1|Page

Session 2018-2019

1.

a. Define the following terminologies: i) System ii) Model iii) Simulation iv) Analytical
solution. Demonstrate different ways in which a system might be studied.
➢ System: A system is defined to be a collection of entities, e.g., people or machines, that
act and interact together toward the accomplishment of some logical end.

ad
Two types of system:
✓ Discrete system: The system in which the state changes abruptly at discrete points in
time are called discrete systems.
Example: A bank is an example of a discrete system.
✓ Continuous system: System in which the state of the system changes continuously with

m
time are called continuous system.
Example: An airplane moving through the air is an example of a continuous system.
➢ Model: A model is a simplified representation of a real-world system or process used to
understand or predict how it behaves.

m
Example: A toy car can be a model of a real car to show how it moves.
✓ Physical Model: A physical model is a tangible, small-scale version of something larger.
Like a model airplane that looks and behaves like a real airplane.
✓ Mathematical Model: A mathematical model uses equations and formulas to represent
ha
how a system works. A simple equation like d=r×t can be used to represent the
relationship between distance, rate (or speed), and time.
➢ Simulation: Simulation is the imitation of reality. Example: Flight simulator.
➢ Analytical solution: An analytical solution is a precise, exact answer to a problem derived
A
using mathematical methods.
Example: Solving the equation x+2=5 to get x=3.
b. What are the steps of simulation for a single-server queueing system? Describe the
problem statement and intuitive explanation of a single-server queueing system.
1. Problem statement:
an
as
H
2|Page

2. Intuitive explanation. Try to include at least first two process and last process in exam.

ad
m
m
ha
3. Program organization and logic. (First two step for second part question
answer)
A
an
as

4. C Program: C program for simulating an M/M/1 queue, emphasizing the use of

H

clear variable names and function prototyping. It also notes that floating-point
inaccuracies in different compilers can affect simulation results, especially when
events are close in time.
3|Page

5. Simulation Output and Discussion: The results of an M/M/1 queue simulation,

highlighting that the output numbers, such as average delay, number in the queue,
and server utilization, are influenced by random number generation and input
parameters. It discusses the challenges of choosing a stopping rule for accurate
steady-state estimation and suggests focusing on the delay in the queue rather
than the total waiting time for efficiency in estimating system performance.

ad
6. Alternative stopping rules: The text describes modifying a queueing simulation to
stop after a fixed amount of time, such as 8 hours, by introducing a dummy "end-
simulation" event. This event, scheduled to occur at the specified time, ensures
that the simulation stops when the event’s time is reached. The changes involve
adjusting the program’s external definitions, main function, and initialization and

m
reporting functions to handle this fixed run length and manage the end of the
simulation properly.
7. Determining the Events and Variables: The text discusses the event-graph method for

m
modeling complex systems, where events are represented as nodes connected by arrows
that show how they affect each other. It explains that this technique helps simplify the
model by eliminating unnecessary events and ensuring correct initialization. The method
is useful for managing and visualizing event sequences in simulations, especially in complex

2.
ha
systems, and can also help detect errors and predict computational demands

a. What is random variable? Compute variance and covariance based on a set of random
variates. Why stochastic process is used to simulate the output data? Illustrate correlation
A
function of the process of delays in queue D1, D2, ...... for the M/M/1 queue.
an

Variance:
as
H
4|Page

ad
m
m
ha
A
an

A stochastic process is used to simulate situations where outcomes are uncertain or

random. It helps predict possible results when we can't know for sure what will
happen, but we can estimate based on probabilities.

Example: Imagine you're trying to guess how many customers visit a shop each day.
as

Some days there are 20, other days 50, and it changes randomly. A stochastic process
helps simulate these changes, giving you a range of possible outcomes instead of just
one fixed number.
H
H 5|Page

as
an
A
ha
m
m
ad
6|Page

b. Explain hypothesis testing including null and alternative hypothesis for mean. What is the
difference type I error and type II error? Explain and illustrate the strong law of large
number.
When one performs a hypothesis test, two types of errors can be made. If one rejects H0 when
in fact it is true, this is called a Type I error. The probability of Type I error is equal to the
level a and is thus under the experimenter’s control. If one fails to reject H 0 when it is false,
this is called a Type II error. For a fixed level a and sample size n, the probability of a Type

ad
II error, which we denote by b, depends on what is actually true (other than H 0: µ = µ0), and
is usually unknown. We call δ=1-β the power of the test, and it is equal to the probability of
rejecting H0 when it is false. There are four different situations that can occur when one tests
the null hypothesis H0 against the alternative hypothesis H1, and these are delineated in Table
4.2 along with their probabilities of occurrence.

m
m
ha
A
The key differences between Type I and Type II errors:

Aspect Type I Error (False Positive) Type II Error (False Negative)

Failing to reject a false null
Definition Rejecting a true null hypothesis
hypothesis
an

Missing an effect that actually

Interpretation Detecting an effect that doesn't exist
exists
Common Concluding a treatment works when it Concluding a treatment doesn't
Example doesn't work when it does
as

Probability
α (alpha) β (beta)
Notation
Consequence Incorrectly claiming significance Missing a true significant finding
Relation to Lower α reduces chance of Type I error, Higher power (1 - β) reduces the
Power but increases risk of Type II error chance of Type II error
H
H 7|Page

as
an
A
ha
m
m
ad