Computer Simulation Approach

By Dereje Shiferaw
 Contents
• Models experiments and computers
• System, system entity and system state
• Models and model classification
• Simulation as experimentation
• Why simulation?
• When is simulation an appropriate tool
• Field of applications
• Key phases in computer simulation
• Discrete and continuous simulations
• Computing: data driven software and Bespoke program
• Experimentation: Interactive experiment and classical
experiment
 Introduction
Ways To Study A System
 Introduction contd…
• Simulation
Definition: is a numerical technique for conducting experiments on a
digital computer, which involves certain types of mathematical and logical
models over extended period of real time.

• System simulation
Definition: The technique of solving problems by the observation of the
performance ,over time, of the dynamic model of the system. It is an
experiment of physical scenario on computer.
• Consider a situation where an expert has to decide how many and what
type of network or support resources to have available
– Model the system and study its performance before actually setting
the system
– Analyze system
– Make operation or resource policy decisions
 Models, experiments and computers
• Model
– Meaning
• representation of construction and working of system, it is similar
to but simpler than the system it represents
– Purpose
• to enable the analyst to predict the effect of changes to the system
– Good model
• close approximation to the real system
• incorporate most of its salient features
• Not so complex
• judicious tradeoff between realism and simplicity
• Experiment
– Running a given model of a system.
– time taken for the actual system and the model may not be the same
• Computer
– All models are either mathematical or software and will be executed
on a computer. Hence the computer is the means for conducting
experiments on the model
 System, system entity and system state
• System
– A system is defined as a collection of interacting components that
receives input and provides output for some purpose
• System entity
– Is something (customer, objects etc) that changes the state of a system
– It is one basic component of a simulation model
• System state
– is a set of data that captures the salient variables of the system
– allows us to describe system evolution over time
– is stored in one or more program variables that represent various data
structures (e.g., the number of customers in a queue, or their exact
sequence in the queue).
– State definition depends on the modeling needs, particularly statistics
to be collected
 Models and model classification
• A model intended for simulation study can be
– A physical model – simplified or scaled down physical object
• Example: Scaled model of an airplane
– Mathematical or analytical model – a set of equations or relations
among mathematical variables. The mathematical model can be
deterministic or stochastic, static or dynamic
• Example: Mathematical model of a factory workflow
– Computer model – a program description of the system . A computer
model with random elements and an underlying time element is
known as Monte Carlo simulation model
• Example : The operation of a manufacturing unit over a period of
time
 Simulation as experimentation
A simulation of a system is
• the operation of a model of the system.
• tool to evaluate the performance of a system, existing or
proposed, under different configurations of interest and over
long periods of real time.
• The model can be reconfigured and experimented with
– Used to test condition which is impossible, too expensive or
impractical to do in the system it represents.
• Example: desert battle simulation
- one force invading another,
- 10 years ago
- simulation involved the modeling of 66,239 tanks, trucks
and other vehicles on simulated terrain around Kuwait, using
multiple supercomputers in the DoD High Performance
Computer Modernization Program
• Simulation experiment is a test or series of tests in which
meaningful changes are made to the input variables
• From simulation experiment
– The operation of the model can be studied,
– Properties concerning the behavior of the actual system or its
subsystem can be inferred.
• Examples:
- 1-billion-atom model of material deformation
- a 2.64-million-atom model of the complex maker of
protein in all organisms, a ribosome
- the Blue Brain project at EPFL (Switzerland) 2005
Technical Attractions of Simulation
• Ability to compress time, expand time
– Simulation time: representation of physical time
– Physical Time: time in physical system
• Ability to control sources of variation
• Avoids errors in measurement
• Ability to stop and review
• Ability to restore system state
• Facilitates replication
• Modeler can control level of detail
Why is simulation?
•Answer questions like:

What is the best design for a new telecommunications network?

What are the associated resource requirements?

How will a telecommunication network perform when the traffic

load increases by 50%?

How will a new routing algorithm affect its performance?

Which network protocol optimizes network performance?

What will be the impact of a link failure?

• Gaining insight into the system operation

• Developing operation or resource policy to improve system

performance
– Deciding the type and size of network and resource to be used

• Testing new concepts and/or systems before implementation

– Testing an appropriate Time to leave for a TCP/IP packet in a wireless sensor
network for energy optimization
– Testing the effectiveness of a DdoS prevention mechanism

• Gaining information without disturbing the system

– Analysis of security of a given network or Internet
 When is simulation appropriate tool?

• A simulation model is implemented in a computer

program. Compared to analytical modelling, it is
relatively in expensive. It is preferred to analytical
modelling when:
• When analytical model with tractable solution is
unknown
• When underlying model is complex
• When driving analytical model is difficult or finding
analytical solution is difficult or time consuming or
when it is difficult to capture necessary details using
mathematical model
 Field of applications
– Production system, inventory systems, Manufacturing
process. Material handling and logistics
• Estimate a set of productivity measures
– Computer systems and communication networks
• Design and planning of capacity so as to minimize
response time
– Military
– Business process reengineering activities
– Health care operations, financial and banking operation,
transportation and air ports
• Improving performance
 Key phases in computer simulation

1. Model development
2. Experiment design
3. Output analysis
4. Conclusion formulation
5. Making decision to alter system under study
1. Simulation Model development

• Modeling is the most important part of a simulation

study. It is determining factor
• A simulation model consists of
• System entities
• Input variables
• Performance measures
• Functional relationships
Example: Simulation of a single server system

• For the above system

– Server and queue are system entities
– Arrival rate and service rate are input variables
– Mean wait time and maximum queue length are performance
measures
– Time in system is functional relationship

time _ in _ system  wait _ time  service _ time

Steps in model building
• Identify the problem
– List problems of existing system
– Produce requirements for proposed system
• Formulate the problem
– Identify bounds, end user
– Define objectives and performance measure
• Collect and process real system data
– Collect data on system spec, input variables and
performance of existing system
– Identify sources of randomness and select appropriate
probability distribution for each random input (exp,
poison, normal etc)
 Steps cont…
• Formulate and develop a model
– Develop schematic and network diagrams,
translate to programs
• Validate and verify the model
– Verify – check if model is correct to specifications
– Validate- check if model output agree with real
system
• Document model for future use
 2. Experiment design
• A simulation model is a test or series of tests in which meaningful
changes are made to the inputs so that we may observe and
identify the reasons for changes in performance measures
• What data need to be obtained, in what form and how much
• Steps
– A) select appropriate experimental design
• Select performance measure, key inputs and their range
– B) Establish experimental conditions for runs
• Starting conditions, length of run, number of independent runs,
– C) perform simulation runs
Example: Simulation study of DDoS attack detection algorithm
- model traffic and network
- design experiment
 3. Output analysis

• Interpret and present results

– Identify best design among competing alternatives
– This can be using statistical inference
• Recommend further course of actions
– Use the results and analysis to write
recommendations
– A formal written report
 Discrete and continuous simulations
• Discrete simulation
– state trajectory is piecewise constant function whose jumps
are triggered by discrete events.
– Simulation state remains unchanged unless a simulation even
occurs
– When an event occur, the model undergoes a state transition
– The model evolution is governed by a clock and a
chronologically ordered event list.
• Continuous simulation
– Simulation where the system state is continuous over time
– Is used to model fluid flows, aircraft dynamics etc
– Based on mathematical model of system
– Model is digitized and simulated using program
 Computing: data driven software and
Bespoke program
• Data driven software
– Data is the major aspect of system
– Design based on end results and inputs rather
than objects which are behaviors
• Bespoke program
– Custom or tailor made programs used for specific
application
– Easy to use, has fewer errors, increase productivity
and unique out of competitors
Experimentation: Interactive experiment and
classical experiment
• Simulation experiments are interactive
experiments
– Experiments can be repeated any number of times
with various inputs and parameters
• Classical experiments are hard ware based or
use physical objects
– are costly and take longer time