5G will be a truly converged system supporting a wide range of applications from mobile voice and multi‐Giga‐bit‐per‐second mobile Internet to D2D and V2X (Vehicle‐to‐X; X stands for either Vehicle (V2V) or Infrastructure (V2I)) communica- tions, as well as native support for MTC and public safety applications. 3D‐MIMO will be incorporated at BSs to further enhance the data rate and the capacity at the macro‐cell level. System performance in terms of coverage, capacity and EE will be further enhanced in dead and hot spots using relay stations, hyperdense small‐cell deployments or WiFi offloading; directional mmWave links will be exploited for backhauling the relay and/or small‐cell BSs. D2D communications will be assisted by the macro‐BS, providing the control plane. Smart grid is another interesting application envisaged for 5G, enabling the electricity grid to operate in a more reliable and efficient way. Cloud computing can potentially be applied to the RAN,
and beyond that, to mobile users that can form a virtual pool of resources to be managed by the network. Bringing the applications through the cloud closer to the end user reduces the communication latency to support delay‐sensitive real‐time control applications.
It is envisaged that 5G will seamlessly integrate the existing RATs (e.g. GSM, HSPA, LTE and WiFi) with the complementary new ones invented in mmWave bands. MmWave technol- ogy will revolutionise the mobile industry not only because of plenty of available spectrum at this band (readily allowing Gbps wireless pipes), but also because of diminishing antenna sizes, enabling the fabrication of array antennas with hundreds or thousands of antenna elements, even at the UE. Smart antennas with beamforming and phased array capabilities will be employed to point out the antenna beam to a desired location with high precision, rotated electronically through phase shifting. The narrow pencil beams will enable the exploiting of the spatial DOF, without interfering with other users. The small antenna sizes will enable Massive/3D MIMO at BSs and eventually at UEs. The mmWave technology will also provide ultra‐broadband backhaul links to carry the traffic from/to either the small BSs or the relay stations, allowing further deployment flexibility for the operators, compared to the wired (cop- per or fibre) backhaul link. Hyperdense small‐cell deployment is another promising solution for 5G to meet the 1000x capacity challenge. Small cells have the potential to provide massive capacity and to minimise the physical distance between the BS and the UEs to achieve the required EE enhancement for 5G. The traditional sub‐3 GHz bands will be employed for macro‐cell blanket coverage, while the higher frequency bands (e.g. cm‐ and mmWave bands) will be employed for small cells to provide a spectral‐ and energy‐efficient data plane, assisted by a control plane served by the macro‐BS .
Along with the development of new RATs and the deployment of hyperdense small cells, the existing RATs will continue to evolve to provide higher SE and EE. The data plane latency (round‐trip time) of the LTE‐A system is around 20 ms, which is expected to be reduced to less than 1 ms in its future evolutions . Moreover, the SE of the existing HSPA system is 1 b/s/Hz/cell, which is expected to increase 10x by 2020 . The EE of the cellular system is expected to improve 1000x by 2015, compared to the 2010 level . The PHY (physical) and MAC (medium access control) layer techniques will be revisited for carrying short and delay‐sensitive packets for MTCs . Virtualisation will also play a key role in 5G for effi- cient resource utilisation in cellular systems, through a multi‐tenant network where a mobile operator will not need to own a complete set of dedicated network equipment; rather, network equipment (e.g. BS) will be shared among different operators. The existing cloud network concept mainly involves the data centres. Mobile network virtualisation will push this concept towards the backhaul and the RAN to allow sharing of backhaul links and BSs among differ- ent operators. Last but not least, it is envisaged that 5G UEs will be multi‐mode intelligent devices. These UEs will be smart enough to autonomously choose the right interface to connect to the network based on the channel quality, its remaining battery power, the EE of different RANs, and the QoS requirement of the running application. These smart and efficient 5G UEs will be able to support 3D media with speeds up to 10 Gbps.
5G is expected to be deployed around 2020, providing pervasive connectivity with ‘fibre‐like’ experience for mobile users. Apart from the expected 10 Gbps peak data rate, the major chal- lenge for 5G is the massive number of connected machines and the 1000x growth in mobile traffic. The ultra‐broadband and green cellular system will be the driving engine for the future connected society where anyone and anything will be connected at anytime and anywhere. In this chapter, we gave an overview of the potential enablers of 5G along with research and development activities around the globe, including Europe, North America and the Asia‐ Pacific region. Being in the prototype stage, standardisation is the next milestone to achieving 5G, which will be followed by the development phase for two to three years. The last phase is network deployment and marketing, which may take another couple of years, foreseeing a potential commercial deployment by around 2020. In the final section of this chapter, we illustrated the foreseen architecture for 5G, harnessing all the common views on the current technology trends and the emerging applications. In a nutshell, mmWave technology, hyper- dense HetNet, RAN virtualisation and massive MTC are all major breakthroughs being con- sidered for upgrading the cellular system to achieve 5G capability. However, these technology developments need to be fuelled by the allocation of new spectrum for mobile communica- tions, expected to happen in the upcoming WRC meeting.