5

I got an idea for a research paper in the field of Computer Science, the topic related was about computer networks. At the beginning, I was working with a colleague who had a lot of background in programming networks protocols, but the problem was that due to some other tasks of him he had to left the project.

Now the problem is that I do not have anybody with the programming networks skills enough to continue with this research, so I decided to turn it into a theoretical paper; like the ones that are presented in other fields like math.

I would like to know, in a bird's eye, what should I include so that my research would be considered and not rejected because it is missing the experimental part. I have based my article in theoretical parts of CS such as graph theory and so on.

Any advice?

  • 1
    I'm voting to close this question as off-topic because it's not about academia generally, but about a specific concept in computer science. It might be better suited to the Computer Science Stack Exchange site. – Bill Barth Oct 4 '15 at 17:57
  • 4
    @Bill: I respectfully disagree with the close vote. The OP's question is about the academic process. It is not about a specific concept in CS: in fact, she hasn't mentioned any specific concepts in CS but only made brief, broad allusions. – Pete L. Clark Oct 4 '15 at 18:16
  • 2
8

There is an academic field called theoretical computer science (TCS), which lies at the border of math and CS. There are lots of venues for work in this area. However the motto for TCS is certainly not "We couldn't do the implementation, so we decided to make it theoretical"! I am by no means an expert in TCS (I know just a little more than absolute zero), but I don't think that computer networks is a key TCS topic: it sounds more applied than that. (I just checked this wikipedia article, which seems to largely support my position. Obviously an expert opinion would be better here.)

My advice to you is to find someone else to help you with the implementation / experimental part so that you can carry out your research along the lines you originally envisaged and not theoretically-as-a-consolation-prize. How can you do this? You could (try the things at the top of the list first):

  • Ask people in your department. Even if they do not have the skills themselves, there are more of them than you so the chances they know someone who does will increase.

  • Ask about it on the internet, including various SE sites.

  • Attend relevant conferences and seek out people with this expertise.

  • Visit another department and/or a specific person whom you think can help.

  • Write up a preliminary version of your work which only does the theoretical part and makes clear that you are looking for someone to help with the implementation. Post this work on your webpage / the arxiv / some (presumably not top tier) conference.

4

As someone who sits between theory and practice of computer networks research, it is hard to advice what you need to do without the specific topic. I suggest that you model the fundamental problem mathematically, and then evaluate the problem analytically as well as conduct some numerical studies. This still requires some 'implementation', but it is no longer packet-based simulations per-se. I would also look at papers from IEEE INFOCOM or IEEE Trans. on Networking or Elsevier Computer Networks -- these journals tend to carry more 'theoretical' papers. Quite often, they apply tools/concepts from TCS. Hopefully, these papers give you some ideas the level of theory and implementation that you need to present. Note, in general, you need to have a feel of the expectation of a given community. Some venues reject a paper without any actual implementation whilst others will drop a paper if there are no proofs in sight.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.