Method
4.1 Model-Based Policy Optimization
4.2 Dynamic Dependence Modeling
5 Experiments
5.1 Setup
5.2 Results
6 Conclusion
References
Multi-index Federated Aggregation Algorithm Based on Trusted Verification
1 Introduction
2 Related Work
3 TVFedmul
3.1 Weight Calculation
3.2 Aggregation
4 System Model
5 Experimental Validation
5.1 Datasets
5.2 Experimental
6 Conclusion
References
Few-Shot Generative Learning by Modeling Stereoscopic Priors
1 Introduction
2 Method
2.1 Architecture and Notation
2.2 Meta-Learning on Base Classes
2.3 Fine-Tuning on Novel Classes
3 Experiment
3.1 Experimental Setting
3.2 Main Results
4 Conclusion
References
Distributed Fair k-Center Clustering Problems with Outliers
1 Introduction
1.1 Our Contribution
1.2 Organization
2 Related Work
3 Preliminaries
4 Problems and Algorithms
4.1 Fair k-center Problem with Outliers
4.2 Distributed Fair k-Center Problem with Outliers
5 Conclusions
References
Multi-zone Residential HVAC Control with Satisfying Occupants' Thermal Comfort Requirements and Saving Energy via Reinforcement Learning
1 Introduction
2 Related Work
3 Theoretical Background
4 Methodology
4.1 Thermal Comfort Prediction
4.2 Problem Formulation
5 Experiments
5.1 SVR-DNN for Thermal Comfort Prediction
5.2 Performance of Q-Learning, DQN and DDPG
6 Conclusion
References
Approximating BP Maximization with Distorted-Based Strategy
1 Introduction
1.1 Related Work
2 Preliminaries
3 Streaming BP Maximization
3.1 Distorted Threshold Algorithm
3.2 Full Distorted Threshold Algorithm
4 Fairness in BP Maximization
5 Conclusions
References
Streaming Algorithms for Maximization of a Non-submodular Function with a Cardinality Constraint on the Integer Lattice
1 Introduction
1.1 Problem Definition
1.2 Preliminaries
2 The Streaming Algorithm
References
Adaptable Focal Loss for Imbalanced Text Classification
1 Introduction
2 Related Work
2.1 Imbalanced Learning
2.2 Pre-trained Model
3 Method
3.1 Focal Loss
3.2 Adaptable Focal Loss
4 Experiment and Analysis
4.1 Data Sets
4.2 Evaluation Criteria and Parameter Setting
4.3 Experimental Analysis
5 Conclusion
References
Roman Amphitheater Classification Using Convolutional Neural Network and Data Augmentation
1 Introduction
2 The Roman Amphitheater Data Set and Data Preprocessing
2.1 Data Set Collection and Preprocessing
2.2 Data Augmentation
3 CNN Structure and Used Framework
4 Experiments
4.1 Results of Data Augmentation
4.2 Pooling Technique Comparison
4.3 Optimal Structure of CNN
4.4 Classification Performance
5 Conclusion
References
Data-Hungry Issue in Personalized Product Search
1 Introduction
2 Related Work
3 Preliminaries
3.1 Problem Definition
3.2 Non-personalized and Personalized Product Search
4 Data-Hungry Issue
4.1 Does Personalization Cause the Data-Hungry Issue?
4.2 When Does Data-Hungry Issue Degrade the Performance?
4.3 How Does Data-Hungry Issue Degrade the Performance?
5 Experiment
6 Conclusion
References
Jointly Super Resolution and Degradation Learning on Unpaired Real-World Images
1 Introduction
2 Related Work
3 Proposed Method
3.1 Loss Function
3.2 Network Architecture
4 Experiments
4.1 Dataset and Evaluation Metrics
4.2 Training Details
4.3 Result
4.4 Ablation Study
5 Conclusion
References
Enhanced Discriminant Local Direction Pattern Learning for Robust Palmprint Identification
1 Introduction
2 Related Work
2.1 Direction Features of Palmprint
2.2 Discriminant Regression Feature Learning
3 Proposed Methodology
3.1 Convolution Average Vector
3.2 Discriminant Elastic-Net Learning
4 Experiments
4.1 Palmprint Databases
4.2 Palmprint Identification
5 Conclusion
References
Latent Multi-view Subspace Clustering Based on Schatten-P Norm
1 Introduction
2 The Schatten P-Norm Based Multi-view Low Rank Representation Framework
2.1 Introduction to Schatten-P Norm
2.2 Objective Function
2.3 Model Optimization
3 Experiments
3.1 Datasets
3.2 Experimental Results
4 Conclusion
References
Security and Privacy
MOFIT: An Efficient Access Control Scheme with Attribute Merging and Outsourcing Capability for Fog-Enhanced IoT
1 Introduction
2 Background
3 System Model
4 Proposed Scheme: MOFIT
5 Security Proof
6 Performance Analysis
7 Conclusion
References
RepBFL: Reputation Based Blockchain-Enabled Federated Learning Framework for Data Sharing in Internet of Vehicles
1 Introduction
2 Related Work
3 System Model
4 Reputation Model
5 Security Analysis
6 Performance Evaluation
6.1 Experimental Setting
6.2 Numerical Results
7 Conclusion
References
Multimodal Fusion Representation Learning Based on Differential Privacy
1 Introduction
2 Preliminaries
2.1 Multimodal Representation Fusion
2.2 Differential Privacy
2.3 Binary Encoding Methods
3 Our Framework
3.1 Encoding Module
3.2 Perturbation Module
3.3 Aggregation Module
4 Experiments
4.1 Datasets
4.2 Experiment Details
5 Results and Discussion
6 Related Work
7 Conclusions
References
Efficient List Decoding Applied to ECC2
1 Introduction
2 Preliminaries
2.1 Elliptic Code and List Decoding
2.2 ECC2
3 The Module Basis Reduction Interpolation of List Decoding for Elliptic Code
3.1 Basis Construction
3.2 Module Basis Reduction
4 Performance
References
Federated Data Integration for Heterogeneous Partitions Based on Differential Privacy
1 Introduction
2 Related Work
3 Our Method
3.1 Multi-Party Feature-Label Matrix Negotiation Algorithm
3.2 Matrix Block Recognition Algorithm
3.3 DP-based Singular Value Decomposition Algorithm
3.4 Optimal Compression Dimension Model of SVD
3.5 Privacy Budget Allocation Strategy
4 Experiments
5 Conclusion
References
Patient-Chain: Patient-centered Healthcare System a Blockchain-based Technology in Dealing with Emergencies
1 Introduction
2 Related Work
2.1 Emergency Access Control for Personal Healthcare Records
2.2 Blockchain-Based Technology Application in Health Care System
3 Patient-Chain System Architecture
4 Evaluation
4.1 Environment Setting
4.2 The Four Scenarios
4.3 Future Work
5 Conclusion
References
A Differential Privacy Image Publishing Method Based on Wavelet Transform
1 Introduction
2 Differential Privacy Protection for Image Publishing
2.1 Laplace’s Method
2.2 Fourier Transform Method
3 The Method of WIP
3.1 Algorithm Design
3.2 Algorithm Error Analysis
4 Experiments and Analysis
4.1 Experimental Environment
4.2 Usability Analysis
4.3 Validity Analysis
5 Conclusion
References
Traffic Matrix Prediction Based on Differential Privacy and LSTM
1 Introduction
2 Related Work
3 Traffic Matrix Prediction Methods
3.1 Traffic Matrix Prediction
3.2 Long Short-Term Memory
3.3 Differential Privacy With Deep Learning
4 Experiments Results and Analysis
4.1 Performance Metrics
4.2 Results and Analysis
5 Conclusions
References
A Blockchain-Based Continuous Query Differential Privacy Algorithm
1 Introduction
2 Related Work
2.1 Differential Privacy Interactive Framework
2.2 Gaussian Mechanism
3 Basic Knowledge
3.1 Differential Privacy
4 Optimize AMG Variance Calculation Method
5 Continuous Query Differential Privacy Mechanism
5.1 Blockchain
5.2 Noise Reuse in Differential Privacy
5.3 Local Data Integrity Verification
6 Experimental Analysis
6.1 Experimental Environment
6.2 Comparison of Time Efficiency of Algorithm 1
6.3 Time Efficiency Comparison of CQDPM
7 Summary
References
Formalization and Verification of Group Communication CoAP Using CSP
1 Introduction
2 Background
2.1 Overall Architecture of Group Communication CoAP
2.2 Working Mechanism of Group Communication CoAP
2.3 CSP
3 Modeling Group Communication CoAP
3.1 Sets, Messages and Channels
3.2 Overall Modeling
3.3 Client Modeling
3.4 Intruder Modeling
4 Verification and Improvement
4.1 Verification
4.2 Verification Results
4.3 Attack and Improvements
5 Conclusion and Future Work
References
Author Index