They have divided up the new developments into five separate categories, some evolutionary and others more revolutionary.
Some of the changes were easily predictable. For instance, use of much higher carrier frequencies than 4G, to achieve higher channel capacities, with consequent short RF range, very small cells, and higher mode MIMO (also helped by the higher frequencies and short range).
There's also going to be increased emphasis in peer-peer or "device-centric" model, as opposed to just the cellular model of previous G.
The scheme if i remember correctly was used in satellite communications. i have not heard about this scheme being considered for 5G networks. No one is talking about it even METIS and 5GNOW.
I also had not seen where continuous phase modulation was being discussed for 5G. But even that isn't really new, since it's already been in use for quite some time (e.g. even in 2G GSM). A similar scheme to this which uses amplitude instead of phase is MLT-3 (e.g. Ethernet 100BASE-TX). Whose aim is also to keep channel bandwidth down, by reducing the number of transitions.
So then, why use a concatenated code with CPM, instead of the more popular iterative codes du jour, like LDPC?
Mohamed, sorry, what do you mean by D2D? Device to device? Or disk to disk? If device to device, as opposed to cell-based, some things I read say yes. For instance , as applied to vehicle to vehicle and vehicle to infrastructure communications, some claim.
Physical layer enhancements for 5G are namely the use of massive MIMO using millimeter waves for communications. Secondly the use is multi channels techniques other than OFDM namely FBMC, UFMC, GFDM, BFDM.