The growth of data usage is increasing significantly as it won't only be your mobile or tablet or computer that consumes internet in the coming 10 years but plethora of other devices (wearable to machines with unique address and censor) will require connection to create Internet of things (IOT). This IOT approach requires multi-fold increase in the internet speed plus the no of devices to connect also needs to increase significantly, along with the transition to new ip addressing scheme (IPv6). So it is expected that 1 Gbps data connection from 4G wont be enough that time, which requires new radio access technologies, of which some candidates are FBMC (Filter Bank Multi Carrier), NOMA (Non Orthogonal Multiple Access), Massive MIMO, Cognitive radio, C-RAN (Cloud RAN) and multi technology carrier aggregation. As the connection speed demand higher bandwidth, extremely high frequency (MMW) can be expected to use with very high spectral efficiency and low latency.
Another imminent issue will be massive deployment of small cells (with hetNet?) that increases capacity, coverage and require least planning with dynamic configuration of huge no. of cells by full fledged SON (Self Organizing Network or there may be "No more cell concept" with Cloud RAN. 5G is also known to radically change the network performance measurement metric from bits per sec to probably bits per joule or bits per 100 metre of coverage. So with 5G, spectrum efficiency is not only the only criteria but also demands energy efficiency to be considered together.
Better cell edge connectivity also needs to be ensured with IRC(Interference Rejection Combining), ICIC(Inter Cell Interference Coordination) , COMP (Coordinated multi point transmission and reception) like features for seamless connectivity. Cloud computing with 5G will not just be the present day cloud storage but it should encompass real time applications to handle the data from various devices.
With such fiber like speeds and massive no. of connections, 5G should support wide range of service scenarios like including M2M (Machine to Machine), D2D (Device to Device), PPDR (Public Protection Disaster Relief) communications.
As most of the issues talked above are under research and it will take years to get the next generation mobile broadband standard, I will be following the development quietly closely (from the journals) which I also did for its predecessor 4G but did not write that much. Lets wait and see how much I put in for future write-ups on 5G.