Welcome to the student guidelines page for the networking research group at RPI. We want to build a top-class, high-impact networking research group. This means that students who form the group should be top-class researchers. Research has to be creative by definition and your research productivity, creativity and intellectual development is my prime concern. I look forward to the shared learning experience with you. My responsibility is to take you from where you are to where you have never been before from an intellectual, creative and career viewpoint.

However, since we are a growing group, administrative overhead eats into this time and reduces our productivity. Therefore, I wrote this web page to clarify and formalize my expectations of student colleagues and your expectations of me, so that we can maximize the utility of the time we spend together in achieving our research goals.

The following are the goals of the group:

• To solve high impact problems centered around the theme of traffic management.

• To use an optimum mix of theoretical, analytical, simulation and experimentation tools to attack these problems.

• Understand and be able to articulate the "big picture" of the problem space they are concerned with

• Do whatever it takes to attack the problem - if necessary learn new analytical, experimental or simulation techniques to solve the problem. This also pre-supposes you be highly motivated and proactive in your approach to research.

Towards the ends mentioned earlier, my responsibility is to clear all hurdles in your path. The best way is to set up a simple process to care of administratrivia:

• Equipment/Experimental hurdles: we will create a large Linux testbed environment (with about 25-50 machines projected) for your prototyping and experimental needs. Besides this we are trying to provide one production workstation terminal and simulation servers for your general purpose use. We will also be buying simulation packages or development environments/tools if the concrete need arises.

• Books: If you need any book, please let me know immediately and I will order it for you asap. Or if I am not around, order online and I will reimburse you. Do make sure that we do not have it already handy.

• Courses: If you need to take any course to update your toolkit for research purposes, please feel free to enroll in it. I will pick up the bill if you are funded.

• Administrative bottlenecks : If you have any administrative bottleneck which is hampering your focus on research, please let me know immediately.

• Feedback: A fair amount of emotional and creative energy can be lost in misunderstandings or overloading. Please use the anonymous feedback form on my home page to send me feedback about any aspect of our research group activities or any of these processes which you might not want to talk to me directly about. Rest assured that your mail will be confidential/anonymous and I will do whatever it takes to address your problem.

• Ideas: We will have meetings twice a week for every group. The purpose of these meetings is to quickly review past work and spend time tweaking our understanding of the "big picture", and creatively brainstorming about new ideas/solutions to problems. Please come to meetings prepared. Also please take notes and document notes/ideas/action-items immediately on the internal project site (described below). This way we will not lose information. I know from experience that bad information management is the key reason why projects and dissertations take much longer than necessary.

  My benchmark for bad meetings is a string of 4 meetings without any creative output. But if you ever get bottlenecked for lack of ideas. Just get a hold of me, either by banging on my door, sending me email or phone. Feel free to call my home (518-482-1782).

• Knowledge Management: Each project group is responsible for setting up an internal web page which would give a snapshot of the work-in-progress. This should contain the following items. We hope to create a template for this soon.
  • A skeleton of a tech-report in progress. This skeleton should be updated as often as possible with bullets, distilled text, and section headings. In my experience, I have found this to be invaluable a) in writing a paper later, and b) not losing information and having to re-do work in cycles.

  • Bibliography: This should have on-line links to every paper we find on the topic, and a .bib file having bibliographic entries which we can leverage for writing tech-reports/papers. Useful bibliography links are:
    • Essential Congestion Control Literature

    • Henning Schulzrinne's Network Bibliography

    • Network Classic Papers Reading List

  • Outstanding action items with dates: Can immediately let us know what's next on our list todo. Also lets us not forget items of interest. Please keep this up-to-date.

  • Minutes of meetings or Ideas generated elsewhere: Detailed information. Meeting headings should be descriptive to allow search. Please keep this up-to-date.

  • Email list archive Each group should set up a mailing list, preferably on egroups.com which can maintain an archive for you. All mails should be sent only to the mailing list -- this way any discussion will be guaranteed to be archived. Mail headers should be constructed to facilitate search.

  • A search feature : Which can look through the site and find topics of interest. Free search engines are available. My advisor's page (www.cis.ohio-state.edu/~jain) has one.

  • Debugging notes : You can optionally maintain debugging, or coding notes here - just for the convenience of searchability.

• Time management At the beginning of every semester, we will make time lines, high-level action items etc for every group. Get hold of me asap if I have not done it for you. Use calendering services like those offered online by Yahoo to help you manage your progress. It can send you reminder mails about action items/progress etc.

• Debugging help: Just get hold of me and drag me to the code, and I will help you. Of course, you should try to first make use of the help you can get within the lab.

Beyond the above guidelines, I have also have set up some requirements for funded students. Basically, these requirements quantify research output. Funded students are responsible for delivering on these requirements. It is my duty to help you in any of the ways mentioned above, but the responsibility is ultimately yours. These are:

• I will fund a student only if the work ultimately leads to a MS Thesis/Project or PhD dissertation. Transient funding for a semester etc is also subject to this rule. This is the minimum period of time when some impactful research output can be produced.

• The student is responsible for maintaining a conceptual "big-picture" and the current "prioritized action-item list" of the project, and identify ways of enhancing it and find the necessary tools (analytical, simulation, experimental) to solve the problems involved. I should not be reiterating these issues during our interactions. This way we can focus our time on new and creative stuff. The student is also responsible for maintaining all knowledge about the project (including skeleton tech-reports, bibliography, action items etc) on the project web page which can then be searched.

• The student is responsible for managing his/her time. This includes taking on the right amount of teaching, course-work or other load to free up enough time for the project responsibilities. Yahoo calendaring services are great for maintaining action item lists, self-imposed project deadlines, schedules, sending reminder emails etc.

• The student is responsible for a technical report of the quality of a conference paper every funded semester . It is the responsibility of the student to manage their time, course-load, and leverage the knowledgement-management tools, lab and human resources (including colleagues and me) to meet this goal. With a large number of projects, I simply cannot manage this responsibility for each student.

• The student is also expected to publish at least one archival JOURNAL paper by the end of the masters project (or roughly every three semesters).

• Every masters project or thesis, or PhD dissertation should have both a simulation/analytic and an experimental component . New schemes designed should be prototyped and tested on the Linux platform. Experimentation over the real Internet is highly encouraged. From my recent experience, research work which has all these components has higher credibility in the community. Also, networking companies want and pay a significantly higher premium for people with this kind of experience.

• It is recommended that each student make a presentation of their work once a semester to the entire team. One possible way is to take one semester's work and present it at the beginning of the next semester.

I encourage students to go on for a Ph.D. because I believe it is not only a crowning academic achievement, but also gives you a broad as well as deep perspective into the field and opens up unique career opportunities. The first concern of many prospective PhD students is the time it will take to complete a PhD. My rough quantitative expectation is three archival journal papers. Going by the above guidelines, you should meet these requirements in 3 years.

I also have the following incentive structure for PhD students.

• For PhD students who are are generating a lot of research output (publications, software distributions etc), and/or are creating opportunities for new funding, I will buy laptops so that you can be flexible from where and when you work.
• I will make more funding available to attend top conferences , and fund you to attend every conference where you have a paper - be it anywhere in the world.
• If the output level is really high and you deserve to finish earlier, I will fund enough to get rid of the credits requirements so that you can finish earlier.

Finally, I look forward to every student colleague to go out and be highly successful, both intellectually, impact-wise and financially. I have a long list of industry contacts whom you can leverage for job positions or summer intern positions. This includes start-up and small companies as well.