For junior- and senior-level simulation courses in engineering, business, or computer science. While most books on simulation focus on particular software tools, Discrete Event System Simulation examines the principles of modeling and analysis that translate to all such tools. This language-independent text explains the basic aspects of the technology, including the proper collection and analysis of data, the use of analytic techniques, verification and validation of models, and designing simulation experiments. It offers an up-to-date treatment of simulation of manufacturing and material handling systems, computer systems, and computer networks. Students and instructors will find a variety of resources at the associated website, www.bcnn.net/, including simulation source code for download, additional exercises and solutions, web links and errata.
I Introduction to Discrete-Event System Simulation 1 Chapter 1 Introduction to Simulation 3 1.1 When Simulation Is the Appropriate Tool 4 1.2 When Simulation Is Not Appropriate 4 1.3 Advantages and Disadvantages of Simulation 5 1.4 Areas of Application 7 1.5 Systems and System Environment 9 1.6 Components of a System 9 1.7 Discrete and Continuous Systems 11 1.8 Model of a System 12 1.9 Types of Models 13 1.10 Discrete-Event System Simulation 13 1.11 Steps in a Simulation Study 14 References 18 Exercises 19 Chapter 2 Simulation Examples 21 2.1 Simulation of Queueing Systems 22 2.2 Simulation of Inventory Systems 39 2.3 Other Examples of Simulation 46 2.4 Summary 57 References 57 Exercises 57 Chapter 3 General Principles 67 3.1 Concepts in Discrete-Event Simulation 68 3.1.1 The Event Scheduling/Time Advance Algorithm 71 3.1.2 World Views 74 3.1.3 Manual Simulation Using Event Scheduling 77 3.2 List Processing 86 3.2.1 Lists: Basic Properties and Operations 87 3.2.2 Using Arrays for List Processing 88 3.2.3 Using Dynamic Allocation and Linked Lists 90 3.2.4 Advanced Techniques 92 3.3 Summary 92 References 92 Exercises 93 Chapter 4 Simulation Software 95 4.1 History of Simulation Software 96 4.1.1 The Period of Search (1955-60) 97 4.1.2 The Advent (1961-65) 97 4.1.3 The Formative Period (1966-70) 97 4.1.4 The Expansion Period (1971-78) 98 4.1.5 Consolidation and Regeneration (1979-86) 98 4.1.6 Integrated Environments (1987-Present) 99 4.2 Selection of Simulation Software 99 4.3 An Example Simulation 102 4.4 Simulation in Java 104 4.5 Simulation in GPSS 112 4.6 Simulation in SSF 117 4.7 Simulation Software 120 4.7.1 Arena 122 4.7.2 AutoMod 123 4.7.3 Extend 124 4.7.4 Flexsim 124 4.7.5 Micro Saint 125 4.7.6 ProModel 125 4.7.7 QUEST 126 4.7.8 SIMUL8 127 4.7.9 WITNESS 128 4.8 Experimentation and Statistical-Analysis Tools 128 4.8.1 Common Features 128 4.8.2 Products 129 References 131 Exercises 132 II Mathematical and Statistical Models 147 Chapter 5 Statistical Models in Simulation 149 5.1 Review of Terminology and Concepts 150 5.2 Useful Statistical Models 156 5.3 Discrete Distributions 160 5.4 Continuous Distributions 166 5.5 Poisson Process 186 5.5.1 Properties of a Poisson Process 188 5.5.2 Nonstationary Poisson Process 189 5.6 Empirical Distributions 190 5.7 Summary 193 References 193 Exercises 193 Chapter 6 Queueing Models 201 6.1 Characteristics of Queueing Systems 202 6.1.1 The Calling Population 202 6.1.2 System Capacity 204 6.1.3 The Arrival Process 204 6.1.4 Queue Behavior and Queue Discipline 205 6.1.5 Service Times and the Service Mechanism 206 6.2 Queueing Notation 208 6.3 Long-Run Measures of Performance of Queueing Systems 208 6.3.1 Time-Average Number in System L 209 6.3.2 Average Time Spent in System Per Customer w 211 6.3.3 The Conservation Equation: L = lambdaw 212 6.3.4 Server Utilization 213 6.3.5 Costs in Queueing Problems 218 6.4 Steady-State Behavior of Infinite-Population Markovian Models 220 6.4.1 Single-Server Queues with Poisson Arrivals and Unlimited Capacity: M/G/1 221 6.4.2 Multiserver Queue: M/M/c/ / 227 6.4.3 Multiserver Queues with Poisson Arrivals and Limited Capacity: M/M/c/N/ 233 6.5 Steady-State Behavior of Finite-Population Models (M/M/c/K/K) 235 6.6 Networks of Queues 239 6.7 Summary 241 References 242 Exercises 243 III Random Numbers 249 Chapter 7 Random-Number Generation 251 7.1 Properties of Random Numbers 251 7.2 Generation of Pseudo-Random Numbers 252 7.3 Techniques for Generating Random Numbers 253 7.3.1 Linear Congruential Method 254 7.3.2 Combined Linear Congruential Generators 257 7.3.3 Random-Number Streams 259 7.4 Tests for Random Numbers 260 7.4.1 Frequency Tests 261 7.4.2 Tests for Autocorrelation 265 7.5 Summary 267 References 268 Exercises 269 Chapter 8 Random-Variate Generation 272 8.1 Inverse-Transform Technique 273 8.1.1 Exponential Distribution 273 8.1.2 Uniform Distribution 276 8.1.3 Weibull Distribution 277 8.1.4 Triangular Distribution 278 8.1.5 Empirical Continuous Distributions 279 8.1.6 Continuous Distributions without a Closed-Form Inverse 283 8.1.7 Discrete Distributions 284 8.2 Acceptance-Rejection Technique 289 8.2.1 Poisson Distribution 290 8.2.2 Nonstationary Poisson Process 293 8.2.3 Gamma Distribution 294 8.3 Special Properties 296 8.3.1 Direct Transformation for the Normal and Lognormal Distributions 296 8.3.2 Convolution Method 298 8.3.3 More Special Properties 299 8.4 Summary 299 References 299 Exercises 300 IV Analysis of Simulation Data 305 Chapter 9 Input Modeling 307 9.1 Data Collection 308 9.2 Identifying the Distribution with Data 310 9.2.1 Histograms 310 9.2.2 Selecting the Family of Distributions 313 9.2.3 Quantile-Quantile Plots 316 9.3 Parameter Estimation 319 9.3.1 Preliminary Statistics: Sample Mean and Sample Variance 319 9.3.2 Suggested Estimators 321 9.4 Goodness-of-Fit Tests 326 9.4.1 Chi-Square Test 327 9.4.2 Chi-Square Test with Equal Probabilities 329 9.4.3 Kolmogorov-Smirnov Goodness-of-Fit Test 331 9.4.4 p-Values and "Best Fits" 333 9.5 Fitting a Nonstationary Poisson Process 334 9.6 Selecting Input Models without Data 335 9.7 Multivariate and Time-Series Input Models 337 9.7.1 Covariance and Correlation 337 9.7.2 Multivariate Input Models 338 9.7.3 Time-Series Input Models 340 9.7.4 The Normal-to-Anything Transformation 342 9.8 Summary 344 References 345 Exercises 346 Chapter 10 Verification and Validation of Simulation Models 354 10.1 Model-Building, Verification, and Validation 355 10.2 Verification of Simulation Models 356 10.3 Calibration and Validation of Models 361 10.3.1 Face Validity 362 10.3.2 Validation of Model Assumptions 362 10.3.3 Validating Input-Output Transformations 363 10.3.4 Input-Output Validation: Using Historical Input Data 374 10.3.5 Input-Output Validation: Using a Turing Test 378 10.4 Summary 379 References 379 Exercises 381 Chapter 11 Output Analysis for a Single Model 383 11.1 Types of Simulations with Respect to Output Analysis 384 11.2 Stochastic Nature of Output Data 387 11.3 Measures of Performance and Their Estimation 390 11.3.1 Point Estimation 390 11.3.2 Confidence-Interval Estimation 392 11.4 Output Analysis for Terminating Simulations 393 11.4.1 Statistical Background 394 11.4.2 Confidence Intervals with Specified Precision 397 11.4.3 Quantiles 399 11.4.4 Estimating Probabilities and Quantiles from Summary Data 400 11.5 Output Analysis for Steady-State Simulations 402 11.5.1 Initialization Bias in Steady-State Simulations 403 11.5.2 Error Estimation for Steady-State Simulation 409 11.5.3 Replication Method for Steady-State Simulations 413 11.5.4 Sample Size in Steady-State Simulations 417 11.5.5 Batch Means for Interval Estimation in Steady-State Simulations 418 11.5.6 Quantiles 422 11.6 Summary 423 References 423 Exercises 424 Chapter 12 Comparison and Evaluation of Alternative System Designs 432 12.1 Comparison of Two System Designs 433 12.1.1 Independent Sampling with Equal Variances 436 12.1.2 Independent Sampling with Unequal Variances 438 12.1.3 Common Random Numbers (CRN) 438 12.1.4 Confidence Intervals with Specified Precision 446 12.2 Comparison of Several System Designs 448 12.2.1 Bonferroni Approach to Multiple Comparisons 449 12.2.2 Bonferroni Approach to Selecting the Best 454 12.2.3 Bonferroni Approach to Screening 457 12.3 Metamodeling 458 12.3.1 Simple Linear Regression 459 12.3.2 Testing for Significance of Regression 463 12.3.3 Multiple Linear Regression 466 12.3.4 Random-Number Assignment for Regression 466 12.4 Optimization via Simulation 467 12.4.1 What Does Optimization via Simulation Mean? 468 12.4.2 Why is Optimization via Simulation Difficult? 469 12.4.3 Using Robust Heuristics 470 12.4.4 An Illustration: Random Search 473 12.5 Summary 476 References 476 Exercises 477 V Applications 483 Chapter 13 Simulation of Manufacturing and Material-Handling Systems 485 13.1 Manufacturing and Material-Handling Simulations 486 13.1.1 Models of Manufacturing Systems 486 13.1.2 Models of Material-Handling 487 13.1.3 Some Common Material-Handling Equipment 488 13.2 Goals and Performance Measures 489 13.3 Issues in Manufacturing and Material-Handling Simulations 490 13.3.1 Modeling Downtimes and Failures 491 13.3.2 Trace-Driven Models 495 13.4 Case Studies of the Simulation of Manufacturing and Material-Handling Systems 496 13.5 Manufacturing Example: A Job-Shop Simulation 499 13.5.1 System Description and Model Assumptions 499 13.5.2 Presimulation Analysis 502 13.5.3 Simulation Model and Analysis of the Designed System 503 13.5.4 Analysis of Station Utilization 503 13.5.5 Analysis of Potential System Improvements 504 13.5.6 Concluding Words 506 13.6 Summary 506 References 506 Exercises 507 Chapter 14 Simulation of Computer Networks 550 15.1 Introduction 550 15.2 Traffic Modeling 552 15.3 Media Access Control 555 15.3.1 Token-Passing Protocols 556 15.3.2 Ethernet 559 15.4 Data Link Layer 561 15.5 TCP 562 15.6 Model Construction 569 15.6.1 Construction 569 15.6.2 Example 571 15.7 Summary 573 References 574 Exercises 574 Appendix 576 Index 591