Last updated 12/05/2001

CPSC 633a : Seminar on Computer Networks, Fall 2001

Announcements

• [12/05] I will be available tomorrow from 2 to 4pm. Then I will be out of town from Dec. 7 to Dec. 16. I will be reachable through email during this time. You can also talk to me, and my phone number will be 512-232-7490. I will be on campus from Dec. 17 to Dec. 19, and you can setup appointments to talk to me. Good luck in your finals!
• [11/26] The class home page on incentive-compatible computing is here.
• [11/12] I did not address Stan's question clearly in the class. I hope it will not cause more confusion. Before I re-explain the idea in next class, you can check  the slides for update. The link to the paper is here. Also, next class I will discuss a lower bound for secure group rekeying. You are not required to read the paper on lower bound. If you are interested in it, however,  please refer to the schedule page for a link to the paper.
• [10/26] I have changed the paper for this coming Monday from [FGS01] to [LA00]. If you have already read [FGS01], you can submit its review instead. However, it is strongly recommended that you read [LA00].
• [10/23] Please prepare for a brief  talk on your project. Keep the time to about 8 minutes and leave 4 minutes for questions. If you need to use powerpoint, I will bring my laptop to the class. In the talk, please try to address the following questions: What is the problem? Why is the problem interesting? Why is the problem difficult? What is the previous work? What is the idea to solve the problem? What is the methodology to justify your solution? What is the expected deliverable?
• [10/21]  I have added a few more potential topics to the project list. Please be prepared to say a few words about your project. If you are still looking for project partners, it is a good opportunity to say what you are interested in the class.
• [10/20] When reading the paper for Monday's class, try to think the relationship between bandwidth adaptation and playout buffer. Also, you can check the slides for Monday's class, and think about some of the questions on the slides.
• [10/11] We will not have class next Monday. Also, I have linked several tutorials on the reading list (I always find reading tutorials is easier than reading papers).
• [10/09] Try to think of the following problems before Wednesday's class:  How do we regulate people's behavior in a society? How do we allocate resource in an economy? Can we borrow ideas from these fields to provide incentives for users  to regulate their sending rates?
• [10/08] The main TCP security issues we will discuss this Wednesday are related with congestion control. For those who are interested in general security, here is Joncheray's paper about a simple active attack against TCP which takes over a connection after the user authentication has been done (e.g. Kerberos). As an example, here is the source code. You are not encouraged to try it out, but are encouraged to read the code. If you are really curious, you can read a sample log. Also, the paper mentions about nmap, which uses "fingerprint" to detect the os of a machine, here is the link. Since we did not cover the state transition diagram of TCP (which will be covered next semester), you may have to read Steven's textbook in the reading list.
• [10/02] Here are a couple questions for you to  think about for tomorrow's class: 1) The authors considered three classes of schemes: link layer, end to end, and split connection. What is the best scheme that was considered for each class in this paper? What is the best scheme you can think of for each class? 2) From the paper, the authors showed that their best scheme in the link layer class is better than their best scheme in the split connection class. Why?
• [10/01] If you are interested in wireless network, the tutorial, wireless congestion control, is a comprehensive survey on wireless congestion control.
• [09/30] Please take a look at the list of project topics. You are welcome to come up with your own topics. Talk to me before you make your decision.
• [09/26] In Monday's class, we will try to debate on TCP/Vegas. Try to think how and why TCP/Vegas improves over TCP/Reno. What are the advantages? What are the disadvantages? Will it work if we introduce TCP/Vegas into the current Internet?
• [09/13] For those who are interested in wireless, you can find a nice tutorial: routing, MAC, and mobility in Ad Hoc network. For the class next Monday, you can read the part on wireless MAC, which starts from page 254 and ends at page 305.
• [09/12] If you are interested in the possibility of infinite bandwidth and its effects, you can read the interesting book [Gil00].
• [09/12] For the problem of CIDR on IP table lookup, please see the papers in the high performance router section, e.g.,  [SSV99].
• [09/10] Starting from this Wednesday (9/12), we will move our class from AKW 200 to AKW 500.
• [09/10] Instead of I listing the names of the students, everyone attended class today can enroll. 
• [09/10] Please make sure that your reviews are world readable.
• [09/09] The instruction on how to submit your reviews is here. Note that I have also revised the review template.
• [09/09] For those who are wondering between CPSC 633a and CPSC 433b, please take a look at the topics of CPSC 433b. Here are my suggestions on which class you may want to take: CPSC 633a, or CPSC 433b. You can take both. If still in doubt, please talk to me.
• [09/05] Please take the class survey.