Performance Engineering and Stochastic Modeling: 17th European Workshop, EPEW 2021, and 26th International Conference, ASMTA 2021, Virtual Event, ... (Programming and Software Engineering)

✍ Scribed by Paolo Ballarini (editor), Hind Castel (editor), Ioannis Dimitriou (editor), Mauro Iacono (editor), Tuan Phung-Duc (editor), Joris Walraevens (editor)

• Publisher: Springer
• Year: 2021
• Tongue: English
• Leaves: 499
• Category: Library

✦ Synopsis

This book constitutes the refereed proceedings of the 17th European Workshop on Computer Performance Engineering, EPEW 2021, and the 26^th International Conference, on Analytical and Stochastic Modelling Techniques and Applications, ASMTA 2021, held in December 2021. The conference was held virtually due to COVID 19 pandemic.

The 29 papers presented in this volume were carefully reviewed and selected from 39 submissions. The papers presented at the workshop reflect the diversity of modern performance evaluation, with topics ranging from modeling and analysis of network/control protocols and high performance/big data information systems, analysis of scheduling, blockchain technology, analytical modeling and simulation of computer and network systems.

✦ Table of Contents

Preface
Organization EPEW 2021
Organization ASMTA 2021
Contents
EPEW 2021
Workload Prediction in BTC Blockchain and Application to the Confirmation Time Estimation
1 Introduction
2 Related Work
3 Background and Motivation
3.1 The Problem of Predicting the Minimum Fee for QoS
3.2 Background on the ARIMA Model
3.3 Background on the Facebook Prophet Model
4 Evaluation of the Accuracy in Performance Predictions
4.1 Comparison of Time Series Prediction Models
4.2 Simulations
5 Conclusion
References
A Petri Net Formalism to Study Systems at Different Scales Exploiting Agent-Based and Stochastic Simulations
1 Introduction
2 Background
2.1 The ESSN Formalism
2.2 The ABM Formalism
3 Deriving an ABM from the ESSN Representation
4 Case Study: SEIRS
5 Conclusions and Future Work
References
State Space Minimization Preserving Embeddings for Continuous-Time Markov Chains
1 Introduction
2 Related Work
3 Preliminaries
4 Embeddings
5 Preservation of Strong Bisimulation
6 Preservation of Weak Bisimulation
7 Inverse of the Embeddings and Minimization
7.1 Inverse of the Embedding from ACTMCs to SCTMCs
7.2 Inverse of the Embedding from SCTMC to ACTMC
8 Conclusions
References
Multi-timescale Fairness for Heterogeneous Broadband Traffic in Access-Aggregation Networks
1 Introduction
2 Overview of STS-TVF Resource Sharing
3 Multi-timescale Fairness
3.1 Multi-timescale Throughput Value Functions (MTS-TVF)
3.2 MTS Rate Measurement-Based Marker
4 Fluid Simulation of MTS-TVF Resource Sharing
4.1 Fluid Model of Packet Marking and Forwarding
4.2 Fluid Simulator
5 Dimensioning Guidelines
5.1 The Proposed MTS-TVF
5.2 Intuitive Behaviour of the Proposed MTS-TVF
6 Approximate Analysis of the Stationary Behaviour
7 Simulation Results
7.1 Simulation Setup
7.2 Numerical Analysis of the Mean Stationary Behaviour
7.3 Time Series Examples
7.4 Statistical Results
8 Conclusion
References
DiPS: A Tool for Data-Informed Parameter Synthesis for Markov Chains from Multiple-Property Specifications
1 Introduction
1.1 Related Work
2 Methods
2.1 Model Checking
2.2 Data
2.3 Optimisation
2.4 Data Intervals
2.5 Sampling
2.6 Quantitative Sampling
2.7 Space Refinement
2.8 Metropolis-Hastings
2.9 Metropolis-Hastings-guided Refinement – HMH
3 Key Tool Features and Contributions
4 Implementation and Experiments
4.1 Implementation
4.2 Experiments
5 Conclusions and Future Work
References
Modelling a Fair-Exchange Protocol in the Presence of Misbehaviour Using PEPA
1 Introduction
2 Protocol Specification
2.1 An Anonymous and Failure Resilient Fair-Exchange E-Commerce Protocol
2.2 The Extended Protocol for Handling Misbehaviours and Communication Problems
3 PEPA Models of the Extended Protocol
4 Performance Evaluation of the Extended Protocol
5 Conclusion
References
Performance Evaluation of a Data Lake Architecture via Modeling Techniques
1 Introduction
2 Data Lakes
2.1 Data Lake, Data Warehouse and Data Hub
3 Data Lake Architectures
3.1 Smart Data Lake
3.2 Azure Data Lake Store
3.3 Google File Storage
3.4 Amazon S3
3.5 A Generic Data Lake Architecture
4 Case Study
4.1 Introduction
4.2 Considered Scenario
4.3 The Model
4.4 Simulation Results
5 Related Work
6 Conclusions and Future Work
References
Computing Bounds for Delay in a Stochastic Network
1 Introduction
2 Stochastic Bounds Based on Stochastic Monotonicity
3 Bounds Based on Associated Random Vectors
3.1 Links with <st Ordering and Independence
3.2 Distance and Association
3.3 Algorithm for a Lower Bound
3.4 Algorithms for an Upper Bound
4 Numerical Results
5 Concluding Remarks
References
Mixture Density Networks as a General Framework for Estimation and Prediction of Waiting Time Distributions in Queueing Systems
1 Introduction
2 Descriptions of Datasets and Feature Selections
3 Mixture Density Networks
4 Evaluation Metrics
5 Results
5.1 The M/M/1 Queueing System (Dataset 1)
5.2 The M/M/c Queueing Systems (Dataset 2)
5.3 The G/G/c Queueing Systems (Dataset 3 and Dataset 4)
5.4 The Preemptive Queueing System (Dataset 5)
5.5 Three Banks in Nigeria
6 Concluding Remarks
References
ASMTA 2021
Performance Evaluation and Energy Consumption for DVFS Processor
1 Introduction
2 DVFS Pstate Processor
3 A Multi-core Processor with One Pstate
3.1 Mean Number of Jobs and Response Time
3.2 Power and Energy Consumption
3.3 Numerical Comparison of the Six Pstates Configurations
4 A Multi-core Processor with Two Pstates
4.1 Closed Form for the Steady-State Distribution
4.2 Mean Number of Jobs and Response Time
4.3 Power and Energy Consumption
4.4 Optimization Under Energy and Response Time Constraints
4.5 Numerical Results
5 Conclusion
References
Performance Models of NFV-Based Hybrid Systems for Delay-Sensitive Services
1 Introduction
2 System Models
2.1 Markov Model with Random Reneging
2.2 Virtual Waiting Time Process
2.3 Markov Model with Deterministic Reneging
3 Performance Measures
4 Numerical Examples
4.1 Random Reneging Model
4.2 Deterministic Reneging Model
5 Concluding Remarks
References
A Stochastic SVIR Model with Imperfect Vaccine and External Source of Infection
1 Introduction
2 Model Description
3 Time Until M Individuals Get Infected
4 Results
5 Conclusions
References
Analysis of Single Bacterium Dynamics in a Stochastic Model of Toxin-Producing Bacteria
1 Introduction
2 The Mathematical Model
3 Dynamics of a Single Bacterium and Its Progeny
3.1 Lifespan of a Bacterium
3.2 Number of Toxin Molecules Produced by a Bacterium in Its Lifetime
3.3 Number of Division Events in the Lifespan of a Bacterium
3.4 Number of Bacteria in the Genealogy of a Bacterium
3.5 Number of Toxin Molecules Produced by the Genealogy of a Bacterium
4 Results
4.1 Parameter Calibration
4.2 Summary Statistics
5 Conclusions
References
EM Based Parameter Estimation for Markov Modulated Fluid Arrival Processes
1 Introduction
2 Problem Definition
2.1 Fluid Arrival Process
2.2 Measurement Data to Fit
3 The EM Algorithm
3.1 Log-Likelihood as a Function of the Hidden Variables
3.2 The Maximization Step of the EM Method
3.3 The Expectation Step of the EM Method
4 Implementation Details
4.1 Structural Restrictions of MMFAP Models
4.2 Computation of the Double Sided Inverse Laplace Transform
4.3 Reducing Computational Cost for Equidistant Measurement Intervals
4.4 Computation of E(i(k)) and E(Mi,j(k))
4.5 Computation of k and k
5 Numerical Examples
5.1 MMFAP Simulator
5.2 Approximating Q with the EM Method
6 Conclusions
A Maximizing the Model Parameters
B Expected Values of the Hidden Parameters
C Numerical Computation of the Expected Value of the Hidden Parameters
References
Reinforcement Learning with Model-Based Approaches for Dynamic Resource Allocation in a Tandem Queue
1 Introduction
2 Related Works
2.1 Energy and Performance Management in the Cloud
2.2 Control Management for Queueing Models and Markov Decision Process
2.3 Three-Tier Software Architecture
2.4 Reinforcement Learning for Resource Allocation
3 Tandem Queue Model
3.1 Model Description
3.2 Semi Markov Decision Process Description
4 Model-Based Reinforcement Learning
4.1 Reinforcement Learning
4.2 Advanced Model-Based Algorithms
5 Robustness of Model-Based RL
5.1 Markov Modulated Poisson Process
5.2 Partially Observable System
6 Experimental Results
6.1 Comparison Criteria Between Algorithms
6.2 Environments and Simulation Parameters
6.3 Results
7 Conclusion
References
Performance Analysis of Production Lines Through Statistical Model Checking
1 Introduction
2 Production Line Systems
3 HASL Model Checking
4 HASL Based Performance Analysis of Production Lines
4.1 Stochastic Petri Net Encoding of Production Lines
4.2 HASL Performance Indicators for Production Lines
5 Experiments
6 Discussion
7 Conclusion
References
Reliability Reference Model for Topology Configuration
1 Introduction
2 Reference Failure Model
2.1 Notation and Input Parameters
2.2 Probability State Space
2.3 Collapsing Failure Modes
2.4 Collapsing Failure Levels
3 Reference Outage Model
3.1 Notation and Formulation
3.2 Balls in Urns Formulation
3.3 Unequal Combinations
3.4 Outage Rate
4 Example Application
4.1 Requirements and Assumptions
4.2 Capacity and Latency Requirements
4.3 Reference Outage Model and Availability Requirement
4.4 Reportable Outage Requirement
5 Topology Configuration Tool
6 Conclusions
References
Splittable Routing Games in Ring Topology with Losses
1 Introduction
2 The Model
3 The Globally Optimal Solution and the Equilibrium
3.1 Minimizing Average Loss Probability
3.2 Equilibrium
3.3 Best Response
3.4 When Is the Globally Optimal Policy =1 an Equilibrium
4 Numerical Examples
4.1 The Equilibrium
4.2 The Best Response
4.3 Non-equilibrium of the Globally Optimal Policy =1
4.4 The Derivative of the Cost of a Player
5 Conclusions
References
Routing of Strategic Passengers in a Transportation Station
1 Introduction
1.1 Overview and Scope
1.2 Literature Review
1.3 Organization of the Paper
2 The Model
2.1 Mathematical Description and Notation
2.2 Information Structure - Strategies
2.3 Customer's Utility Function
3 Equilibrium Strategies
3.1 Characterization of an Equilibrium Strategy
3.2 Unused Facilities
3.3 Derivation of the Equilibrium Strategy
4 The Social Optimization Problem
5 Discussion - Extensions
References
Coupled Queueing and Charging Game Model with Energy Capacity Optimization
1 Introduction
2 Model Description
2.1 Mathematical Description
2.2 Expected Number of EV Charging
2.3 Stationary Probability of Fully Charged Battery
3 Queueing Game with Two Stations
3.1 Definition of the Nonatomic Game
3.2 Properties of Equilibrium
3.3 Price of Anarchy
4 Energy Capacity
5 Numerical Illustrations
5.1 Probability of Full Battery at Equilibrium
5.2 Price of Anarchy
6 Conclusions and Future Works
References
Hybrid Simulation of Energy Management in IoT Edge Computing Surveillance Systems
1 Introduction
2 Related Work
2.1 Analytical Methods
2.2 Simulation of Energy Management
2.3 Simulation and Analysis of Animal Movements
3 Characteristics of the System
3.1 Activities
3.2 Organization
3.3 Energy Management
4 Modeling Approach
5 Modeling a Scenario
6 Results
7 Conclusions
References
Highly Accurate, Explicit Approximation for Erlang B
1 Introduction
2 M/M/N/Q Queueing System Refresher
2.1 Formulation
2.2 Key Performance Indicators
3 Prior Work
4 Erlang B Approximation
4.1 Properties
4.2 B(N,N)
4.3 limN   B( N,2N )
4.4 Shape and Fit
5 Results and Application
6 Conclusions and Future Work
Appendix
References
The Join the Shortest Orbit Queue System with a Finite Priority Line
1 Introduction
2 Model Description
2.1 Equilibrium Equations and Stability Condition
2.2 Preliminary Results and Decay Rate
3 The Compensation Method
3.1 Construction of the Basis of Product Forms
3.2 The Formal Solution
3.3 Absolute Convergence and the Normalization Constant
4 A Simple Numerical Example
A Proof of Lemma 1
B Proof of Lemma 2
C Proof of Lemma 3
References
A Short Note on the System-Length Distribution in a Finite-Buffer GIX/C-MSP/1/N Queue Using Roots
1 Introduction
2 Description of the Model GIX/C-MSP/1/N
3 Analysis of the System-Length Distributions
3.1 Steady-State Distribution at a Pre-arrival Epoch
3.2 Stationary Distribution at Arbitrary Epoch
3.3 Performance Measures
4 Numerical Results
5 Conclusion
References
Stationary Analysis of Infinite Server Queue with Batch Service
1 Introduction
2 Stationary Distribution of M/M(b)/ Queue
3 Other Performance Measures for M/M(b)/ Queue
4 Numerical Examples
5 Conclusion
References
Performance Evaluation of Stochastic Bipartite Matching Models
1 Introduction
2 Model and Preliminary Results
2.1 Model and Notation
2.2 Discrete-Time Markov Chain
2.3 Stability Conditions and Stationary Distribution
3 Performance Evaluation by State Aggregation
3.1 Partition of the State Space
3.2 Normalization Constant
3.3 Waiting Probability
3.4 Mean Number of Unmatched Items and Mean Waiting Time
3.5 Time Complexity and Related Work
4 Numerical Evaluation
References
Analysis of Tandem Retrial Queue with Common Orbit and Poisson Arrival Process
1 Introduction
2 Mathematical Model
3 The System of Differential Kolmogorov Equations
4 The First Order Asymptotic: Fluid Limit
5 The Second Order Asymptotic: Diffusion Limit
6 Stationary Distribution of the Diffusion Process and Queue Length Approximation
7 Approximations Accuracy and Their Application Area
8 Conclusion
References
Queueing Analysis of a Mixed Model of Public and Demand Responsive Transportations
1 Introduction
2 The CRS System
2.1 Queueing Model
2.2 The Difficulty of this Model
3 Approximation Model
3.1 Vehicle Model
3.2 Road Model
4 The Required Time Distribution at the Spot
5 Numerical Experiments
6 Conclusion
References
An Analytical Framework for Video Quality and Excess Data Distribution in Multiple-Quality Video Under Dynamic Channel Conditions
1 Introduction
2 System Model
2.1 Markov Fluid Queues
2.2 Multi-regime Markov Fluid Queue Model for the Playout Buffer
2.3 The Distribution of Excess Data
3 Numerical Results
4 Conclusion
References
Author Index