A while ago I started my CS PhD, but I'm really doing pure maths as it's in cryptography.

For (complicated) reasons, my funding may soon be moved to EPSRC CASE, however as part of this I need to put together a document with a plan of work "identifying phases" in my project, and giving target dates for completing "deliverables".

I feel that this is unreasonable given the sort of research I'm doing. I'm trying to find out if a certain thing in cryptography exists, prove it, and produce a proof of concept if so. The direction of my research will change significantly, and be driven almost entirely by, my results along the way and the papers I read. I haven't planned more than a few months ahead and don't think I can or should.

My supervisor says he agrees with me, but this sort of thing is fairly common when applying for funding/dealing with management, and I should view it as practise.

I've tried to put something together, but not knowing which areas of maths I'll be working with more than three months later(probably including some I haven't heard of yet) couldn't even _fabricate_ anything.

Can anyone help me? Thanks!

Most of us haven't any real idea of the direction our research will go in, it's heavily paper driven and so forth. You just need to propose a framework. Consider ethnographers having to put forward proposals before they've gone to the field!