What are the 5G candidate technologies?

By Dimitris Mavrakis

Telecoms.com periodically invites expert third-party contributors to submit analysis of a key topic affecting the telco industry. In this piece Dimitris Mavrakis, Principal Analyst, Intelligent Networks, at analyst firm Ovum, looks at the technologies likely to contribute to the still-undefined 5G standard.

The LTE ecosystem gathered recently in Amsterdam to discuss the commercialization and future of mobile networks. As expected, 5G was a hot topic of discussion but there is still considerable confusion in the market regarding all of its aspects: technology, commercial opportunity, application in verticals, and many others.

This confusion is accentuated by the fact that the majority of mobile operators have still not come across a successful way to monetize their LTE Networks. But if we look to the past, can we identify what the next step towards 5G should be?

Setting the scene

The purpose of this article is not to discuss the requirements of a 5G standard, but to identify technology candidates, relevant organizations, and how the 5G standard is likely to be developed. The requirements of a 5G network have been identified and are outlined by the 5G-PPP.

Making up new definitions in the telecoms market is generally frowned upon and in many cases futile: ITU defined 4G to be IMT-Advanced (100Mbps when user is moving, 1Gbps when stationary) but the market has decided otherwise. LTE – and even LTE-Advanced – cannot provide anything remotely close to these requirements, but on the other hand, T-Mobile US called HSPA 4G and Free France did the same, both for marketing and competitive reasons.

A new mobile network generation usually refers to a completely new architecture, which has traditionally been identified by the radio access: Analog to TDMA (GSM) to CDMA (UMTS) and finally to OFDM (LTE). It is clear that 5G will require a new technology – and a new standard to address current subscriber demands that previous technologies cannot answer. However, 5G – driven by current traffic trends – necessitates a complete network overhaul that cannot be achieved organically. Software-driven architectures, fluid networks that are extremely dense, higher frequency and wider spectrum, billions of devices and Gbps of capacity are a few of the requirements that cannot be achieved by LTE and LTE-Advanced.

Technology candidates for 5G

Current technology developments and user demands are merely providing a glimpse of the nature of 5G networks. At the moment, cost is not a major driver of 5G technology discussions, allowing a much wider list of candidate technologies to be considered. In order to discuss these technologies meaningfully, cost is also ignored in this article. It is the radical and disruptive changes to the existing network that are outlined.

It is clear that a new air interface will be needed. ZTE went as far as to suggest that a 5G network will consist of several air interfaces coexisting in the same network. From a theoretical perspective, this is ideal (e.g. OFDM does not lend itself to small cells and hetnets and there are other, better candidates), but from an operational and economic perspective, this would mean significant development costs and deployment effort.

A selection of technology candidates is outlined below:

Extreme densification

Network densification is not new. As soon as 3G networks were congested, mobile operators realized the need to introduce either new cells into the system or more sectors. This has evolved to include many flavors of small cells, which essentially move the access point much closer to the end user. There is simply no other way to increase the overall system capacity of a mobile network significantly. 5G networks are likely to consist of the several layers of connectivity that hetnets are currently suggesting: a macro layer for lower data speed connectivity; a very granular layer for very high data speeds; and many layers in between. Network deployment and coordination are major challenges to be addressed here, as they increase exponentially with respect to the number of network layers.

Multi-network association

Several networks are currently providing connectivity for end-user devices: cellular, Wi-Fi, mm-wave, and device-to-device are a few examples. 5G systems are likely to tightly coordinate the integration of these domains to provide an uninterrupted user experience. However, bringing these different domains together has proven to be a considerable challenge and Hotspot 2.0/Next Generation Hotspot are perhaps the first examples of cellular/Wi-Fi integration. Whether a 5G device will be able to connect to several connectivity domains remains to be seen, and a major challenge is the ability to successfully switch from one to another.

Full duplex

All mobile communication networks have relied on a duplex mode to manage the uplink and downlink. There are frequency duplex – FDD (such as LTE, where uplink and downlink are separated in frequency) – and time duplex schemes – TDD (where the transmitter and receiver transmit at different points in time, as in TD-LTE). A duplex mode is necessary to coordinate uplink and downlink, but full duplex technologies are now being discussed. In these schemes, a device both transmits and receives at the same time, thus achieving almost double the capacity of a FDD or TDD system.

There are major technology challenges to achieving what is essentially self-interference cancellation and major changes in both networks and devices are essential for full duplex. However, the potential increase in overall capacity is substantial, making full duplex a very important technology for the future of mobile networks.

Mm-wave

Lower frequency spectrum (450MHz-2.6GHz) – which is currently relevant for mobile communications – is almost fully congested. There are, however, massive amounts of spectrum in the higher spectral bands, which may reach as high as 300GHz. Naturally, network design for such high frequencies is much more complicated than the operators are accustomed to: as frequency increases, building penetration becomes more difficult, to the point where a simple wall becomes an opaque barrier for mm-wave signals. However, there are tens of GHz available in these bands, which may be used for short-range, point-to-point, line-of-sight connections, providing limitless speeds of wireless connectivity.

Mm-wave could be used by indoor small cells (in line with the extreme densification principle outlined above), which will provide very high-speed connectivity in confined areas. The high-frequency nature of mm-wave means antennas can be very small, thus only creating a small impact on device real estate. Nevertheless, Ovum believes mm-wave is an extremely radical technology and may require many years of R&D to be cost-effective for the mass market.

It is interesting to note that developments in mm-wave are not new: the WiGig alliance is focusing on 60GHz spectrum. Google also announced the acquisition of Altiostar (a start-up founded by ex-Clearwire engineers), which has been developing technology for the 60GHz band.

Massive MIMO

MIMO has been deployed in LTE-Advanced networks, where the base station and end-user device uses more than a single antenna to increase link efficiency. Massive MIMO refers to the network, where the base station employs a much higher number of antennas that create localized beams around each connected device. The gains in capacity are enormous but so are the technical challenges associated with this concept. However, there is new interest in the market for these concepts, exemplified by a startup called Artemis, which has developed a product called pCell, based on this technology.

Virtualization, software control and cloud architectures

A parallel evolutionary trend to 5G is software and cloud, where the network is driven by a distributed set of data centers that provide service agility, centralized control, and software upgrades. SDN, NFV, cloud, and open ecosystems are likely to be the foundations of 5G and there is an ongoing discussion about how to take advantage of these architectures. Although they are not new – and likely deployed for 5G – all of these concepts are necessary to provide the increased capacity and connectivity of billions of devices that 5G specifications promise.

What happens next?

It’s clear that the technologies outlined above hold great potential and are a revolutionary step when compared with existing network technologies. The process for selecting which technology will be finally implemented is likely to be a long one and depend on performance, implementation, cost, politics, and many other issues. But it is reasonable to assume that it will be the technologies that cost the least that are likely to be implemented, as has been the case with LTE-Advanced.

Ovum’s Intelligent Networks practice is a multinational team of experts assessing the present and future of mobile and fixed telecoms networks. Through constant communication with vendor and service provider communities, Ovum’s telecoms network analysts have a firm understanding of current network trends as well as the future evolution of network infrastructure. Several disruptive topics, including 5G, NFV, and APIs, are shaping their research agenda, ultimately leading to their vision of the telecoms world in 2020. Visit their webpage at www.ovum.com