Now I am trying to choose my Ph.D. topic. Keywords are: multithreading software, software verification (include model checking), deadlocks. I want to create some interesting technique to make the verification process of multithreading programs easy.
My collegues and I have already developed a mathematical model of deadlocks, an algorithm for detecting deadlocks based on this model, and a tool for creating and detecting deadlocks in models of multithreading software. But our research is more theretical and less practical. Now I try to find the way to make our research more useful, and I want some advice from ResearchGate's members. And do you think that this topic is actual and interesting for Ph.D. research?
Thank for your help :)
Although I do not have any experience in that field of work I find your topic of choice most useful. I'm guessing that, since you wish to go practical, you intend to build a library (module) for one or several programming languages.
Since your project is quite a bit out of my league I can only use my experience and therefore the difficulties I've found to serve as example. Most of my programming is done with the high-level language Python and I've actually considered (to gain performance) doing some multithreading programming but the documentation for this libraries is not that good and I've never heard of an IDE (but please consider that I've not used that many IDE either) that had good support and analysis for this kind of algorithms.
Having such a tool that would simplify developing that kind of software in any language is, in my opinion, an excellent contribution to, at the very least, the scientific community, specially the part of the community that lack advanced programming skills (like me actually!!!). I hope I've interpreted correctly your intentions.
Good luck for your project, and should you choose to pursue it , my gratitude also. Best Regards...:)
Thank you for you answer and interesting opinion. Good luck in your research, too :)
Dear Ivan
Your topic is of interest, especially in complex systems modeling. From my little experience in modeling and simulation of "functional landscape modeling" as contributor and scientific animator for modeling/simulation ecosystem plateform projects (see publications related to LandMod 2010 congress), difficulties are often understood as technical, while they lie in the non explicit model working frames. More precisely, heterogenous approaches do not explicit the spatial and temporal resolution they are working in. Another key point raises from the data exhanged from the various process, and more precisely their mutual conflicting updates in case of dynamic systems. There is, on the application part of complex system, a luck of formal approaches, despite the fact that some do exist.
So, if your interest relates to complex system model application frames, aiming to develop distributed/parallel implementation frames I do encourage you to look close to DEVS formalism (Ziegler in the 70's) and more precisely to the past half decade renew of this formalism from new developments (applications, parralel implementation, ...)
You'll find applications from these but also a theoritical mathematical frame to be compared with, this can drastically help you in the positionning of your own approach and help, to clarify your PhD subject.
Hope that 2012 bring you Health, Joy and further inspiration ! Sincerly ..:)
Thank you for your answer and good ideas. I will read more about DEVS formalism.
Best Regards
Hi, I strongly suggest that you look for litterature in academic search engines like citeseerX. In addition, you can find scores of general resources, advice and tools to organize your PhD at Academic Joy : http://www.academicjoy.net. In this site, there is a comprehensive list of PhD disserations databases from around the world. Good luck.
- Badges
- Science topic
- Similar topics
- Engineering