Introduction to Computer Performance Modeling

ECE/CSC/OR 579, Spring 2018

https://moodle-courses1718.wolfware.ncsu.edu/course/view.php?id=6398

The aim of this course is to present queueing theory and simulation
techniques as tools for modeling and studying the performance of communication networks
and computer systems. The students will be introduced to classical tools and
methodology in probability theory and stochastic modeling as well as simulation
techniques, all of which are essential tools for students to conduct advanced
research in the area of network performance modeling and analysis.

Students will participate and learn by doing assignments before coming to
class, by asking and answering questions during in-class discussions, by
performing simulation projects, and by preparing for in-class exams.

At the end of this course, students should be able to

- Apply simulation techniques to develop models of
computer and communication systems
- Apply queueing-based models to characterize computer
and communication systems
- Use appropriate analytic tools to compute performance
measures of interest (e.g., delay, throughput) for a given queueing system
- Design (or choose) the system parameters (e.g., server
or link capacity) to achieve a given level of performance
- Evaluate the relative merits of alternative system
design solutions
- Engage in research in the field of performance analysis
and evaluation via Markov chains for general networked systems

Mon &Wed., 11:45AM--1:00PM, 1010 EB1

Do Young Eun,
Professor

Office: 3064 EB2

Phone: 919-513-7406

Email: dyeun@ncsu.edu

**Office
hours:** Mon. 2--3PM & Wed. 10:30AM--11:30AM (or
by appointment) in 3064 EB2

- MA 421 Probability Theory or equivalent
- C, C++, or other programming language

**Queueing Systems, Volume 1**: Theory, by Leonard Kleinrock, Wiley-Interscience, 1975, ISBN: 0471491101

- Probability, Markov
Chains, Queues, and Simulation: The Mathematical Basis of Performance
Modeling, by William J. Stewart,
Princeton University Press, 2009, ISBN: 0691140626

**Performance Modeling and Design of Computer Systems: Queueing Thoery in Action**, by Mor Harchol-Balter, Cambridge University Press, 2013,**ISBN-13:**978-1107027503 (e-book available on NCSU) (author website)**Markov chains**, by J. P. Norris, Cambridge, 1998 ISBN-10: 0521633966 (author website)**Performance Analysis of Communications Networks and Systems,**by Piet Van Mieghem, Cambridge University Press, 2006, ISBN-10: 0521855152 (author website)

Note:

- References are *
**not*** required to buy, but you may want to have some of them on your bookshelf. They are on reserve in the Hunt library for this course. - Lecture will be mostly based on class notes, to
be posted prior to each class

There will be one midterm
exam, one final exam, projects (simulation), and homework assignments.

Homework: 20%

Simulation Projects: 15%

Midterm exam: 25%

Final exam: 40%

**Note:** All exams will be open books and open notes.

Homework grading

- Homework will be due in class at the
**beginning**of the lecture. - Late homework assignments will not receive any credit.
- Each question in each homework assignment will have an equal weight, unless otherwise specified

- Only hardcopy (handwritten or typed) will be accepted in class. No online submission.

**Audit students must earn a B average on the homeworks. **

- There will be no make-up exams
- New assignments and deadlines will be announced in class. Hard copies of handouts, assignments etc. will usually not be distributed. Updates and copies of assignments will be available only from the web and/or e-mail. It is your responsibility to check whether anything new has been issued if you miss a class.
- Objections to grading of assignments or exams must be filed in writing within one week after they have been returned to you.
- No cheating allowed. Any form of cheating will result in an immediate failure of the course, and may be reported to school for further action.
- During the lecture please turn off any cellular phones, laptops, tablets, etc.

- Review of probability theory and random variables
- Review of z-transforms and Laplace transforms
- Poisson processes
- Markov Chains (both continuous and discrete time)
- Birth-Death Processes
- M/M/1 queue and variants
- M/Er/1, Er/M/1, and Erlang distribution
- M/G/1 queue, P-K formular
- Priority queueing
- Finite State Markov Chains
- Other aspects of Performance analysis on networked system and applications
(random walk on graph, dynamics on network, Markov Chain Monte Carlo, etc.), if time permits

Reasonable accommodations will be made for students with verifiable disabilities. In order to take advantage of available accommodations, students must register with Disability Services for Students. See https://dso.dasa.ncsu.edu/ for more information. more information on NC State's policy on working with students with disabilities, please see http://policies.ncsu.edu/regulation/reg-02-20-01

All the provisions of the NC State University's Code of Student Conduct and University Policy on Academic Integrity apply to this course. In addition, it is my understanding and expectation that your signature on any test or assignment means that you neither gave nor received unauthorized aid.