EE 538 Communication Network Analysis

(Visit for announcements throughout the semester)

Description: This is a first-year graduate course on the performance analysis and design of communication networks. The emphasis is on analytical techniques and on systems issues that arise in the control of these networks. Requirements are EE 533 (or an equivalent course on stochastic processes) or consent of instructor. CS 421 is recommended but not required.

Time and place: Monday 8:40 9:40, Wednesday 15:40 16:40, Room EA 222 Engineering Faculty (valid as of Feb. 11).

Monday, Feb. 18: There will a one-hour lecture starting at 9:40.

Final exam questions can be picked up from Room EA-122 on Wednesday, May 8. Answers should be handed in by 9:00 on Monday, May 13.

Office Hours: Monday 10:30--11:30, Wednesday 17:30--18:30, Room EA 122.

Topics will be selected from:

1. Review of Markov and renewal processes
2. Basic queueing theory; M/M/1, M/G/1, G/M/1, G/G/1 queues, Little's Law, PASTA property, Kingman's bound
3. Multiple access techniques; Aloha, Ethernet, tree algorithms
4. Reversibility and loss models, Erlang fixed point approximation
5. Networks of queues; Burke's theorem, open and closed networks
6. Flow control; TCP
7. Broadband networks; Leaky-bucket regulators, Cruz's network calculus, effective bandwidths, admission control and pricing
8. Markov decision processes and dynamic programming; routing, admission control
9. Macroscopic network behavior; fluid and diffusion limits

References:

• D. Bertsekas and R.G. Gallager, Data Networks, 2nd Ed., Prentice-Hall, 1992 (Recommended) (2,3,4,5)
• L. Kleinrock, Queueing Systems Volume 1, Wiley, 1975 (1,2)
• F.P. Kelly, Reversibility and Stochastic Networks, Wiley, 1979 (available online) (4,5)
• M. Schwartz, Broadband Integrated Networks, Prentice-Hall, 1996 (6,7)
• P. Varaiya and J. Walrand, High Performance Communication Networks, Morgan Kauffman, 1996 (7)
• J. Walrand, An Introduction to Queueing Networks, Prentice-Hall, 1988 (2,5,8)
• D. Bertsekas, Dynamic Programming and Optimal Control, Athena Scientific, 1995 (8)
• R.G. Gallager, Discrete Stochastic Processes, Kluwer, 1996 (1,2)
• W. Feller, An Introduction to Probability Theory and its Applications, Volume 1, Wiley, 1968 (1)
• K.W. Ross, Multiservice Loss Models for Broadband Telecommunication Networks, Springer, 1995 (not available in the Library) (4)

Grading:15% homework, 20% midterm I, 20% midterm II, 45% final.

Course policies: There will be roughly 6 homeworks. Homework due dates are nonnegotiable. You can collaborate with fellow students in solving each homework question, provided that 1) you write your own answer, and 2) you indicate the name(s) of your collaborator(s) on your solution. You are allowed to bring one crib sheet of notes to the first exam, two sheets to the second exam, and three sheets to the final. Exams are otherwise closed book.