Ants are widely used in algorithms to resolve complicated problems via many iteration of a simple behavior, there is even a conference held every year an this subject, this year's is the seventh : http://iridia.ulb.ac.be/ants2010/
Based on this, we can hope that the field is pretty much covered by specialists, and there is not so much left for us poor students.
But if we leave algorithms for just a moment and look at real ants, we see that the "ant behavior" used in algorithms is not very close to reality, and the ants are much more interesting than just wandering around and putting some pheromones behind them. So it could be very interesting to try simulating a behavior closer to how real ants act, from the biological and emergence point of view : how single beings without superior authority achieve great things like ants colonies ?
And what if the model that emerges from our work allows some applications in algorithm then ? This totally not impossible, and maybe we won't fall to far away from our first concern : algorithms.
So it seems that we will be working on real ants behavior rather than looking in the well studied field of basic ants algorithms.
for this we will have to read/listen/watch some things about ants. It seems that the book called "The Ants", by Bert Hölldobler and Edward O. Wilson is (was) a reference on this topic in 1990, so it could be interresting to look at it, some parts are available on google books :
and I also found a more recent one, in French, referring to "The Ants" as its inspiration, and willing to gather knowledge about ants adding research issues since 1990. I red the first chapter, which is interesting and with a lot of references :