Project for Computer Networks

There are several project options for the Fall 2003 semester of Computer Networks. The deliverables for all projects are a paper in standard IEEE conference paper format, source code, and a demonstration. The purpose of the project is for a student to work with sockets programming and explore an interesting idea or research direction (i.e., to not implement some already existent and well-know application or some "useless" application). I believe all of the projects offered below have the potential for leading to further investigation beyond this semester. Please read this page very carefully to understand all the parameters of the class project.

Rules:

1. All source code and written material must be your own. Certainly, you may and should look at source code (e.g., open source Linux), but what you submit must be your work entirely.
2. Can you work together? Yes. But, see rule #1.
3. All submitted source code must comply with Christensen C style guidelines. Code that does comply to these guidelines will be discarded. All programs must be implemented in straight C (yes, some of the projects could probably be better done in Perl) and must compile with the Borland bcc32 command-line compiler. All programs must be console-mode (no GUI). All programs must be portable between Windows and Unix unless there is some good reason (which you should discuss with me) that this cannot be accomplished.
4. The submitted paper must comply with IEEE conference paper formatting rules. Author instructions from the IEEE are here (PDF). A paper I wrote for ICC 2004 is here (ZIP). You may use my paper as your template if you are using Word as your word processor. You have a five page limit. Any submitted paper that does not comply with these formatting rules will be discarded.
5. I am very eager to help you with your projects. However, I firmly believe that last minute efforts are counter-productive for both you and me. Thus, after the "final project review date" stated in the syllabus I will not help you anymore. We are covering sockets programming early in the semester so there is no reason or excuse for back-ending your project.
6. The project submission date is stated in the syllabus. Late submissions will not be accepted.

Project ideas:

Project Idea #1:On average any two people in the world are separated by six people. That is, from me to any person in the world I need only go through six people that know each other. Google "six degrees of separation" to learn more (or see here. Your challenge is to determine how many degrees of link separation there are between random personal homepages on the web. You will need to write a program to roughly do the following:
1. grab a page,
2. parse out the links (store them in a list),
3. follow the links
4. repeat at step (2) until have "enough links"
You are to complete a reasonably comprehensive study of the degrees of the separation question (i.e., your program is the tool for the study). This is a two-person project. Most important is that you THINK.

Project Idea #2:One of my graduate students, Sai Sakamuri is studying ways of solving the spam problem. We have an idea called email++ where all email messages are validated via an out-of-band TCP connection between the receiving and sending user agent. For more information see:

• Email++
A two-person project is to implement an email++ POP3 user agent. A three-person project is to implement an email++ POP3 user agent that also implements a puzzle solving feature where the sender user agent "pays" for received email in CPU cycles. A four-person project is to implement an SMTP server that implements email++ as a proxy for a user agent. This will likely entail implementing both a user agent and a minimum-function SMTP server.

Project Idea #3:One of my graduate students, Graciela Perera is investigating distributed search techniques in peer-to-peer networks. One idea being investigated is to implement a hybrid search procedure that considers flooding as well as direct access based on download history. Nodes maintain statistics on what other nodes they download files from. When a node is searching for a file it first directly checks nodes from which is has had good success (in finding files) in previous searches. If this fails to find the desired file, first then is the request flooded. This scheme trades-off response time for a reduction in network traffic. For further information, see:

• P2P search with reduced flooding
A two-person project is to develop is a Gnutella servent that maintains statistics on what nodes it finds files. A three-person project is to build a servent that maintains statistics (as above) and uses these statistics to attempt a direct search before resorting to flooding. For both projects, an experimental evaluation of the servent must be completed. For the two-person case, the evaluation is a characterization of file-node relationships For the three-person case, the evaluation is a measurement of how many searches can be satisfied with a direct (as opposed to flooding) search.

Project Idea #4: There is no project idea #4, but you are very welcome to propose your own project. A good project will emphasize sockets programming and implement some new idea or concept, or be a tool for an interesting study. Clearly, "a Visual Basic front-end on a database" is not within the scope or interest of this class! Nothing in Java is of interest unless you can make a case why Java, and not C, is needed to implement your idea.