CSC7290: Advanced Computer Networking (Winter 2009)

QUICK LINKS: Lectures | Exercises | Projects | | Related resources

Course mailing list: general, web section only

Instructor: Dr. Hongwei Zhang
hzhang AT cs.wayne.edu
+1 313 577 0731
Class timings: MW 3:00pm-4:20pm
Class webpage: http://www.cs.wayne.edu/~hzhang/courses/7290/7290.html
Office hours: MW 4:30pm-5:30pm in 454 State Hall, or by appointment

Teaching Assistant: TBA
TA Office hours:

TBA

Overview (flyer)

This course is designed for students who are interested in the principles and techniques of network protocol and system design. Topics span three broad areas: 1) Performance evaluation of networked systems: techniques and metrics, experiment design, data analysis, and statistical modeling; 2) Modeling, analysis, and design of network protocols: formal specification and analysis of network protocol properties, design of scalable and fault-tolerant network protocols; 3) Stochastic analysis of networked systems: stochastic process, queuing theory, and their applications to network modeling and analysis.

In short, the objective of this course is to help students understand the foundational principles and techniques of network design and analysis, to help students appreciate why networks have been designed as they are today, and to build up students' capability in enhancing the state of the art in computer networking.

Prerequisites

Basic knowledge of computer networks (e.g., materials covered in CSC6290 or equivalent), elementary probability theory, statistics, and mathematical logic. Or consent of instructor.

References (Tentative, may change before semester starts)

Required:

[R0] Raj Jain, The Art of Computer Systems Performance Analysis: Techniques for Experimental Design, Measurement, Simulation, and Modeling, John Wiley & Sons, Inc., 1991. (ISBN: 0471503363)

Strongly recommended:

[R1] Dimitri Bertsekas and Robert Gallager, Data Networks (2nd edition), Prentice Hall, 1992. (ISBN: 0132009161) (From 1st edition: Queueing chapter, Routing chapter; courtesy of Prentice Hall.)

Recommended:

[R2] Mohamed G. Gouda, Elements of Network Protocol Design (1st edition), John Wiley & Sons. (ISBN: 0471197440)
[R3] Sheldon M. Ross, Introduction to Probability Models, 9th edition, Academic Press, 2006. (ISBN: 9780125980623)
[R4] Robert G. Gallager, Discrete Stochastic Processes, Kluwer Academic Publishers, 1996. (ISBN: 0792395832)

[R5] Thomas G. Robertazzi, Computer Networks and Systems: Queueing Theory and Performance Evaluation (3rd edition), Springer. (ISBN: 0387950370)
[R6] Anurag Kumar, D. Manjunath, Joy Kuri, Communication Networking: An Analytical Approach, Morgan Kaufmann, 2004. (ISBN: 0124287514)
[R7] Ravindra K. Ahuja, Thomas L. Magnanti, James B. Orlin, Network Flows: Theory, Algorithms, and Applications, Prentice Hall, 1993.

Flowchart of topics

Prelude: how to have a bad career in research; general principles on systems research; review of basic probability theory and statistics
Performance evaluation techniques and their applications to computer networks

Common mistakes and how to avoid them
Selection of techniques and metrics
Experiment design and analysis
Summarizing measured data
Comparing systems using sample data
Mathematical regression models

Formal models of distributed algorithms and their application to network protocol analysis

Network process: syntax and semantics
Models and analysis of network processes: specification and proof techniques
Applications to network protocol design and analysis

Stochastic models of computer networks and their applications

Stochastic processes and queuing theory
Applications to protocol and system analysis

Lectures

Part 0(a): Introduction

Webcasting:

Part 0(b): Computer Networking: Art and Science

Webcasting:

Part 1 (a): [Performance Evaluation] Common Mistakes and How to Avoid Them (Chapter 2 of R0)

Webcasting:

Part 1 (b): [Performance Evaluation] Selection of Techniques and Metrics (Chapter 3 of R0)

Webcasting:

Part 2(a): Review of Proability Theory (Chapter 1 of R4)

Webcasting:

Part 2(b): Workload Characterization Techniques (Chapter 6 of R0)

Webcasting:

Part 3(a): [Experiment design & analysis] Introduction to Experimental Design (Chapter 16 of R0)

Webcasting:

Part 3(b): [Experiment design & analysis] 2^k Factorial Designs (Chapter 17 of R0)

Webcasting: 0

Part 3(c): [Experiment design & analysis] One-factor Experiments (Chapter 20 of R0)

Webcasting:

Part 3(d): [Experiment design & analysis] General Full Factorial Design with k Factors (Chapter 23 of R0)

Webcasting:

Part 4: Introduction to Simulation (Chapter 24 of R0)

Webcasting:

Part 5(a): [Data analysis] Summarizing Measured Data (Chapter 12 of R0)

Webcasting:

Part 5(b): [Data analysis] Comparing Systems Using Sample Data (Chapter 13 of R0)

Webcasting:

Part 5(c): [Data analysis] Simple Linear Regression Models (Chapter 14 of R0)

Webcasting:

Part 5(d): [Data analysis] Other (non-SL) Regression Models (Chapter 15 of R0)

Webcasting:

Part 6(a): [Queueing analysis] Introduction to Queuing Analysis (Sections 3.2 and 3.3.1 of R1)

Webcasting:

Part 6(b): [Queueing analysis] Review of Markov Chain Theory (Appendix A to Chapter 3 of R1)

Webcasting:

Part 6(c): [Queueing analysis] M/M/* Queues (Sections 3.3 and 3.4 of R1)

Webcasting:

Time Reversibility and Burke's Theorem
Open Jackson Networks
Closed Jackson Networks
M/G/1 Systems

Exercises

TBA

Projects

Project options

Evaluate the performance and/or analyze the properties of selected protocols/systems in the following fields:

Vehicular sensor networks

intra-vehicle sensing and control
inter-vehicle sensing and control
urban/participatory/opportunistic sensing via vehicles

Sensor networks in

Healthcare: http://www.agingtech.org/browse.aspx?CA=1
Engineering: structural health monitoring, factory automation & industrial control, etc.
Scientific study: environmental engineering, social sciences, etc.
Homeland security and military, or
Daily life: urban sensing (http://research.cens.ucla.edu/portal/page?_pageid=56,948798&_dad=portal&_schema=PORTAL), security monitoring, etc.

Can focus on issues such as MAC, routing, transport control, data storage and querying, and localization.

Mobile networks in

traffic control: real-time road traffic condition detection and control, etc.
auto safety: DSRC, etc.
homeland security, or
social networks

Heterogeneous networks

integrated wireless networks (sensor networks, WiFi, cellular) and the Internet etc.

Networking technologies for emerging economies

network properties: wireless, mobility, intermittent connectivity ...
network services: telemedicine, mobile banking, e-retailing, stored data and voice messaging, remote education, local content and news, security, policing, etc.

Other topics of your choice (with conscent of instructor)

Rules:

Each student is expected to select his project and discuss with the instructor to finalize the scope of the project.
Students are allowed to form groups in doing projects, but the number of students per group should be no more than 3.

Deliverables:

In-class presentation. 1) The slides for your presentation should be sent via email to the class at least one day before your presentation, so that everyone can go over your slides before coming to class. 2) Your presentation should include overview of related work, the problem definition, methodology, and analysis.
Written project report. A sample of network performance evaluation article/report is here.

Timeline:

Select the topic and form your project group by 01/31/2010.
Submit your detail project plan and timeline by 02/28/2010.
Present your project in class according to this schedule.
Submit your project report electronically by midnight 05/01/2010.

Evaluation criteria:

Your performance in project will be evaluated based on the following metrics:

Bredth and depth of your project, as evidenced by your project report and presentation.
Presentation quality (e.g., clarity, readability, and conciseness) of your project report and in-class talk.
Whether or not you are able to stick to the project timeline.

Miscl.:

Related resources

TinyOS

TinyOS Community Forum, http://www.tinyos.net/
Phil Levis's book on TinyOS Programming
TinyOS documentation
Resources on Using TinyOS and motes

Network simulators:

TOSSIM
ns-2
qualnet/glomosim
opnet

Links to journal & conferences related to computer networking, by Hongwei Zhang

Computer System Performance Evaluation, by Dr. John S. C. Lui.
Network Design and Performance Analysis, by Dr. Santosh Kumar.
Computer Systems Analysis, by Dr. Raj Jain.

Policies

Lecture: Attendance at lectures is required. If a student has to skip a lecture due to hard constraints, he/she is required to inform the instructor beforehand.

Homework: Homework assignments will be designed to stimulate independent thinking among the students. They will be due at the beginning of class, usually a week after they are given. Homework assignments will not be accepted after the due date. An exception to this rule is that you give in advance a strong and convincing reason.

Exam: Exams will be scheduled in advance. Unless prior arrangements are made, a grade of zero will be recorded for missed exams.

Grading: The tentative grade weighting for the semester will be:

Class participation: 10%
Quizzes: 45%
Project: 45%

* These weights are subject to minor adjustments.
* Letter grades will be assigned based on performance relative to other students. A tentative grading scale is as follows:

A: 93-100
A-: 90-92
B+: 85-89
B:   80-84
B-: 75-79
C+: 70-74
C:   65-69
C-: 60-64
F:    0-60

* A regrading request will cause the entire exam/homework/project to be regraded, and thereby the overall grade can increase or decrease.

Miscellaneous

I expect you to carefully read all material handed out in class. I also expect you to read the book according to the reading assignments announced in class. You are encouraged to discuss the material presented in class with other students, but definitely do not collaborate with anyone in solving the homework problems. The Wayne State University Student Code of Conduct applies. Feel free to discuss our expectations and grading criteria with the grader or me during the semester.