Discrete Event System

Simulation PDF
Jerry Banks

Scan to Download
Discrete Event System Simulation
Mastering Discrete-Event Simulation for Uncertain
System Design
Written by Bookey
Check more about Discrete Event System Simulation
Summary
Listen Discrete Event System Simulation Audiobook

Scan to Download
About the book
This book serves as an essential guide to discrete-event
simulation, a vital tool in operations research and management
science for navigating system design amid uncertainty. It
thoroughly explores the critical aspects of data collection,
analytical techniques, model verification and validation, and
the thoughtful design of simulation experiments. Accessible to
readers with a foundational knowledge of differential and
integral calculus, probability, and elementary statistics, it
incorporates practical examples using C++ and the latest
software packages for simulation and output analysis. Key
topics include modeling and random-variate generation from
the lognormal distribution, as well as clear distinctions
between terminating and steady-state simulations, and within-
and across-replication statistics. Featuring current applications
in manufacturing and material handling systems, the book also
highlights how simulation techniques vary across different
hierarchical levels of computing systems. Ideal for those eager
to deepen their understanding of system simulation.

Scan to Download
About the author
Jerry Banks is a prominent figure in the field of operations
research and simulation, renowned for his substantial
contributions to the development and application of discrete
event system simulation methodologies. With a robust
academic background and extensive experience in both
industry and academia, Banks has played a pivotal role in
advancing the understanding of complex systems through
simulation techniques. He has authored and co-authored
several influential texts, including the widely used "Discrete
Event System Simulation," which serves as a cornerstone
resource for students and professionals alike. His work not
only emphasizes theoretical foundations but also practical
applications, making simulation accessible and relevant across
various disciplines. Through his teaching and research, Banks
has inspired countless practitioners to harness the power of
simulation for solving real-world problems.

Scan to Download
Summary Content List
Chapter 1 : 1. Introduction to Simulation

Chapter 2 : 2. Simulation Examples in a Spreadsheet

Chapter 3 : 3. General Principles

Chapter 4 : 4. Simulation Software

Chapter 5 : 5. Statistical Models in Simulation

Chapter 6 : 6. Queueing Models

Chapter 7 : 7. Random-Number Generation

Chapter 8 : 8. Random-Variate Generation

Chapter 9 : 9. Input Modeling

Chapter 10 : 10. Verification, Calibration, and Validation of

Simulation Models

Chapter 11 : 11. Estimation of Absolute Performance

Chapter 12 : 12. Estimation of Relative Performance

Chapter 13 : 13. Simulation of Manufacturing and

Material-Handling Systems

Scan to Download
 Chapter 1 Summary : 1. Introduction to
Simulation

Section Summary

Introduction to Simulation imitates real-world processes to analyze behaviors and characteristics using a model with
Simulation mathematical or logical assumptions.

When Simulation Is Simulation is fitting for studying complex interactions, observing changes, gaining system insights,
the Appropriate Tool testing designs, and training, among other uses.

When Simulation Is Ten rules indicate inappropriateness for simulation, such as if common sense, analytical solutions, or
Not Appropriate cheaper experimentation suffices.

Advantages and Advantages: Allows exploration without disruption, tests designs, validates hypotheses, and identifies
Disadvantages of bottlenecks. Disadvantages: Requires training, can yield difficult-to-interpret results, and may be costly
Simulation or time-consuming.

Areas of Application Simulation is widely used in sectors including manufacturing, healthcare, logistics, and military
operations. The Winter Simulation Conference (WSC) showcases advancements in these areas.

Examples of Recent Notable uses include Turkish Army's fuel optimization, PLATO's Olympic Games management, and
Applications of Schlumberger's seismic survey improvements.
Simulation

Systems and System Systems encompass interconnected entities with a common purpose influenced by external factors.
Environment Understanding system boundaries is vital for modeling.

Components of a Key components include entities, attributes, activities, states, and events that define the system's
System behavior and characteristics.

Discrete and Systems may be discrete (with specific state changes) or continuous (with fluid state changes), often
Continuous Systems showing characteristics of both.

Model of a System Models simplify complex systems while maintaining essential characteristics for analyzing
relationships and predicting outcomes.

Discrete-Event This focuses on using numerical analysis for discrete-event systems, moving away from traditional
System Simulation mathematical methods.

Scan to Download
Section Summary

Steps in a The 12-step process includes problem formulation and implementation, highlighting the need for user
Simulation Study involvement to ensure validity and trust in results.

Introduction to Simulation

Simulation is the imitation of a real-world process or system

over time, either manually or through computer-based
methods. It generates an artificial history to analyze system
behavior and characteristics. A simulation model involves
assumptions expressed mathematically or logically. After
validating a model, various inquiries about system
performance can be explored, including investigating
potential changes before real-world implementation. In
complex systems where mathematical solutions are
impractical, numerical simulation is employed, collecting
data to estimate performance measures.

When Simulation Is the Appropriate Tool

Simulation has gained popularity due to easier programming

languages and powerful computing capabilities. It is suitable
for:
1. Studying complex system interactions.

Scan to Download
2. Observing effects of organizational and environmental
changes.
3. Gaining insights for system improvements.
4. Experimenting with variable interactions.
5. Serving educational purposes.
6. Testing designs or policies before implementation.
7. Verifying analytical solutions.
8. Determining machine requirements.
9. Training without operational disruption.
10. Visualizing system operation via animation.
11. Addressing complex systems where internal interactions
are significant.

When Simulation Is Not Appropriate

Ten rules outline when simulation is not suitable:

1. When common sense can resolve the problem.
2. If analytical solutions exist.
3. If direct experimentation is cheaper.
4. When costs exceed potential savings.
5. Lacking necessary resources or time.
6. Unavailability of needed data.
7. Insufficient time or personnel to verify and validate
models.

Scan to Download
8. Managers' unrealistic expectations.
9. Complexity difficulties in modeling system behavior.

Advantages and Disadvantages of Simulation

Advantages:

1. Explore new procedures without disrupting real

operations.
2. Test hardware designs without committing resources.
3. Validate hypotheses on phenomena.
4. Alter time dynamics of investigations.
5. Analyze variable interactions.
6. Identify bottlenecks in processes.
7. Understand operational systems better.
8. Answer “what if” scenario questions.

Disadvantages:

1. Requires specialized training for model building.

2. Results can be difficult to interpret due to randomness.
3. Modeling can be time-consuming and expensive.
4. Suitable analytical solutions may exist for some problems.

Scan to Download
Areas of Application

Simulation finds extensive applications across various

sectors including manufacturing, business processing,
military operations, healthcare, and logistics. The Winter
Simulation Conference (WSC) is a premier event for sharing
advances in simulation applications and methodologies.

Examples of Recent Applications of Simulation

- Turkish Army optimized fuel supply using simulation.

- PLATO helped design and manage Olympic Games
operations.
- Schlumberger used simulation to improve seismic survey
accuracy.
- Comprehensive applications across sectors like healthcare
and logistics demonstrate the broad impact of simulation
technology.

Systems and System Environment

A system is defined as a group of interconnected entities

working towards a common purpose, affected by external

Scan to Download
environmental changes. Understanding system boundaries is
crucial for effective modeling.

Components of a System

Key terms include:

- Entity: An object of interest.
- Attribute: A property of an entity.
- Activity: A defined action over time.
- State: The collection of variables describing a system’s
status.
- Event: An occurrence that alters the system state.

Discrete and Continuous Systems

Systems can be categorized as discrete (state changes at

specific points) or continuous (state changes fluidly over
time). Many practical systems exhibit characteristics of both.

Model of a System

Models represent systems for studying relationships and

predicting operational outcomes. They simplify real systems
while retaining essential characteristics. Various models exist

Scan to Download
including mathematical, physical, static, dynamic,
deterministic, and stochastic.

Discrete-Event System Simulation

This text focuses on discrete-event system simulation,

emphasizing numerical analysis over traditional
mathematical problem-solving due to the inherent
complexity of most systems.

Steps in a Simulation Study

The simulation study consists of 12 steps ranging from

problem formulation to implementation, emphasizing the
need for a systematic approach to ensure valid, efficient
results. The involvement of model users throughout the
process enhances trust and application of the model findings.

Scan to Download
Example
Key Point:The importance of simulation in
understanding complex systems.
Example:Imagine you're tasked with improving the
efficiency of your company's logistics. Instead of
disrupting real operations with trial and error, you can
create a simulation model that mimics your current
processes. This permits you to experiment with different
variables—like delivery routes, order sizes, or staffing
levels—without incurring the costs and risks associated
with actual changes. Through this simulation, you could
identify bottlenecks and test potential improvements,
gaining insights that translate into real-world
enhancements, all while ensuring your operational
stability.

Scan to Download
Critical Thinking
Key Point:The importance of evaluating when
simulation is appropriate and when it is not.
Critical Interpretation:The author emphasizes the
criteria for proper simulation usage, advocating for
careful consideration of its applicability. However, this
perspective may overlook instances where intuition or
simple solutions could be more effective, leading to
potential over-reliance on simulation tools. It remains
imperative for readers to weigh such claims against
alternative methodologies like analytical modeling or
direct experimentation, which may yield more
straightforward insights in certain contexts. Research on
decision-making in simulation applications (e.g., Law &
Kelton's 'Simulation Modeling and Analysis') can
provide additional insights.

Scan to Download
 Chapter 2 Summary : 2. Simulation
Examples in a Spreadsheet

Section Summary

Introduction This chapter provides an overview of discrete-event system simulation using spreadsheet
examples, focusing on simple models rather than extensive development.

Key Elements of Simulation

Random Numbers: Essential for representing uncertainty; discusses generation

methods like RAND() and VBA functions.
Components: Inputs, activities, events, states, and outputs.
Examples Covered: Coin tossing, queueing simulations, inventory models, project
network analysis.

1. Basics of Spreadsheet Offers practical methods for generating random numbers and emphasizes the importance of
Simulation understanding random variables.

2. Spreadsheet Simulation
Models
Simulation Table: Custom designs to track system states and compute outputs.
Examples: Includes coin tosses, service times, and customer arrival times.

3. Key Simulation Examples

Coin Toss Simulation: A Monte Carlo simulation illustrating random outcomes.

Queueing Models: Highlights management of service times and arrival processes.
Inventory Simulations: Manages stock levels and replenishment dynamics.

4. Reliability and Event-Based

Models
Replacement Policies: Compares strategies for cost-effective machine component

Scan to Download
Section Summary

replacement.
Hitting a Target: Models random errors during a bombing mission using normal
distributions.

5. Estimating Lead-Time
Demand and Project Activity
Network Lead-Time Demand: Simulates customer demand estimation during inventory
reordering.
Project Activity: Models project completion times using random variables.

6. Conclusion Highlights the significance of input characteristics, random variable generation, and
response analysis; discusses the use of various statistical distributions and the need for
multiple trials.

References A bibliography for further reading and resources related to simulation modeling.

Exercises Offers problems and case studies for practical application of simulation concepts with
spreadsheets.

Summary of Chapter 2: Simulation Examples in a

Spreadsheet

Introduction

This chapter provides an overview of simulation using

spreadsheet examples to illustrate key concepts in
discrete-event system simulation. The focus is on simple
models that can be executed manually or with a spreadsheet,
rather than extensive model development.

Scan to Download
Key Elements of Simulation

-
Random Numbers
: Random numbers are essential to represent uncertainty. The
chapter discusses methods to generate random samples, using
functions such as RAND() and user-defined VBA functions.
-
Components of a Simulation Model
: These include inputs, activities, events, states, and outputs.
-
Examples Covered
: Various applications including coin tossing, queueing
simulations, inventory models, and project network analysis.

1. Basics of Spreadsheet Simulation

-
Generating Random Numbers
: The chapter outlines practical methods to generate random
numbers and sample distributions for simulations.
-
Randomness Simulation
: Emphasizes the importance of understanding random

Scan to Download
variables and generating them correctly.

2. Spreadsheet Simulation Models

-
Simulation Table
: Custom designs for each simulation indicate how to track
system states over time and compute outputs.
-
Examples
: Discusses individual examples such as simulating coin
tosses, service times for calls, and generating arrival times
for customers.

3. Key Simulation Examples

-
Coin Toss Simulation
: Illustrates a Monte Carlo simulation through random
outcomes of a coin toss.
-
Queueing Models
: Single-server and two-server queue examples highlight
managing service times and arrival processes.

Scan to Download
-
Inventory Simulations
: Explains how to manage stock levels and replenishment
under various conditions.

4. Reliability and Event-Based Models

-
Replacement Policies
: Compares different strategies for replacing machine
components based on cost-effectiveness.
-
Hitting a Target
: Uses normal distributions to model random errors in hitting
a target during a bombing mission.

5. Estimating Lead-Time Demand and Project

Activity Network

-
Lead-Time Demand
: Focuses on simulations to estimate customer demand during
inventory reordering.
-

Scan to Download
Project Activity
: Demonstrates how to model project completion times using
random variables to estimate overall durations.

6. Conclusion

The chapter highlights the importance of understanding input

characteristics, generating random variables, and analyzing
responses. Various statistical distributions are covered, and
the chapter introduces the necessity of conducting multiple
trials for accurate simulation results.

References
: A bibliography is provided for further reading and resources
for simulation modeling.

Exercises
: Includes a variety of problems and case studies for practical
application of simulation concepts using spreadsheets.

Scan to Download
 Chapter 3 Summary : 3. General
Principles
Section Key Points

Overview This chapter establishes a framework for discrete-event simulation, outlining key components such as
entities, attributes, activities, and events. It emphasizes concepts applicable to various simulation
packages.

1. Concepts in - Dynamic and Stochastic Systems: Systems changing over time with random elements.
Discrete-Event - Major Concepts:
Simulation • System: Collection of entities with specific goals.
• Model: Abstract representation of relationships.
• System State: Variables at any given time.
• Entity: Objects needing representation (e.g., customer, machine).
• Attributes: Properties of entities (e.g., job routing).
• Event: Occurrences that alter the system state.
• Event List: Chronological listing of future events.
• Activity/Delay: Processing time description.

1.1 Event This algorithm advances simulation time in chronological order using a future event list (FEL) with
Scheduling/Time operations for updating the system state and generating future events.
Advance Algorithm

1.2 World Views Three primary world views in simulation modeling:

- Event Scheduling: Events drive system changes.
- Process Interaction: Focus on entity life cycles and interactions.
- Activity Scanning: Checks conditions for activities at each time increment.

2. List Processing - Methods for managing lists in simulations.

- Basic Operations: Adding/removing records and maintaining order.
- Use of Arrays: Simplifications and limitations for list management.
- Dynamic Allocation and Linked Lists: Efficient management of records.
- Advanced Techniques: Sophisticated strategies for list processing optimization.

3. Summary Recaps essential building blocks of simulation: entities, attributes, events, activities, and world views,
establishing a foundation for understanding sequence and timing in simulations.

References Sources for further reading: Cormen et al., Schriber & Brunner, Sedgewick.

Exercises Exercises to reinforce concepts discussed, encouraging readers to develop models and apply
simulation logic manually and with software tools.

General Principles

Scan to Download
Overview

This chapter provides a foundational framework for

modeling complex systems using discrete-event simulation.
It outlines the key components of discrete-event simulation
models, including entities, attributes, activities, and events,
focusing on changes in system state at discrete points in time.
The chapter emphasizes underlying concepts applicable
across various simulation packages, independent of specific
terminology or constructs that may vary between packages.

1. Concepts in Discrete-Event Simulation

-
Dynamic and Stochastic Systems
: The chapter deals with systems that change over time and
exhibit random elements.
-
Major Concepts
:
-
Install Bookey App to Unlock Full Text and
System
: A collection of interactingAudio
entities meeting specific goals.
-

Scan to Download
Chapter 4 Summary : 4. Simulation
Software

Simulation Software

History of Simulation Software

- Simulation software is in its middle-age phase, developed

over various decades.
- Key historical periods are divided as follows:
1.
1955–1960: The Period of Search
- Simulation conducted using general-purpose programming
languages without dedicated routines.
2.
1961–1965: The Advent
- Intro of simulation programming languages (SPLs) like
GPSS and SIMSCRIPT.
3.
1966–1970: The Formative Period
- Major languages matured; GPSS underwent significant

Scan to Download
revisions.
4.
1971–1978: The Expansion Period
- Advances from external sources, GPSS/H became widely
used.
5.
1979–1986: The Period of Consolidation and
Regeneration
- Extension to desktop environments with packages like
SLAM and SIMAN.
6.
1987–2008: Integrated Environments
- Growth of user-friendly environments with graphical
interfaces and animation.
7.
2009–2011: The Future
- Potential for agent-based modeling and enhanced user
interfaces.

Features and Attributes of Simulation Software

- Considerations for purchasing simulation software include

cost, ease of learning, modeling power, and animation
capabilities.

Scan to Download
- Software types:
1. General-purpose programming languages (e.g., C, Java).
2. Simulation programming languages (e.g., GPSS/H).
3. Simulation environments with graphical interfaces and
built-in programming languages.
- Important components for analysis include output
analyzers, data management, scenario definitions, and run
management.

Selection of Simulation Software

- Evaluate software based on:

- Modeling accuracy and detail.
- Execution speed.
- Vendor support and documentation.
- Flexibility of input methods.
- Check for features such as animation capabilities, statistical
analysis tools, and output reporting functionalities.

Example Simulation: Checkout Counter

- A simulation model for a grocery checkout counter was

developed using Java, GPSS/H, and the Scalable Simulation
Framework (SSF).

Scan to Download
- The model runs until a specified number of customers are
served, analyzing interarrival and service times.

Simulation in Different Languages

-
Java
: Simulation requires explicit programming for
event-scheduling and statistics.
-
GPSS/H
: Provides block structures for processing with drawing
capabilities for visual representation.
-
SSF
: Allows for simpler object-oriented process-view
simulations, beneficial for parallel processing.

Simulation Environments Overview

- Examples include:
-
AnyLogic
: Multi-method support including discrete-event and

Scan to Download
agent-based modeling.
-
Arena
: Graphical model-building with object-based design.
-
Flexsim
: 3D animation and collaboration for complex systems.
-
SIMUL8
: Focused on service industries with user-friendly modeling
features.

Experimentation and Statistical-Analysis Tools

- Packages offer tools for statistical analysis, optimization

techniques, and the ability to define scenarios and export data
for further analysis.
This chapter provides insights into the evolution of
simulation software, its critical features, and considerations
for selection and implementation across various
programming languages and environments.

Scan to Download
Chapter 5 Summary : 5. Statistical
Models in Simulation

Statistical Models in Simulation

Introduction

In modeling real-world phenomena, the actions of entities

within systems often exhibit probabilistic behavior rather
than deterministic predictability due to various sources of
variation. This chapter focuses on the development and
application of statistical models that can effectively describe
such variability.

1. Review of Terminology and Concepts

-
1.1 Discrete Random Variables
: Defined as variables with finite or countably infinite
possible values. A probability mass function (pmf) describes
their distribution.

Scan to Download
-
1.2 Continuous Random Variables
: Defined by intervals with associated probability density
functions (pdfs), which provide probabilities through
integration.
-
1.3 Cumulative Distribution Function (cdf)
: Represents the probability that a random variable takes on a
value less than or equal to a specific point.
-
1.4 Expectation
: The mean or expected value of a random variable,
calculated differently for discrete and continuous cases.
-
1.5 Mode
: The value that occurs most frequently in a distribution,
applicable in both discrete and continuous contexts.

2. Useful Statistical Models

Statistical models aid simulation by introducing probabilistic

events across different domains:
-
2.1 Queueing Systems

Scan to Download
: Characterized by interarrival and service times that can be
modeled as probabilistic events.
-
2.2 Inventory Systems
: Incorporate demands and lead times as random variables,
commonly modeled by distributions like the gamma
distribution.
-
2.3 Reliability Models
: Use exponential, gamma, and Weibull distributions to
model time-to-failure and service times.
-
2.4 Limited Data Situations
: When comprehensive data is unavailable, uniform,
triangular, or beta distributions can help in modeling.
-
2.5 Other Distributions
: Bernoulli, binomial, hyper-exponential distributions might
apply depending on the context.

3. Discrete Distributions

Focuses on random variables that take integer values, with

distributions including the Bernoulli, binomial, geometric,

Scan to Download
negative binomial, and Poisson distributions.

4. Continuous Distributions

Describes random phenomena where values exist over

intervals. Key distributions addressed include uniform,
exponential, gamma, Erlang, normal, Weibull, triangular,
lognormal, and beta distributions.

5. Poisson Process

Details the characteristics of a Poisson process for counting

events over time, including properties such as stationary and
independent increments.

6. Empirical Distributions

Empirical distributions are constructed from actual observed

data, reflecting real-world scenarios without assuming a
specific theoretical distribution.

7. Summary

The chapter reinforces the importance of probabilistic models

Scan to Download
in simulations, emphasizing the variability of real-world
phenomena and the necessity of understanding distributional
forms to effectively analyze input data in simulation
contexts.

References

Citations of resources consulted throughout the chapter,

including textbooks and research papers that support the
concepts discussed.

Exercises

A series of exercises designed to apply the concepts learned

in the chapter to practical problems in simulation and
statistical modeling.

Scan to Download
Critical Thinking
Key Point:The Necessity of Probabilistic Models in
Simulation
Critical Interpretation:One significant point made in the
chapter is the assertion that probabilistic models are
essential for accurately simulating real-world systems
due to their inherent variability. This perspective
encourages a deeper exploration into the necessity of
embracing randomness in modeling, but it's important to
question whether the reliance on statistical distributions
may oversimplify complex systems. Critics may argue
that deterministic approaches could also yield valuable
insights, particularly in systems where unpredictability
is minimal. Scholars like Robert C. Schwebel in
'Modeling and Simulation of Systems Using MATLAB
and Simulink' emphasize the potential of combining
both deterministic and stochastic methods for a more
robust understanding of system behavior. Thus, while
Banks highlights the critical role of statistical models, it
is prudent for readers to remain aware that other
modeling paradigms may also be relevant and could
potentially enrich simulation practices.

Scan to Download
Chapter 6 Summary : 6. Queueing
Models

Queueing Models

Introduction to Queueing Models

Queueing Models are used to analyze systems where entities,

referred to as customers, arrive to request services. These
models are applicable across various fields, including service
facilities, production, and transport systems. The main
performance measures for queueing systems include server
utilization, waiting line lengths, and customer delays, with a
focus on tradeoffs between server efficiency and customer
satisfaction.

Characteristics of Queueing Systems

1.
Key Elements:
Customers (any entities needing service) and servers

Scan to Download
(resources providing service) define the system. Various
examples depict different types of queueing systems with
defined customer-server relationships.

2.
Calling Population:
The potential customers can be finite or infinite. The arrival
rate is affected by the system's current state, particularly in
finite population scenarios.
3.
System Capacity:
Many systems limit the number of customers that can wait
or be served, impacting arrival rates.
4.
Arrival Process:
Characterization of how customers arrive, often modeled as
Poisson processes with interarrival times following an
exponential distribution.
5.
Queue Behavior and Discipline:
Refers to customer actions while waiting and the order of
Install
service Bookey
(FIFO, App to Unlock Full Text and
LIFO, etc.).
6. Audio
Service Times:

Scan to Download
Chapter 7 Summary : 7.
Random-Number Generation

Random-Number Generation

Random numbers are essential for simulating discrete

systems and are generated by various programming
languages. This chapter discusses random number generation
and its testing for statistical properties.

1 Properties of Random Numbers

For random numbers \( R_1, R_2, \ldots \), uniformity and

independence are crucial. Each \( R_i \) should come from a
continuous uniform distribution on the interval [0, 1]. The
probability density function is defined, and its expected value
and variance are calculated. Key consequences include the
expected number of observations in divided intervals and
independence of generated values from previous ones.

2 Generation of Pseudo-Random Numbers

Scan to Download
"Pseudo" indicates that these numbers aren't truly random
due to their generation by defined methods. Common issues
arise, such as non-uniformity or dependence within generated
sequences. Reliable generation methods should be fast,
portable, and have long cycles while approximating ideal
statistical properties.

3 Techniques for Generating Random Numbers

The linear congruential method is a popular technique,

producing a sequence of integers based on a recursive
relationship, which can be transformed into random numbers
in [0, 1]. Properties such as maximum density and maximum
period depend on the parameters chosen. Example
calculations illustrate its application.

3.2 Combined Linear Congruential Generators

As simulation complexity increases, generators with longer

periods are needed. Combining several multiplicative
congruential generators can yield sequences with desirable
statistical properties and extended periods.

3.3 Random-Number Streams

Scan to Download
The initial seed of the generator establishes a stream of
random numbers. Using strategically spaced streams
enhances system comparisons in simulations.

4 Tests for Random Numbers

Testing for uniformity and independence are essential. Two

primary tests include:
-
Frequency Test
: Compares generated numbers to uniform distribution using
the Kolmogorov–Smirnov or chi-square test.
-
Autocorrelation Test
: Evaluates correlation among generated numbers to indicate
independence.

5 Summary

This chapter covers random number generation and its

statistical validation. While many simulation tools generate
random numbers automatically, understanding their
properties is important to ensure reliability. Ongoing research

Scan to Download
continues to improve random number generation techniques,
emphasizing the complexity of achieving truly random
sequences.

Scan to Download
Chapter 8 Summary : 8.
Random-Variate Generation

Random-Variate Generation

Overview

This chapter focuses on the methods for sampling from

various continuous and discrete distributions, emphasizing
the generation of random variates for simulation models. It
highlights how many real-world variables, such as
interarrival times and service times, can be unpredictably
modeled using random variables with specified probability
distributions.

Key Techniques

1.
Inverse-Transform Technique

- Can be applied to exponential, uniform, Weibull, and

Scan to Download
triangular distributions, as well as empirical distributions.
- To generate random variates, the cumulative distribution
function (cdf) is inverted.
- Example for Exponential Distribution:
- The random variate generator is derived as \(X =
-\frac{1}{\lambda} \ln(R)\), where \(R\) is a uniformly
distributed random number on [0,1].
2.
Uniform Distribution

- Random variables can be generated as \(X = a + (b - a)R\)

for uniformly distributed variables.
3.
Weibull Distribution

- Generating values involves using the cdf and similarly

deriving \(X\) as \(X = \alpha[-\ln(1 - R)]^{1/\beta}\).
4.
Triangular Distribution

- The generation method adjusts depending on value

intervals relative to the distribution parameters.
5.
Empirical Continuous Distributions

Scan to Download
 - If no theoretical distribution fits the data, an empirical
distribution utilizing observed data can be used.
- Random variates are generated through interpolation
between observed data points.
6.
Discrete Distributions

- Employed similarly using the inverse-transform technique

or through empirical distributions.

Acceptance-Rejection Technique

- Utilized for distributions like the Poisson and Gamma when

efficient generation is required.
- It involves determining if a generated random number
meets defined criteria, allowing or rejecting the value
accordingly.

Special Properties

- Specific properties or transformations can simplify the

generation process for distributions, such as the Box-Muller
transform for normal distributions.

Scan to Download
- Convolution methods for Erlang and binomial distributions
leverage relationships between randomly generated variables
to create new variates efficiently.

Summary

The chapter encompasses various techniques essential for

random-variate generation crucial in the simulation modeling
process, providing a foundation for analysts to apply these
methods in real-world scenarios. Techniques range from the
inverse-transform and acceptance-rejection methods to the
use of special properties for specific distribution types.
Understanding these methods allows model builders to
generate the necessary random inputs for simulations
effectively.

Scan to Download
Chapter 9 Summary : 9. Input Modeling

Input Modeling

Overview

Input models play a crucial role in simulation by providing

essential data that drives simulations across various fields
such as queueing systems, reliability problems, and supply
chains. The reliable selection and fitting of input models are
vital for obtaining actionable simulation results.

Steps for Developing Input Models

1.
Data Collection
: Gathering accurate data can be resource-intensive;
situations may arise where data cannot be collected,
necessitating the use of expert opinions or estimates.
2.
Identifying Probability Distributions
: Analyze the collected data (often through histograms) to

Scan to Download
select appropriate probability distributions.
3.
Parameter Estimation
: Determine specific parameters for the chosen distribution,
usually based on sample data.
4.
Goodness-of-Fit Evaluation
: Validate the fit of the chosen distribution using statistical
tests.

Challenges in Data Collection

-
Stale Data
: Old datasets may not reflect current realities, leading to
misleading conclusions.
-
Unexpected Data Variability
: Data might reveal distinct distributions or dependencies not
initially recognized.
-
Install Bookey
Time-Varying Data App to Unlock Full Text and
Audio
: Fluctuations in data collection (e.g., daily call volumes)
may necessitate non-stationary models.

Scan to Download
Chapter 10 Summary : 10. Verification,
Calibration, and Validation of
Simulation Models

Verification, Calibration, and Validation of

Simulation Models

Introduction

The chapter addresses the crucial tasks of verifying and

validating simulation models, emphasizing the skepticism of
managers and engineers regarding model outputs. The goal
of validation is twofold: to ensure the model accurately
reflects system behavior and to build user confidence in the
model's credibility.

Verification vs. Validation

1.
Verification
: Focuses on ensuring that the model is implemented

Scan to Download
correctly according to the conceptual design. It involves
checking that input parameters and logical structures are
accurately represented in the simulation.

2.
Validation
: Confirms that the model faithfully represents the real
system. This is done through comparing model outputs with
actual system behavior, often requiring calibration to address
discrepancies.

Model Development Process

Model building is iterative and includes:

- Observing the real system.
- Constructing a conceptual model based on assumptions.
- Implementing an operational model using simulation
software.
- Continual verification and validation throughout
development.

Verification Techniques

Several practical techniques aid in verifying the simulation

Scan to Download
model:
- Independent checks from simulation experts.
- Flow diagrams of logical structures.
- Analyzing model outputs for reasonableness.
- Documenting variables and logic within the operational
model.
- Using graphical interfaces for clarity.

Calibration and Validation

While calibration and validation are distinct, they often occur

simultaneously. This process entails:
- Comparing the model with a real system.
- Adjustments made iteratively until model accuracy is
acceptable.
- Evaluating face validity, assumptions, and input-output
transformations.

Techniques for Validation

-
Face Validity
: Engaging knowledgeable users in ensuring model realism.
-

Scan to Download
Assumptions Validation
: Testing structural and data assumptions under various
conditions.
-
Input-Output Transformation Validation
: Using existing data to predict system behavior with the
model.

Statistical Testing

Objective statistical methods (e.g., t-tests) are recommended

to validate model predictions against real system data. A
well-structured test can help quantify the model's predictive
ability and guide necessary adjustments.

Turing Test for Validation

Involving knowledgeable personnel in comparing model

outputs with real system outputs can identify inadequacies
and enhance model credibility.

Summary

Validation of simulation models is essential, as decisions

Scan to Download
based on these models can greatly impact outcomes. By
employing a variety of techniques, including statistical
testing and expert reviews, model builders can assure model
accuracy and foster trust among users and decision-makers.

References

A list of foundational texts and papers relevant to verification

and validation techniques is provided for further reading and
in-depth understanding.

Exercises

The chapter concludes with exercises designed to reinforce

concepts related to statistical validation and model
verification, encouraging practical application of learned
skills.

Scan to Download
Critical Thinking
Key Point:The Importance of Verification and
Validation in Simulation Modeling
Critical Interpretation:A key point made in this chapter
is the significance of verifying and validating simulation
models to ensure their accuracy and reliability, which is
fundamental in building trust among users. However,
one must critically consider that reliance on these
techniques does not eliminate the inherent uncertainties
associated with modeling real-world systems. Different
authors, such as Sargent (2013) and Law & Kelton
(2000), suggest that while validation methods can
enhance credibility, they cannot guarantee absolute
correctness of model behavior, especially under varying
conditions. Therefore, readers should remain cautious
about the author's perspective, acknowledging that
models are simplifications and may not capture every
nuance of the systems they aim to simulate.

Scan to Download
Chapter 11 Summary : 11. Estimation of
Absolute Performance

Estimation of Absolute Performance

Overview

Output analysis in simulation aims to predict the

performance of a system or compare different designs.
Estimating absolute performance involves measuring one or
more system performance metrics, which are influenced by
the random variability of the simulation output when using
random-number generators. Statistical methods estimate the
standard error or confidence intervals for performance
measures derived from simulation data.

Key Concepts

-
Random Variables
: Output variables, like total cost per week in an inventory

Scan to Download
system, are treated as random variables with unknown
distributions.
-
Autocorrelation
: Due to dependencies between observations (e.g.,
consecutive weeks in an inventory system), standard
statistical methods assuming independence need adjustments.
-
Initialization Bias
: Initial conditions can heavily influence results; proper
selection is crucial for reliable long-term performance
estimation.

Types of Simulations

1.
Terminating Simulations
: Run for a designated period until a stopping condition is
met. Examples include operational studies like banks or
manufacturing systems.
2.
Steady-State Simulations
: Focus on the long-term behavior of a system, often with
continuous outputs and designs that avoid initialization bias.

Scan to Download
Statistical Techniques

-
Point Estimation
: Calculating mean estimates from discrete observations or
continuous processes.
-
Confidence Intervals
: Providing ranges around point estimates to quantify
reliability.
-
Batch Means Method
: Separating data into batches to mitigate autocorrelation and
refine estimates.

Practical Applications

1.
Simulated Scenarios
: Various systems (e.g., call centers, production lines) are
analyzed to understand performance variabilities and to
deduce mean behavior over time.
2.

Scan to Download
Sample Size Determination
: Estimating the required number of replications and
observations needed to achieve a specified error tolerance in
estimates.

Conclusion

The chapter emphasizes the need for rigorous statistical

analysis in simulation output to derive reliable performance
estimates. Different simulation types necessitate specific
analytical strategies, especially in handling initialization and
correlation issues. Techniques such as the batch means
method and careful design of experiments are crucial for
achieving precise simulations.

Scan to Download
Critical Thinking
Key Point:The dependence on statistical methods can
lead to misleading conclusions.
Critical Interpretation:While the chapter highlights the
importance of rigorous statistical analysis in estimating
performance metrics from simulations, it's essential to
recognize that reliance on these methods may mask
underlying assumptions and limitations. For instance,
the adjustment for autocorrelation can sometimes
oversimplify complex dependencies within data.
Moreover, the initialization bias issue suggests that
early states can unduly influence long-term estimates,
which challenges the notion of steady-state conditions
being truly representative. Critics argue that statistical
frameworks can obscure the model’s fidelity and
question the foundational premises of randomness
assumed by simulation techniques, as noted in sources
like 'Simulation Modeling and Analysis' by Law and
Kelton. This calls for a more critical approach in
interpreting simulation outputs, reminding readers that
not all results derived from statistical models are to be
taken at face value.

Scan to Download
Chapter 12 Summary : 12. Estimation of
Relative Performance

Estimation of Relative Performance

Overview

This chapter focuses on the statistical methods used to

compare the relative performance of different system designs
through simulation. Since response variables exhibit random
variability, statistical analysis helps distinguish real
differences in performance from those caused by inherent
randomness.

1. Comparison of Two System Designs

The chapter discusses methods for comparing two system

designs using replication. It introduces independent sampling
and the common random numbers (CRN) technique, which
reduces variance in the estimated performance measure
difference.

Scan to Download
*
Example: Software Made Personal (SMP) Call
Center

SMP compares customer service times between two designs

of its call center—one with specialized operators and one
with cross-trained operators. The analysis utilizes
simulations to estimate mean response times and compute
confidence intervals for performance measure differences.
*
Confidence Interval Analysis

Three scenarios are defined based on the confidence interval

results, indicating if one system is statistically significantly
better than the other or if no difference can be conclusively
determined.

1.1 Independent Sampling

Independent sampling involves using separate random

number streams for each design, which assumes the
Install Bookey
observations from eachApp to are
system Unlock Full independent.
statistically Text and
Audio intervals and standard
Procedures to compute confidence
errors are detailed.

Scan to Download
Chapter 13 Summary : 13. Simulation of
Manufacturing and Material-Handling
Systems

Simulation of Manufacturing and

Material-Handling Systems

Overview

Simulation plays a crucial role in the design and analysis of

manufacturing and material-handling systems, aiding
engineers in the assessment of new facilities, equipment
investments, and improvements to existing systems. It allows
for comprehensive evaluations without disrupting current
operations.

1. Manufacturing and Material-Handling

Simulations

Scope and level of detail are key considerations in simulation

projects. Scope defines what components are included in the

Scan to Download
model, while level of detail explores how thoroughly each
component is analyzed. Proper selection hinges on the
objectives of the study and available data.
*
1.1 Models of Manufacturing Systems

Key characteristics considered include physical layout,

labor schedules, equipment rates, and production schedules.
*
1.2 Models of Material Handling Systems

Effective management of material handling is vital as it

typically consumes a large amount of system time. This
involves evaluating subsystems like conveyors and
transporters.
*
1.3 Common Material-Handling Equipment

Includes various transporters (e.g., forklifts, AGVs) and

conveyor systems, essential for efficient manufacturing and
distribution.

2. Goals and Performance Measures

Scan to Download
The primary goal of simulation is to provide insights rather
than just quantitative data. Common performance measures
focus on throughput, cycle time, resource utilization,
queueing, and system bottlenecks.

3. Issues in Manufacturing and Material-Handling

Simulations

Key issues include the accurate modeling of downtimes and

choosing between actual input data or statistical models.
Proper treatment of these factors significantly influences
simulation outcomes.
*
3.1 Modeling Downtimes and Failures

Strategies for modeling both scheduled and unscheduled

downtimes are discussed, highlighting the impact on overall
system performance.
*
3.2 Trace-Driven Models

Utilizing actual historical data leads to more accurate

models as opposed to purely statistical distributions.

Scan to Download
4. Case Studies of the Simulation of Manufacturing
and Material Handling Systems

A range of published studies illustrates the diverse

applications and benefits of simulation across various sectors
including aerospace manufacturing and semiconductor
processing.

5. Manufacturing Example: An Assembly-Line

Simulation

A detailed ongoing simulation of a production line is

presented, focusing on throughput, bottlenecks, operator
efficiency, and WIP storage effects on overall system
performance. This example emphasizes the necessity of
iterative simulations to optimize design.

6. Summary

This chapter underscores the significance of simulation in

manufacturing and material handling, outlining key
principles, approaches, and challenges involved. The
accurate representation of downtimes and leveraging
historical data through trace-driven models are emphasized
as crucial for achieving reliable simulation outcomes.

Scan to Download
Best Quotes from Discrete Event System
Simulation by Jerry Banks with Page
Numbers
View on Bookey Website and Generate Beautiful Quote Images

Chapter 1 | Quotes From Pages 6-27

1.A simulation is the imitation of the operation of a
real-world process or system over time.
2.Once developed and validated, a model can be used to
investigate a wide variety of 'what if' questions about the
real-world system.
3.Simulation can also be used to study systems in the design
stage, before such systems are built.
4.Simulation enables the study of, and experimentation with,
the internal interactions of a complex system or of a
subsystem within a complex system.
5.Changing simulation inputs and observing the resulting
outputs can produce valuable insights about which
variables are the most important and how variables interact.
6.Simulation has many advantages, but some disadvantages.

Scan to Download
7.The art of modeling is enhanced by an ability to abstract
the essential features of a problem, to select and modify
basic assumptions that characterize the system.
8.A simulation model of a system is a simplification of the
system.
Chapter 2 | Quotes From Pages 28-93
1....our main purpose is to introduce and illustrate
some of the key concepts in simulation; it is not to
teach how to formulate or develop a spreadsheet
model, but rather to use a few simple spreadsheet
models to illustrate, by example, some of the key
elements of any simulation.
2....although some types of real-world simulation models,
such as those for risk analysis, financial analysis, some
reliability problems and others, are amenable to a
spreadsheet formulation and solution, the spreadsheet has
severe limitations for most complex real-world dynamic,
event-based simulations.
3.By specifying the seed, the user can make a random

Scan to Download
 number generator repeat a sequence of numbers. This is
valuable for debugging simulation models and repeating
simulation experiments.
4.The greatest simplicity possible, as long as the model logic
is correct, assists in an ease of understanding that is of
tremendous value, especially when trying to verify that a
model’s logic is indeed correct.
5....the average waiting time of those who wait is 3.54
minutes, computed as follows: Average waiting time of
those who wait (minutes) = total time customers wait in
queue (minutes) / total number of customers who wait =
163 / 46 = 3.54 minutes.
6.The model outputs could be used to compute measures of
system performance (also known as model responses, or
simply responses).
Chapter 3 | Quotes From Pages 94-123
1.In discrete-event simulation, a system is modeled
in terms of its state at each point in time; of the
entities that pass through the system and the

Scan to Download
 entities that represent system resources; and of the
activities and events that cause the system state to
change.
2.Discrete-event models are appropriate for those systems for
which changes in system state occur only at discrete points
in time.
3.A discrete-event simulation proceeds by producing a
sequence of system snapshots (or system images) that
represent the evolution of the system through time.
4.The future event list (FEL) contains all previously
scheduled future events and their associated event times...
Time t is the value of CLOCK, the current value of
simulated time.
5.To keep track of activities and their expected completion
time, at the simulated instant that an activity duration
begins, an event notice is created having an event time
equal to the activity’s completion time.
6.An activity typically represents a service time, an
interarrival time, or any other processing time whose

Scan to Download
 duration has been characterized and defined by the
modeler.
7.In summary, the process-interaction approach has been
adopted by the simulation packages most popular in the
USA.
8.The completion of an activity is an event, often called a
primary event, that is managed by placing an event notice
on the FEL.

Scan to Download
Chapter 4 | Quotes From Pages 124-175
1.If you were about to purchase simulation software,
what would concern you? Would it be the cost, the
ease of learning, the ease of use, or would it be the
power to model the kind of system with which you
are concerned?
2.Many simulation environments contain a simulation
programming language, but some take a graphical approach
similar to process-flow diagramming.
3.Today very few people writing discrete-event simulation
models are using programming languages alone; however,
in certain application areas, some people are using
packages based on Java or on some other general-purpose
language.
4.The period of Search (1955–1960) marked significant
exploration in the development of reusable routines to
facilitate simulation.
5.In summary, agent-based modeling is a method for
simulating the actions and interactions of autonomous

Scan to Download
 individuals (the agents) in a network, with a view to
assessing their effects on the system as a whole.
Chapter 5 | Quotes From Pages 176-231
1.To the model builder, these variations appear to
occur by chance and cannot be predicted.
2.An appropriate model can be developed by sampling the
phenomenon of interest.
3.The exponential distribution is often used to model service
times that are highly variable.
4.A major task in simulation is the collection and analysis of
input data.
5.Through continued efforts in the selection of an appropriate
distribution form, a postulated model could be accepted.
6.The cumulative distribution function (cdf) measures the
probability that the random variable X assumes a value less
than or equal to x.
7.The expectation of a random variable is also referred to as
the mean, ¼, or the first moment of X.
8.The normal distribution is used so often that the notation X

Scan to Download
 "< N(¼, Ã^2) has been adopted.
Chapter 6 | Quotes From Pages 232-279
1.Queueing models, whether solved mathematically
or analyzed through simulation, provide the
analyst with a powerful tool for designing and
evaluating the performance of queueing systems.
2.From a certain degree, some of the input parameters are
under management’s direct control.
3.For relatively simple systems, these performance measures
can be computed mathematically—at great savings in time
and expense as compared with the use of a simulation
model.
4.The primary long-run measures of performance of
queueing systems are the long-run time-average number of
customers in the system (L) and in the queue (LQ), the
long-run average time spent in the system (w) and in the
queue (wQ) per customer, and the server utilization.
5.An important application of mathematical queueing models
is determining the minimum number of servers needed at a

Scan to Download
 work station or service center.
6.In general, it is the variability of service times and the
variability of the arrival process that causes waiting lines to
build up and congestion to occur.
7.For most systems, if the arrival rate increases, or if the
service rate decreases, or if the variance of service times or
interarrival times increases, then the system will become
more congested.
8.These estimators could be severely biased (either high or
low) if the initial conditions are unrepresentative of steady
state or if the simulation run length is too short.
9.Thus, if the model predicts that average waiting time w will
be 12.7 minutes, then average waiting time in the real
system is likely to be less than 12.7 minutes.
10.These results do provide a compelling case for the
addition of one or more staff, and provide a baseline
against which to check a detailed simulation model.

Scan to Download
Chapter 7 | Quotes From Pages 280-301
1.The ultimate test of the linear congruential
method, as of any generation scheme, is how
closely the generated numbers R1, R2, . . .
approximate uniformity and independence.
2.The goal of any generation scheme, however, is to produce
a sequence of numbers between 0 and 1 that simulates, or
imitates, the ideal properties of uniform distribution and
independence as closely as possible.
3.Inventing techniques that seem to generate random
numbers is easy; inventing techniques that really do
produce sequences that appear to be independent,
uniformly distributed random numbers is incredibly
difficult.
4.Even if generated numbers pass all the tests (those covered
in this chapter and those mentioned in the chapter), some
underlying pattern might have gone undetected without the
generator’s having been rejected as faulty.
5.The method should be fast. Individual computations are

Scan to Download
 inexpensive, but simulation could require many millions of
random numbers.
6.The method should be portable to different
computers—and, ideally, to different programming
languages. This is desirable so that the simulation program
will produce the same results wherever it is executed.
Chapter 8 | Quotes From Pages 302-335
1.In practice, most simulation modelers will use
either existing routines available in programming
libraries or the routines built into a simulation
language.
2.Even though the chance of this happening is small, it is
nevertheless worthwhile to understand how random-variate
generation occurs.
3.To generate a value X1 with cdf F(x), a random number R1
between 0 and 1 is generated, then a horizontal line is
drawn from R1 to the graph of the cdf, then a vertical line
is dropped to the x axis to obtain X1, the desired result.
4.If the modeler has been unable to find a theoretical

Scan to Download
 distribution that provides a good model for the input data,
then it may be necessary to use the empirical distribution of
the data.
5.To obtain normal variates X i with mean ¼ and variance à 2
, we then apply the transformation X i = ¼ + Ã Z i,
Chapter 9 | Quotes From Pages 336-389
1.Regardless of the sophistication of the analyst,
faulty models of the inputs will lead to outputs
whose interpretation could give rise to misleading
recommendations.
2.Data collection is one of the biggest tasks in solving a real
problem and it is one of the most important and difficult
problems in simulation.
3.Simply having data is not enough to support effective input
modeling. Data can be out of date or 'dirty' (containing
errors).
4.The first step in the development of a useful model of input
data is to collect data from the real system of interest.
5.Understanding which input models are the most and least

Scan to Download
 reliable is important for judging the reliability of the
conclusions drawn from the simulation study.
6.It is sensible to run the simulation with several plausible
input models to see if the conclusions are robust or highly
sensitive to the choices.
7.When data are not available, expert opinion and knowledge
of the process must be used to make educated guesses.
8.It is still important to understand what the software does, so
that it can be used appropriately.
9.When input models are derived without the benefit of data,
it is particularly important to examine the sensitivity of the
results to the models chosen.

Scan to Download
Chapter 10 | Quotes From Pages 390-419
1.It is the job of the model developer to work closely
with the end users throughout the period of
development and validation to reduce this
skepticism and to increase the model’s credibility.
2.The goal of the validation process is twofold: (a) to
produce a model that represents true system behavior
closely enough for the model to be used as a substitute for
the actual system for the purpose of experimenting with the
system, analyzing system behavior, and predicting system
performance; and (b) to increase the credibility of the
model to an acceptable level, so that the model will be used
by managers and other decision makers.
3.Validation should not be seen as an isolated set of
procedures that follows model development, but rather as
an integral part of model development.
4.Validation is concerned with building the correct model. It
attempts to confirm that a model is an accurate
representation of the real system.

Scan to Download
5.The purpose of model verification is to assure that the
conceptual model is reflected accurately in the operational
model.
6.Documentation is also important as a means of clarifying
the logic of a model and verifying its completeness.
7.No model is ever totally representative of the system under
study. In addition, each revision of the model, as pictured
in Figure 3, involves some cost, time, and effort.
8.The ultimate test of a model, and in fact the only objective
test of the model as a whole, is the model’s ability to
predict the future behavior of the real system when the
model input data match the real inputs and when a policy
implemented in the model is implemented at some point in
the system.
9.A necessary condition for the validation of input-output
transformations is that some version of the system under
study exist, so that system data under at least one set of
input conditions can be collected to compare to model
predictions.

Scan to Download
10.If the current version of the bank model produces
estimates of average delay Y2 that are not close enough to
real system behavior, the source of the discrepancy is
sought, and the model is revised in light of this new
knowledge.
Chapter 11 | Quotes From Pages 420-465
1.The purpose of the statistical analysis is to acquire
some assurance that these estimates are
sufficiently precise for the proposed use of the
model.
2.A common goal in simulation is to estimate ¸ = E ( 1 n n"
i=1 Y i ).
3.Initialization bias... can be severe if run length is too short,
but generally it decreases as run length increases.
4.The statistical precision of point estimators can be
measured by a standard-error estimate or by a confidence
interval.
5.A performance measure that does not fit this common
framework is a quantile or percentile.

Scan to Download
6.The effect of the autocorrelation on the estimator S2/n is
derived by an examination of Equations (26) and (28).
7.By increasing the run length, the sample size can be
increased to n observations, Y1, Y2, . . . , Yn, based on a
run length of n weeks.
8.The results of 4 replications of the SMP simulation; each
row presents the average number of minutes callers spent
on hold, and the average number of callers on hold, during
the course of a simulated 8-hour day.
9.The method used in each case is the method of independent
replications.
10.Significant problems arise concerning the specification of
these initial and stopping conditions, problems that we
discuss later.
Chapter 12 | Quotes From Pages 466-509
1.... statistical analysis is needed to discover whether
any observed differences are due to differences in
design or merely to the random fluctuation
inherent in the models.

Scan to Download
2.Confidence intervals provide a measure of the precision of
the estimator of ¸1 " ¸2 , we want to shrink it until we can
either conclude that there is a difference that matters...
3.There is an important distinction between statistically
significant differences and practically significant
differences in systems performance.
4.Common random numbers (CRN) results in a positive
correlation between the performance measures, leading to
lower variance of the difference and more precise
estimates.
5.... it is critical to perform a second set of experiments on
the top designs identified by the heuristic... to avoid two
types of errors...
6.Most likely, if enough additional data were collected, the
c.i. would shift, and definitely shrink in length, until
conclusion 1 or 2 would be drawn.
7.Statistical significance answers the question: Is the
observed difference larger than the variability in the
measures?

Scan to Download
8.The confidence interval gives a measure of the precision of
the point estimator, thereby allowing us to assess practical
significance.

Scan to Download
Chapter 13 | Quotes From Pages 510-567
1.The purpose of simulation is insight, not numbers.
2.Simulation has been used successfully as an aid in the
design of new production facilities, warehouses, and
distribution centers.
3.Models can be developed in an iterative fashion, adding
detail for peripheral operations at later stages if such
operations are later judged to affect the main operation
significantly.
4.The insights gained from simulation can be critical in
identifying bottlenecks and areas for improvement in
manufacturing and material-handling systems.
5.Should the model simulate each conveyor section or
vehicle movement, or can some be replaced by a simple
time delay?

Scan to Download
Discrete Event System Simulation
Questions
View on Bookey Website

Chapter 1 | 1. Introduction to Simulation| Q&A

1.Question
What is simulation in the context of discrete-event
systems?
Answer:Simulation is the imitation of the operation
of a real-world process or system over time. It
involves generating an artificial history of the
system to observe and draw inferences about its
operating characteristics. The process allows
analysts to perform a variety of 'what if'
investigations regarding the behavior of the system
under different conditions.

2.Question
How can simulation benefit system design and analysis?
Answer:Simulation serves as both an analysis tool for
existing systems and a design tool for new systems. It helps

Scan to Download
predict the performance implications of potential changes
and allows for experimentation with the internal interactions
of complex systems without impacting actual operations.

3.Question
Under what circumstances is simulation not appropriate?
Answer:Simulation is not appropriate when the problem can
be solved through common sense or analytical methods,
when direct experiments are less costly, when resources or
time are insufficient, or when the complexity of the system
behavior cannot be defined.

4.Question
What are the advantages of using simulation?
Answer:Advantages of simulation include the ability to
explore new policies without disrupting actual operations,
test hardware designs without resource commitment, obtain
insights about variable interactions, and address 'what if'
analyses particularly beneficial in designing new systems.

5.Question
What challenges exist in interpreting simulation results?

Scan to Download
Answer:Simulation results can be difficult to interpret due to
their random nature. It can be challenging to distinguish
whether observed outcomes are due to underlying system
interactions or merely random variation, making careful
statistical analysis necessary.

6.Question
What steps are involved in conducting a simulation
study?
Answer:The steps include: 1) Problem formulation, 2)
Setting objectives and project planning, 3) Model
conceptualization, 4) Data collection, 5) Model translation,
6) Verification, 7) Validation, 8) Experimental design, 9)
Production runs and analysis, 10) Documentation and
reporting, and 11) Implementation.

7.Question
What role does validation play in simulation?
Answer:Validation is crucial because it ensures that the
simulation model accurately replicates the real-world
system's behavior. It typically involves iterative calibration,

Scan to Download
comparing model outputs with actual system performance
until an acceptable accuracy level is achieved.

8.Question
Can you connect simulation to real-world applications?
Provide an example.
Answer:An example of real-world application is in the
Turkish Army, which utilized simulation to model and
optimize its fuel supply system. This approach allowed for
measurable performance analysis under various scenarios,
resulting in significant savings.

9.Question
What is the importance of user involvement in the
simulation process?
Answer:Involving model users throughout the simulation
process enhances the quality of the model and increases the
users' confidence in using the results, leading to more
successful implementation and better decision-making.

10.Question
How has simulation software evolved in recent years?
Answer:Simulation software has advanced significantly,

Scan to Download
integrating new methodologies like agent-based modeling
and becoming more user-friendly through template models
that require only data inputs. This evolution allows for more
efficient simulation modeling and analysis.
Chapter 2 | 2. Simulation Examples in a
Spreadsheet| Q&A
1.Question
What are the key components of a discrete event
simulation model as discussed in Chapter 2?
Answer:The key components of a discrete event
simulation model include inputs, activities, events,
states, outputs, and responses. Inputs are
independent variables that influence the system such
as demand rates, activities are the actions that occur
within the process, events are specific occurrences
that change the state of the system, states represent
the status of the system at any point in time, outputs
are results derived from the simulation, and
responses measure system performance.

2.Question

Scan to Download
How do random numbers play a role in simulations
according to the chapter?
Answer:Random numbers are crucial for simulating
uncertainty within discrete event systems. They are used to
generate random variables that follow certain statistical
distributions representing various characteristics of the
system, such as service times, interarrival times, or demand.
The chapter emphasizes the importance of randomness in
creating realistic simulations that can help predict system
behavior.

3.Question
What is the significance of using a simulation table in a
spreadsheet for discrete event simulations?
Answer:The simulation table serves as the backbone of the
spreadsheet model, organizing data into rows and columns
that facilitate tracking system states, generating random
observations, and calculating outputs. It allows the modeler
to systematically design and conduct experiments, leading to
better analysis and insight into system performance.

Scan to Download
4.Question
Why is it recommended to conduct multiple trials or
experiments in simulations?
Answer:Conducting multiple trials or experiments allows for
a better understanding of the variability and reliability of the
simulation results. It provides a range of outputs that can be
analyzed statistically to determine average behavior,
distribution shapes, and the likelihood of different outcomes,
which can lead to more informed decision-making.

5.Question
Can you explain the Monte Carlo simulation method used
in the chapter? Give an example.
Answer:The Monte Carlo simulation method involves using
random sampling to estimate the properties of a model. For
example, simulating the number of successful coin tosses
over several trials utilizes generated random numbers to
represent each toss. By performing a high number of trials,
one can graph the distribution of heads and tails, thus
estimating probabilities statistically.

Scan to Download
6.Question
What are the limitations of using spreadsheet simulations
for complex real-world problems as highlighted in the
chapter?
Answer:Spreadsheets have limitations in handling complex
dynamic systems due to their ad hoc nature. They struggle
with intricate dependencies and multiple interacting
components, which often require more sophisticated
approaches such as discrete event simulation software or
programming languages that can manage complexity more
effectively.

7.Question
In the context of the coin tossing game discussed in the
chapter, what does the outcome reveal about probability?
Answer:The outcomes from the coin tossing game highlight
the non-intuitive nature of probability, where results from
multiple trials can show surprising distributions. For
example, it may be more likely for one player to dominate
the lead (lostly) in the majority of tosses rather than evenly
split the lead, demonstrating the influence of randomness in

Scan to Download
sequential trials.

8.Question
Explain how descriptive statistics are utilized in
simulation analysis as discussed in Chapter 2.
Answer:Descriptive statistics are used to summarize and
interpret the results of simulation runs, calculating averages,
medians, and ranges of outputs to provide a clear picture of
system performance. By summarizing data, one can assess
the effectiveness of different strategies and predict outcomes
in similar future scenarios.

9.Question
What are some practical applications of the concepts
introduced in this chapter on simulations?
Answer:The concepts introduced can be applied across
various domains including queueing systems (like banks or
stores), inventory management (like newsstands), project
scheduling, risk analysis, and even in operations research
scenarios, where decisions must be made under uncertainty.

10.Question
How does one generate random samples in a spreadsheet

Scan to Download
simulation as per the techniques described?
Answer:Random samples in a spreadsheet can be generated
using built-in functions like RAND() for uniform distribution
or custom VBA functions for more complex distributions.
The generated random values can then be converted into the
desired range or probability distribution.
Chapter 3 | 3. General Principles| Q&A
1.Question
What are the fundamental building blocks of
discrete-event simulation models?
Answer:The fundamental building blocks of
discrete-event simulation models include entities,
attributes, activities, events, and the system state.
Entities are the objects within the system, attributes
are the properties of these entities, activities refer to
the tasks that entities undertake, and events are
occurrences that cause changes in the system state.

2.Question
How does discrete-event simulation differ from

Scan to Download
continuous simulation?
Answer:Discrete-event simulation specifically models
systems where changes only occur at distinct points in time
(events), while continuous simulation represents systems
where changes can happen at any time.

3.Question
What is the significance of the Future Event List (FEL) in
discrete-event simulation?
Answer:The Future Event List (FEL) is crucial in organizing
and managing events that are scheduled to occur in the
simulation. It ensures that events are executed in
chronological order, allowing the simulation to accurately
represent the dynamic behavior of the system over time.

4.Question
What methodologies can be used to handle lists in
discrete-event simulation software?
Answer:Methodologies for handling lists in discrete-event
simulation include basic list operations like adding and
removing records, as well as more advanced techniques like

Scan to Download
dynamically allocated records and linked lists for efficiency
in managing complex data.

5.Question
What is the difference between an activity and a delay in
discrete-event simulation?
Answer:An activity has a defined duration known in
advance, such as a service time; while a delay has an
undefined duration that is determined by system conditions,
such as waiting in line.

6.Question
Can you give an example of how events trigger activities
in a simulation model?
Answer:In a call center simulation, when a caller arrives (an
event), it may trigger the start of a service activity if a server
is available. If all servers are busy, the caller experiences a
delay until one becomes free.

7.Question
In terms of system state, what information is necessary to
describe a system at any time?
Answer:The system state is defined by a collection of

Scan to Download
variables that include the number of entities in the system,
their attributes, the contents of any lists (such as queues), and
the status of ongoing activities.

8.Question
What are the three prevalent world views in
discrete-event simulation?
Answer:The three prevalent world views in discrete-event
simulation are: event scheduling, which focuses on events
and their impact on state; process interaction, which views
the life cycles of entities flowing through a system; and
activity scanning, which checks conditions enabling
activities to begin.

9.Question
How do you define an event in the context of simulation?
Answer:An event is an instantaneous occurrence that alters
the state of a system, such as a customer’s arrival or the
completion of a service.

10.Question
What factors determine whether an activity can begin or
if it must wait?

Scan to Download
Answer:An activity can begin based on the availability of
resources, the completion of previous activities, and the
specific conditions defined by the model. If conditions are
not met, the activity must wait until those conditions change.

Scan to Download
Chapter 4 | 4. Simulation Software| Q&A
1.Question
What are the main considerations when selecting
simulation software?
Answer:1. Cost: Affordability within budget
constraints.
2. Ease of Learning: Simplicity of the learning curve
for newcomers.
3. Ease of Use: User-friendly interfaces and
functionalities.
4. Modeling Power: Ability to model complex
systems relevant to specific needs.
5. Animation Capabilities: Quality and clarity of
visualizations provided.

2.Question
How has simulation software evolved over the years?
Answer:The history of simulation software can be divided
into several periods:
1. 1955-1960: The Period of Search focused on developing

Scan to Download
initial programming languages and reusable routines.
2. 1961-1965: The Advent saw the emergence of simulation
programming languages like GPSS and SIMSCRIPT.
3. 1966-1970: The Formative Period refined language
features and extended capabilities.
4. 1971-1978: During the Expansion Period, interactive and
faster versions of existing languages debuted.
5. 1979-1986: The Period of Consolidation and Regeneration
marked the adaptation of software for microcomputers.
6. 1987-2008: Integrated Environments emerged with
advanced functionalities and graphical interfaces.
7. 2009-2011: The future pointed towards improved
interfaces, better animations, and the adoption of agent-based
modeling.

3.Question
What is agent-based modeling and its significance in
simulation?
Answer:Agent-based modeling simulates the actions and
interactions of autonomous individuals (agents) to

Scan to Download
understand their effects on the whole system. Each agent
makes decisions based on local conditions and can exhibit
behaviors such as learning and adaptation. Its significance
lies in capturing the complexity of systems where individual
behaviors lead to emergent phenomena, making it especially
useful in fields like economics, social sciences, and
environmental studies.

4.Question
What advice is offered for evaluating simulation
software?
Answer:1. Avoid focusing solely on one feature. Consider
overall usability, accuracy, vendor support, and application
relevance.
2. Prioritize execution speed for efficient debugging and
development.
3. Scrutinize vendor claims and advertisements for true
representations of capabilities.
4. Test the software with a small version of your problem
before purchasing.

Scan to Download
5.Question
What statistical tools are commonly integrated into
simulation environments?
Answer:Common statistical tools include:
1. Summary statistics: Means, variances, etc.
2. Confidence intervals: For assessing the precision of
estimates.
3. Hypothesis testing tools: For comparing different
simulation scenarios.
4. Design of experiments: For systematic exploration of input
parameter effects.
5. Optimization techniques: Such as genetic algorithms or
other heuristic methods to find near-optimal solutions.

6.Question
Give an example of how to simulate using a programming
language such as Java.
Answer:In Java, a discrete-event simulation involves:
1. Defining a simulation clock to track time.
2. Setting up event lists to manage events based on

Scan to Download
time-stamps.
3. Implementing event routines to handle actions when
events occur, updating the simulation state.
4. Gathering statistics on state changes and system
performances.
5. Running simulations until a defined stopping criterion,
while continuously updating system states based on event
triggers.
Chapter 5 | 5. Statistical Models in Simulation| Q&A
1.Question
What role do statistical models play in discrete event
simulation?
Answer:Statistical models are crucial in discrete
event simulation as they help represent the
probabilistic nature of real-world phenomena. Since
many events occur at random and are subject to
variation, statistical models provide a framework
for understanding and predicting these occurrences,
thereby allowing for more accurate simulations.

Scan to Download
2.Question
Why is it important to understand both discrete and
continuous random variables in simulation?
Answer:Understanding both discrete and continuous random
variables is essential as they apply to different types of data.
Discrete random variables deal with countable outcomes,
such as the number of arrivals in a queue, while continuous
random variables represent data that can take any value in an
interval, like time to failure of a component. This knowledge
helps in accurately modeling different scenarios in
simulations.

3.Question
How does the Poisson distribution relate to event arrival
in simulations?
Answer:The Poisson distribution is key in modeling the
number of events that occur in a fixed interval of time or
space, especially when the events happen independently and
with a constant mean rate. It is widely used in discrete event
simulations for modeling arrivals, service times, and system

Scan to Download
breakdowns, providing a clear understanding of system
performance.

4.Question
What is the significance of the memoryless property
characteristic of certain distributions?
Answer:The memoryless property, particularly seen in the
exponential distribution, implies that the future probability of
an event does not depend on how much time has already
passed. This is significant in simulations as it simplifies
calculations, particularly in queuing models where the timing
of future arrivals does not influence their probability.

5.Question
In practical terms, how might a simulation analyst go
about selecting appropriate statistical distributions for a
model?
Answer:A simulation analyst typically starts by collecting
data relevant to the phenomena being modeled. They then
analyze this data, potentially visualizing it through
histograms to identify patterns. Based on the observed
patterns, they can hypothesize the appropriate statistical

Scan to Download
distributions, validate their assumptions through
goodness-of-fit tests, and refine their choices using statistical
software.

6.Question
Can you explain the differences between empirical
distributions and parametric distributions?
Answer:Empirical distributions are based solely on observed
data without assuming any underlying distribution, making
them flexible but potentially limited if the sample does not
cover the entire range of possible values. In contrast,
parametric distributions are defined by a small number of
parameters (like mean and variance) and can describe data
that fits a particular known form, providing a framework for
statistical inference.

7.Question
What are some common applications of statistical models
in discrete event simulations?
Answer:Common applications include modeling queueing
systems (like customer arrivals and service times), inventory

Scan to Download
systems (demand patterns and lead times), reliability and
maintainability analysis (time to failure), and event
occurrences in telecommunications and computer networks.
Each of these applications relies on appropriate statistical
models to accurately represent the processes involved.

8.Question
Discuss how variance is interpreted in the context of
simulation studies. How does it help in decision-making?
Answer:Variance measures the spread of data points in
relation to the mean, indicating the degree of uncertainty in
the estimates provided by the simulation. A high variance
suggests a wide range of possible outcomes, which can
influence risk assessments and decision-making. By
analyzing variance, simulation analysts can derive insights
about the stability and reliability of models, thereby guiding
informed strategic decisions.

9.Question
What are the advantages and limitations of using the
uniform distribution in modeling when data is limited?

Scan to Download
Answer:The uniform distribution is advantageous when there
is no prior knowledge about the shape of the distribution, as
it assumes all values within a specified range are equally
likely. However, this 'maximum ignorance' approach can
lead to significant inaccuracies if actual data presents a
non-uniform distribution. The limitations arise when using
uniform distribution inappropriately without further data to
justify its assumed randomness.

10.Question
How do empirical distributions assist in understanding
complex systems during simulations?
Answer:Empirical distributions allow analysts to create
models based directly on observed data, which helps in
accurately reflecting the complexities of real-world systems.
They adapt to the specific characteristics of the data
collected, leading to models that can capture unpredictable
variations and behaviors inherent in the system being
studied.
Chapter 6 | 6. Queueing Models| Q&A

Scan to Download
1.Question
What is the significance of queueing models in the
analysis of systems?
Answer:Queueing models are important as they
provide a framework for analyzing the behavior and
performance of various service systems, helping to
understand factors affecting customer wait times,
service efficiency, and overall system performance.
They allow analysts to evaluate trade-offs between
service quality and resource utilization.

2.Question
How do input parameters influence the performance
measures of a queueing system?
Answer:Input parameters like the arrival rate of customers,
service demands, and server arrangement directly affect
performance measures such as server utilization, waiting line
lengths, and delays. Understanding these relationships allows
management to optimize performance based on controllable
factors.

Scan to Download
3.Question
What are the trade-offs faced by analysts when designing
or improving queueing systems?
Answer:Analysts often face trade-offs between maximizing
server utilization and minimizing customer wait times or
delays. This balance is crucial for enhancing customer
satisfaction while maintaining operational efficiency.

4.Question
What is the role of simulation in queueing analysis
compared to analytical methods?
Answer:Simulation allows for the modeling of complex
queueing systems that are difficult to evaluate
mathematically, providing insights into system dynamics and
potential performance under various conditions. While
analytical methods offer quicker estimates through
mathematical formulas, simulation can capture variability
and non-linear behavior in real systems.

5.Question
Can you explain the difference between finite and infinite
calling populations in queueing systems?

Scan to Download
Answer:In finite calling populations, the arrival rate is
affected by the number of customers already being served or
waiting, meaning when all customers are in the queue, new
arrivals cannot occur. In contrast, infinite populations assume
that the arrival rate remains constant regardless of how many
customers are in the system.

6.Question
What measures can be taken to reduce congestion in
queueing systems?
Answer:Congestion can be reduced by increasing the number
of servers, improving service speeds, managing arrival rates
(e.g., through scheduling), and minimizing variability in
service times.

7.Question
Why can increasing server utilization lead to longer
waiting times?
Answer:Higher server utilization means that servers are
busier and less available to handle incoming customers,
which can lead to queues forming, as customers wait longer

Scan to Download
for service, thereby increasing average waiting times and
system congestion.

8.Question
How can the understanding of coefficient of variation (cv)
impact queue management decisions?
Answer:The coefficient of variation helps assess the
variability of service times relative to the mean. A higher cv
indicates more unpredictability in service duration, which
can lead to longer queues and waiting periods. Thus,
reducing cv through process improvements can enhance
service reliability and customer experience.

9.Question
What is an example of a real-world application of
queueing models?
Answer:Queueing models are used in various settings, such
as hospitals for managing patient flows, banks for optimizing
teller service times, call centers for routing incoming calls
efficiently, and manufacturing for managing production lines
to balance workloads and reduce idle times.

Scan to Download
10.Question
What does Little's Law tell us in the context of queueing
systems?
Answer:Little's Law states that the average number of
customers in a queuing system (L) is equal to the arrival rate
(») multiplied by the average time a customer spends in the
system (w). This fundamental relationship helps derive key
performance metrics from basic input parameters.

Scan to Download
Chapter 7 | 7. Random-Number Generation| Q&A
1.Question
Why is randomness important in discrete event system
simulations?
Answer:Randomness is crucial in discrete event
system simulations because it allows for the
modeling of uncertain and unpredictable future
events, which helps in accurately predicting system
behavior under various conditions.

2.Question
What are the two main statistical properties that a
sequence of random numbers should exhibit?
Answer:A sequence of random numbers should exhibit
uniformity and independence. Uniformity ensures that
numbers are evenly distributed across the interval [0, 1],
while independence guarantees that each number drawn does
not influence the next.

3.Question
What does it mean when we say that pseudo-random
numbers are generated?

Scan to Download
Answer:Pseudo-random numbers are generated by an
algorithm that produces numbers appearing random but are
actually deterministic and can be replicated if the starting
conditions (seed) are known.

4.Question
What are some of the common errors when generating
pseudo-random numbers?
Answer:Common errors include non-uniform distribution,
discrete values instead of continuous, incorrect mean (too
high or too low), high or low variance, and dependencies
shown by autocorrelation.

5.Question
Why should random-number generation methods be
computationally efficient?
Answer:Because simulations often require millions of
random numbers, inefficient generation methods can
significantly slow down the entire simulation process, so fast
methods help manage total computational costs.

6.Question
What is the significance of a long cycle in

Scan to Download
random-number generators?
Answer:A long cycle ensures that the sequence of generated
numbers does not repeat too soon, providing a greater variety
of random values and enhancing the reliability of
simulations.

7.Question
How can we test the randomness of generated numbers?
Answer:Randomness can be tested using frequency tests such
as the Kolmogorov-Smirnov test or chi-square test to check
for uniformity, and autocorrelation tests to verify
independence.

8.Question
What is the linear congruential method for generating
random numbers?
Answer:The linear congruential method generates random
integers using a recursive relation based on a seed and
parameters (multiplier, increment, and modulus), which can
then be transformed into random floating-point numbers
between 0 and 1.

Scan to Download
9.Question
What is the role of the seed in pseudo-random number
generation?
Answer:The seed initializes the random-number sequence
and allows for the generation of the same sequence of
numbers when reused, which can be useful for debugging or
testing.

10.Question
What cautions should one take when interpreting the
results of randomness tests?
Answer:Even if a set of generated numbers passes all tests
for randomness, an underlying pattern may still exist that
hasn't been detected. Therefore, it’s important to continue
validating generators and consult established literature on the
statistical properties of random number generators.
Chapter 8 | 8. Random-Variate Generation| Q&A
1.Question
What is the importance of random variate generation in
discrete event system simulation?
Answer:Random variate generation is crucial in

Scan to Download
discrete event system simulation because it allows
modelers to create samples from various probability
distributions that represent uncertain and
unpredictable phenomena, such as interarrival times
and service times. These samples serve as input for
simulation models, enabling realistic modeling of
systems where randomness plays a significant role.

2.Question
How does the inverse-transform technique for random
variate generation work?
Answer:The inverse-transform technique involves the
following steps: 1) Compute the cumulative distribution
function (CDF) of the desired random variable; 2) Generate a
uniform random number R from the interval [0, 1]; 3) Solve
the equation F(X) = R to find the desired random variable X.
For example, for an exponential distribution, it results in X =
-(1/») ln(R), where » is the rate parameter.

3.Question
Can the acceptance-rejection technique be applied to any
probability distribution?

Scan to Download
Answer:The acceptance-rejection technique can be applied to
a variety of distributions, particularly when generating
samples from those distributions is challenging. However, its
efficiency depends on minimizing the number of rejections,
and it is particularly effective for generating random variates
from distributions that are difficult to sample directly.

4.Question
Why might a modeler choose to construct their own
random variate generation routine?
Answer:Modelers might opt to construct their own random
variate generation routines if existing programming libraries
lack routines for certain distributions, or if specific
distributions have unique requirements that standard libraries
do not meet. Understanding the underlying mechanisms can
also provide deeper insight into the simulation model's
behavior.

5.Question
What is a specific example of generating an exponential
variate using the inverse-transform technique?

Scan to Download
Answer:To generate an exponential variate using the
inverse-transform technique, first compute the CDF: F(x) = 1
- e^(-»x). Generate a uniform random number R. Solve for X
in the equation R = 1 - e^(-»X). The solution yields X = -
(1/») ln(1 - R) or equivalently, X = - (1/») ln(R), since R and
(1-R) are identically distributed.

6.Question
Why is understanding random variate generation
methods beneficial for simulation modelers?
Answer:Understanding random variate generation methods
helps simulation modelers make informed decisions about
which techniques to use for their specific applications. It
allows them to tailor the generation process to their model's
needs, potentially improve efficiency, and handle situations
where built-in functions may not meet the requirements.

7.Question
What role does the gamma distribution play in random
variate generation?
Answer:The gamma distribution is often used to model

Scan to Download
waiting times and is characterized by its shape and scale
parameters. Techniques for generating gamma random
variates include acceptance-rejection and convolution
methods. Understanding these methods is critical, especially
when dealing with processes represented by sums of
exponential random variables.

8.Question
How can modelers generate values from empirical
continuous distributions?
Answer:Modelers can generate values from empirical
continuous distributions by arranging observed data points,
assigning probabilities to intervals, and using interpolation
methods. The empirical cumulative distribution function
(CDF) can be constructed from the observed data, allowing
for random sampling based on the probabilities assigned to
different intervals.

9.Question
What considerations should modelers keep in mind when
choosing a random variate generation method?

Scan to Download
Answer:When choosing a random variate generation method,
modelers should consider the numerical efficiency of the
method, the computational resources available, the precision
required, the characteristics of the distribution, and whether
the method is suitable for the specific input data
characteristics.

10.Question
What is the benefit of generating empirical distributions
when theoretical distributions are inadequate?
Answer:Empirical distributions provide a way to accurately
reflect the observed data when theoretical distributions do
not fit well. They allow for a more data-driven approach,
capturing the real behavior of the system being modeled,
which can lead to more accurate simulations and insights into
system performance.
Chapter 9 | 9. Input Modeling| Q&A
1.Question
Why are input models crucial in simulation, and what are
the risks associated with using faulty models?

Scan to Download
Answer:Input models are essential because they
provide the driving force for simulations, such as the
distributions of time between arrivals in queueing
systems or demand in supply-chain simulations. If
input models are incorrect or based on outdated,
inaccurate, or insufficient data, the simulation
outputs could lead to misleading recommendations,
resulting in poor decision-making and potential
failure in real-world applications.

2.Question
What are the four steps in developing a useful model of
input data, as outlined in the chapter?
Answer:The four steps are: 1. Collect data from the real
system, which can be time-consuming and
resource-intensive. 2. Identify an appropriate probability
distribution to represent the input process by analyzing the
collected data. 3. Choose specific parameters for the
identified distribution, potentially estimated from the data. 4.
Evaluate the fit of the chosen distribution through

Scan to Download
goodness-of-fit tests to ensure it accurately represents the
data.

3.Question
Can you describe a scenario where 'stale data' might lead
to incorrect conclusions in a simulation study?
Answer:In a simulation study at a cancer screening clinic, if
analysts use data from 2002 on how nurses took patient
histories, that could be problematic if the current method has
changed (e.g., moving from paper to digital entries in 2008).
Relying on old data might create a model that misrepresents
current processes, leading to ineffective changes based on
inaccurate assumptions.

4.Question
What is the importance of checking for relationships
between variables in data collection, and how can this
impact the resulting model?
Answer:Investigating relationships, such as correlations
between demand and lead time, is critical because treating
variables as independent when they are not can create
inaccurate simulations. For instance, in supply chain

Scan to Download
scenarios, increased demand may extend lead times,
necessitating a multivariate input model, rather than simple
independence assumptions, to accurately reflect the system's
dynamics and dependencies.

5.Question
How can expert opinion be leveraged in cases where data
collection is impossible?
Answer:When data cannot be collected due to time
constraints or prohibitive costs, expert opinions can be
sought to make educated guesses about the distribution
characteristics. Experts can provide insights into expected
values, variability, and the range of possible outcomes,
potentially creating a more informed but tentative input
model until actual data can be gathered.

6.Question
What is the role of goodness-of-fit tests in input modeling,
and what considerations should be made when
interpreting their results?
Answer:Goodness-of-fit tests, like chi-square and
Kolmogorov-Smirnov tests, help evaluate whether the chosen

Scan to Download
distribution accurately represents the data. Important
considerations include sample size—small samples are less
likely to reject incorrectly hypothesized distributions, while
large samples might reject many distributions and lead to
assumptions that don't account for practical model
application. Thus, the results should be one of several pieces
of evidence leading to the final modeling decision.

7.Question
How does the NORTA (normal-to-anything)
transformation facilitate the modeling of dependent input
processes with non-normal distributions?
Answer:NORTA allows for the generation of correlated
random variables that do not follow normal distributions by
starting from a bivariate normal distribution and
transforming the marginals using their respective cumulative
distribution functions (CDFs). This enables the creation of
multivariate models that maintain specified correlations
while conforming to desired marginal distributions,
ultimately leading to more accurate simulations.

Scan to Download
Chapter 10 | 10. Verification, Calibration, and
Validation of Simulation Models| Q&A
1.Question
What is the significance of verification and validation in
simulation models?
Answer:Verification and validation are crucial for
ensuring the accuracy and credibility of simulation
models. Verification confirms that the model has
been implemented correctly, while validation
establishes that the model accurately represents the
real system and can produce reliable predictions and
recommendations. This process fosters trust among
users, which is essential for informed
decision-making.

2.Question
How does the iterative process of calibration improve
model accuracy?
Answer:Calibration involves repeatedly comparing the
model's predictions with actual system behavior, adjusting
the model in response to any discrepancies observed. This

Scan to Download
iterative process helps refine the model and increases its
accuracy over time, as each cycle allows for deeper insights
into the system's complexities.

3.Question
What are some common techniques used during the
model verification process?
Answer:Common techniques include peer review by an
expert in simulation software, creating flow diagrams to trace
logical paths, examining model outputs for reasonableness,
animating the model to visualize behavior, and documenting
every variable and assumption clearly.

4.Question
How can graphical interfaces assist in the verification and
validation of models?
Answer:Graphical interfaces provide a visual representation
of the model, making it easier to understand and navigate.
They help users intuitively grasp complex systems and
facilitate the identification of errors or irregularities during
the verification and validation processes.

Scan to Download
5.Question
What role does user involvement play in building credible
simulation models?
Answer:User involvement is vital as it ensures that
real-world knowledge and insights are integrated into the
model from its conception through implementation.
Engaging stakeholders enhances face validity, as they can
evaluate the model's outputs for reasonableness, and fosters
acceptance of the model by demonstrating its relevance and
accuracy.

6.Question
Why is it important to perform sensitivity analysis in
simulation models?
Answer:Sensitivity analysis helps to determine how changes
in input variables affect model outputs. It provides insights
into the stability and robustness of the model, guiding users
on how various factors impact system performance and
thereby aiding in strategic decision-making.

7.Question
What are the consequences of failing to properly validate

Scan to Download
a simulation model?
Answer:Failing to validate a model can lead to misguided
decisions based on inaccurate predictions. This can result in
operational inefficiencies, wasted resources, and the potential
failure of projects or systems that rely on these flawed
models, ultimately undermining trust in simulation as a
decision-support tool.

8.Question
How can statistical tests contribute to the validation
process of simulation models?
Answer:Statistical tests provide an objective means to
compare model outputs against observed data from the real
system. They help to establish whether the model's
predictions are consistent with actual performance, allowing
modelers to quantify discrepancies and assess the model’s
validity.

9.Question
What is the concept of 'face validity' in simulation
modeling?

Scan to Download
Answer:Face validity refers to the extent to which a model
appears credible and reasonable to users or experts familiar
with the real system. It is an initial subjective assessment that
helps ensure the model is grounded in realistic assumptions
and behaviors before more formal validation processes are
conducted.
Chapter 11 | 11. Estimation of Absolute
Performance| Q&A
1.Question
What is the purpose of output analysis in simulation
studies?
Answer:Output analysis in simulation studies aims
to evaluate the performance of a system based on
data generated by simulations. This includes
predicting system performance and comparing
different designs to determine the most effective
option.

2.Question
Why is random variability an important consideration in
simulation output data?

Scan to Download
Answer:Because simulations often rely on random-number
generators, the output data can exhibit randomness, meaning
that using different random numbers can lead to different
performance outcomes. Recognizing and analyzing this
variability is crucial for making accurate estimates of system
performance.

3.Question
What is the distinction between absolute performance
and relative performance in simulation?
Answer:Absolute performance refers to estimating specific
performance measures of a single system, while relative
performance involves comparing the performance of two or
more systems.

4.Question
How does autocorrelation affect the analysis of simulation
output data?
Answer:Autocorrelation implies that output data points are
not statistically independent, which complicates traditional
statistical analysis methods. It can affect the validity of

Scan to Download
confidence intervals and standard errors calculated using
common statistical techniques.

5.Question
What are the consequences of poor initial conditions in
simulations?
Answer:Beginning the simulation with unrealistic initial
conditions can bias the output, causing it to misrepresent the
steady-state behavior of the system. This necessitates careful
selection and management of initial conditions to improve
reliability.

6.Question
What methods can be used to mitigate initialization bias
in steady-state simulations?
Answer:Methods to mitigate initialization bias include
intelligent initialization, where the system starts in a state
closer to steady conditions, and the separation of the
simulation into an initialization phase and a data-collection
phase.

7.Question
Why is the method of independent replications preferred

Scan to Download
in simulations?
Answer:The method of independent replications helps
achieve statistically independent observations, allowing for
valid statistical inferences and reliable estimates of
performance measures.

8.Question
What statistical tools can be used to estimate confidence
intervals for simulated performance measures?
Answer:Statistical tools include confidence intervals based
on the sample mean and standard error; these can be
calculated using formulas involving t-distribution or normal
distribution approximations, depending on the data
characteristics.

9.Question
How do batch means improve analysis in long-run
simulations?
Answer:Batch means, which involve grouping observations
into batches and analyzing their means, can help reduce the
impact of autocorrelation and improve the estimation of

Scan to Download
variability and confidence intervals.

10.Question
What role does the choice of sample size play in
simulation studies?
Answer:The sample size affects the precision of the
estimates. Larger sample sizes typically reduce uncertainty in
confidence intervals but require more computation time;
thus, a balance must be achieved based on the simulation
goals and resource constraints.

11.Question
Why is understanding the stochastic nature of simulations
essential for analysts?
Answer:Because simulations often model inherently
unpredictable systems, understanding the stochastic nature
allows analysts to make informed decisions based on
statistical methods tailored to handle variability and
randomness.

12.Question
What potential issues arise from improper application of
statistical methods to simulation data?

Scan to Download
Answer:Using statistical methods that assume independence
on correlated data can lead to incorrect conclusions, as it may
underestimate or overestimate the precision of performance
measures, resulting in misleading interpretations.

13.Question
What implications does the design of the simulation have
on output analysis?
Answer:The design of a simulation, including aspects like
initial conditions, run lengths, and replication strategies,
greatly influences the validity of the output analysis, as poor
design can lead to biased results and unreliable estimates.

14.Question
What is the importance of reporting both point estimates
and measures of error in simulation results?
Answer:Reporting point estimates along with measures of
error, such as confidence intervals and standard errors,
provides a clearer picture of the reliability of the estimates,
helping stakeholders to understand the uncertainty associated
with the simulations.

Scan to Download
15.Question
How can software tools enhance the simulation analysis
process?
Answer:Software tools can simplify the implementation of
complex statistical analyses, automate calculations, and
provide visualizations of results, enabling users to more
efficiently interpret data and derive meaningful insights.

16.Question
What are the common statistical methods for output
analysis of terminating versus steady-state simulations?
Answer:For terminating simulations, methods often focus on
independent replications and sample statistics. In contrast,
steady-state simulations might use batch means or
time-weighted averages to reduce initialization bias and
analyze long-run behaviors.

17.Question
Why is it critical to address both error measurement and
risk in simulation analysis?
Answer:While error measurement assesses the accuracy of
estimates, risk reflects the inherent variability and

Scan to Download
uncertainty in the system's performance. Addressing both is
essential for robust decision-making based on simulation
results.

18.Question
What role does the design choice of run length play in
reducing bias in simulations?
Answer:Increasing run length allows for more data collection
beyond the transient phase where biases may be present,
thereby yielding estimates that are more reflective of the
steady-state behavior of the system.

19.Question
How do replications contribute to the reliability of
simulation models?
Answer:Replications ensure that a diverse range of scenarios
is tested, thus capturing the variability in system performance
and facilitating the creation of statistically valid conclusions
regarding output measures.

20.Question
Can correlations among data points affect simulation
results, and how should they be handled?

Scan to Download
Answer:Yes, correlations can lead to biased estimates and
misleading confidence intervals. They should be handled
through methods like modifying statistical techniques to
account for autocorrelation or utilizing strategies like batch
means.

21.Question
How does the concept of confidence intervals differ
between terminating and steady-state simulations?
Answer:In terminating simulations, confidence intervals are
typically based on independent replications, while in
steady-state simulations, special care is needed due to
potential autocorrelation, requiring more sophisticated
methods for accurate estimation.
Chapter 12 | 12. Estimation of Relative
Performance| Q&A
1.Question
What is the primary focus of Chapter 12 in 'Discrete
Event System Simulation'?
Answer:The primary focus of Chapter 12 is the
estimation of relative performance through

Scan to Download
statistical methods to compare the performance of
two or more system designs using simulation.

2.Question
Why is statistical analysis necessary in simulating system
designs?
Answer:Statistical analysis is necessary because observations
of response variables in simulation contain random variation,
and it helps determine whether observed differences in
performance are due to actual design differences or just
random fluctuations.

3.Question
What are the two sampling techniques discussed for
comparing two system designs?
Answer:The two sampling techniques discussed are
independent sampling and common random numbers (CRN),
where CRN uses the same random numbers in simulations of
different system designs to reduce variance in estimates.

4.Question
How does the common random numbers (CRN)
technique improve simulation results?

Scan to Download
Answer:The CRN technique improves simulation results by
reducing the variance of the estimated difference between
performance measures, leading to more precise confidence
intervals and estimates.

5.Question
What conclusions can be drawn from the confidence
intervals in comparing two systems?
Answer:The conclusions based on confidence intervals can
indicate whether one system design is statistically
significantly better than another, depending on whether the
confidence interval for the mean difference lies entirely
above or below zero, or if it includes zero.

6.Question
What distinction is made between statistical significance
and practical significance in system performance
comparisons?
Answer:Statistical significance refers to whether the
observed difference is greater than the variability in
estimates, while practical significance concerns whether the
difference is large enough to matter in real-world

Scan to Download
decision-making.

7.Question
How can the number of replications affect the conclusions
drawn from simulation experiments?
Answer:Increasing the number of replications usually leads
to narrower confidence intervals, which can shift conclusions
regarding which system design is better, especially if the
initial data was not sufficient.

8.Question
What is the 'Select-the-Best Procedure' mentioned in the
chapter?
Answer:The Select-the-Best Procedure is a statistical
methodology used to identify the best system design from a
set of alternatives while ensuring a specified probability of
correct selection.

9.Question
Why is metamodeling discussed in the context of
simulation?
Answer:Metamodeling is discussed as a way to approximate
complex simulation models with simpler mathematical

Scan to Download
relationships, making it easier to analyze performance across
different input variables.

10.Question
What is a significant challenge in optimization via
simulation according to the chapter?
Answer:A significant challenge is that the performance of
system designs can only be evaluated through simulation,
introducing random variability into the estimates, which
complicates the optimization process and may lead to
uncertainties in selection.

Scan to Download
Chapter 13 | 13. Simulation of Manufacturing and
Material-Handling Systems| Q&A
1.Question
What role does simulation play in designing
manufacturing and material-handling systems?
Answer:Simulation serves as a vital tool in the
design of new production facilities and warehouses.
It allows engineers to evaluate the impact of changes
before they are implemented, serving as a 'test
drive' for proposed improvements. This minimizes
disruption to existing processes while providing
insights into potential performance outcomes.

2.Question
How can proper scope and level of detail impact a
simulation project?
Answer:Scope defines the boundaries of what is included in
the model, while the level of detail determines how
thoroughly components are represented. A well-defined
scope and appropriate detail are crucial for answering
specific questions and achieving accurate insights into

Scan to Download
system performance, while a lack of clarity can lead to
incomplete or irrelevant results.

3.Question
What is the significance of treating downtimes and
failures in simulation models?
Answer:Understanding and accurately modeling downtimes
are crucial as they can significantly impact system
performance. Ignoring random unscheduled downtimes leads
to unrealistic simulations. Statistical distributions for time to
failure and repair should be used to replicate real-world
variability.

4.Question
What is a trace-driven model and its benefits in
manufacturing simulations?
Answer:A trace-driven model utilizes actual historical data
for inputs, ensuring more accurate representation of
processes like order profiles, which enhances the credibility
and reliability of the model's outputs. It eliminates
misestimations that could arise from using independent

Scan to Download
distributions.

5.Question
How can simulation help managers evaluate staffing
needs?
Answer:Simulation assists in quantifying staffing
requirements by modeling various operating conditions and
determining the impact of different staffing levels on system
performance. It helps visualize workloads and identify
optimal staffing patterns to meet production demands
without excess labor costs.

6.Question
What are the common measures of system performance
in manufacturing simulations?
Answer:Common measures include throughput, cycle time,
resource utilization, bottlenecks, queue lengths, and the
efficiency of scheduling and control systems. These metrics
help identify areas for improvement and inform strategic
decisions.

7.Question
How does the setup and management of a simulation

Scan to Download
environment influence its outcomes?
Answer:The design of the simulation environment, including
the selection of appropriate models, tools, parameters, and
input data, can significantly affect the simulation's accuracy
and relevance. A well-thought-out environment enables
better analysis and clearer insights into system behavior.

8.Question
Explain the iterative nature of developing simulation
models. Why is this approach beneficial?
Answer:Developing simulation models iteratively allows
analysts to start with a basic model and progressively add
details as needed, ensuring that the model evolves in
response to the insights gained during testing. This approach
helps focus efforts on relevant aspects without
overcomplicating the initial design.

9.Question
Can you give an example of how simulation was used to
identify performance improvement areas in
manufacturing?

Scan to Download
Answer:In one case, simulation was used to analyze the
performance of a production line where bottlenecks were
suspected. By simulating different configurations, engineers
could visualize how changes in workstation capacity and
operator allocation impacted throughput, leading to targeted
improvements.

10.Question
What challenges do analysts face when determining the
level of detail for a manufacturing simulation?
Answer:Analysts must balance the need for comprehensive
data against the potential complexity and resource
requirements of the simulation. Limited data and uncertainty
about how components interact can complicate decisions on
how detailed the model should be, potentially impacting the
validity of the results.

Scan to Download
Discrete Event System Simulation Quiz
and Test
Check the Correct Answer on Bookey Website

Chapter 1 | 1. Introduction to Simulation| Quiz and

Test
1.Simulation exclusively relies on analytical
solutions to model real-world processes.
2.Simulation is a valuable tool for understanding complex
system interactions and improving system performance.
3.Simulation is always the most cost-effective method to
solve problems in every scenario.
Chapter 2 | 2. Simulation Examples in a
Spreadsheet| Quiz and Test
1.Random numbers are essential to represent
uncertainty in discrete-event system simulation.
2.The chapter emphasizes that simulation models should
always be complex to yield accurate results.
3.Queueing models are discussed in the chapter as examples
of managing service times and arrival processes.

Scan to Download
Chapter 3 | 3. General Principles| Quiz and Test
1.Dynamic and stochastic systems are the only types
of systems discussed in the chapter.
2.In discrete-event simulation, an event is defined as an
occurrence that alters the system state.
3.The event scheduling world view focuses on the life cycles
of entities rather than events driving system changes.

Scan to Download
Chapter 4 | 4. Simulation Software| Quiz and Test
1.Simulation software has been developed over
various decades and is currently in its middle-age
phase.
2.The major language GPSS underwent significant changes
during the period from 1961 to 1965.
3.AnyLogic supports both discrete-event and agent-based
modeling.
Chapter 5 | 5. Statistical Models in Simulation| Quiz
and Test
1.Discrete random variables can take on a
continuous range of values.
2.The expectation of a random variable is always equal to its
mode.
3.Empirical distributions are built from real observed data,
reflecting actual scenarios without assuming a theoretical
distribution.
Chapter 6 | 6. Queueing Models| Quiz and Test
1.Queueing Models are only applicable in service

Scan to Download
 facilities and cannot be used in production or
transport systems.
2.Kendall's notation is used to classify queue systems based
solely on the number of customers.
3.The arrival process of customers in queueing systems is
often modeled as Poisson processes.

Scan to Download
Chapter 7 | 7. Random-Number Generation| Quiz
and Test
1.Uniformity and independence are crucial
properties of random numbers generated for
discrete event simulation.
2.Pseudo-random numbers are truly random and do not rely
on defined methods for generation.
3.The linear congruential method can produce sequences of
integers which are then converted into random numbers in
the interval [0, 1].
Chapter 8 | 8. Random-Variate Generation| Quiz
and Test
1.The inverse-transform technique can be applied to
both continuous and discrete distributions.
2.Random variates generated from a uniform distribution are
defined as X = a + (b - a)R.
3.The acceptance-rejection technique can only be used for
generating random variates from the normal distribution.
Chapter 9 | 9. Input Modeling| Quiz and Test
1.Input models are fundamental for providing

Scan to Download
 essential data in simulations.
2.Collecting data for input modeling is simple and does not
require significant resources.
3.Multivariate models are unnecessary if data variables are
dependent on each other.

Scan to Download
Chapter 10 | 10. Verification, Calibration, and
Validation of Simulation Models| Quiz and Test
1.Verification focuses on ensuring that the model
accurately reflects the real system's behavior.
2.Calibration and validation are distinct processes that do not
occur simultaneously.
3.Engaging knowledgeable users is part of the face validity
technique used in validation.
Chapter 11 | 11. Estimation of Absolute
Performance| Quiz and Test
1.Output analysis in simulation aims to predict the
performance of a system or compare different
designs.
2.Autocorrelation refers to the independence of observations
in a simulation study.
3.Initialization bias has no significant impact on long-term
performance estimation.
Chapter 12 | 12. Estimation of Relative
Performance| Quiz and Test
1.Independent sampling assumes that observations

Scan to Download
 from two different system designs are statistically
dependent.
2.The Common Random Numbers (CRN) technique helps to
increase the variance of estimated performance measures in
simulation.
3.The Bonferroni approach is used to maintain an overall
confidence level when making multiple comparisons
among system designs.

Scan to Download
Chapter 13 | 13. Simulation of Manufacturing and
Material-Handling Systems| Quiz and Test
1.Simulation allows for a detailed evaluation of new
facilities and equipment investments without
disrupting current operations.
2.The primary goal of simulation in manufacturing systems
is solely to generate quantitative data.
3.Utilizing historical data in trace-driven models leads to
more accurate simulations compared to using statistical
distributions.

Scan to Download