Well, this is hot in the sense that everyone uses this sentense "... is unkonwn."
If you are not a genius, you'd better keep an eye on the development on this topic, but never work on it. The reason is simple, over last thirty years, many able resesearcher have worked on this and yet they still did not get it. Hard, hard hard.....
Recently, there are some new development in the field, for example, the degree-of-freedom analysis (which is approximation of the exact problem by focusing on high SNR regime). Google, for example, S.Jafar (at UCI) for his papers. Another advance is due to Tse (at UC Berkeley). He proposed a deterministic model, which turns out to shed insight on the real problem.