This is a broad question.

The algorithms to achieve high spectral efficiency are essentially the same in Massive MIMO as in conventional multi-user MIMO, which has been a research topic for 20 years. Linear detection based on ZF/MMSE and linear precoding based on ZF/MMSE are typically used, so there is not much novelty there. The main novelty of massive MIMO lies in the channel hardening behavior which makes it possible to derive ergodic spectral efficiency expressions that only depend on the large-scale fading and not on the small-scale fading realizations. This can be utilized to derive novel power control algorithms that utilize only the large-scale fading, which makes them easier to implement in practice.

The same thing applies for energy efficiency optimization. We can optimize the long-term energy efficiency, instead of the short-term energy efficiency, by using the ergodic spectral efficiency expressions.

What I find particularly interesting and novel in massive MIMO is the possibility to not only optimize the power allocation, but also the hardware configuration. This is discussed in the following paper:

Emil Björnson, Luca Sanguinetti, Jakob Hoydis, Mérouane Debbah, “Optimal Design of Energy-Efficient Multi-User MIMO Systems: Is Massive MIMO the Answer?,” IEEE Transactions on Wireless Communications, vol. 14, no. 6, pp. 3059-3075, June 2015. (http://arxiv.org/pdf/1403.6150) 

If you have any more detailed questions I can try to answer them here.

thank you bjornson sir, for your valuable time and answer