The great 5G bake-off: where are we and where are we going?

Nice and easy does it – at least to start

CCS Insight principal analyst for operators, Kester Mann, says the 5G story has advanced a long way in the past year, but there is still a long way to go.

“We had the first standards locked down in December [2017], and that paves the way for very early launches that we will probably see in 2019,” he says. “We may see something at the back end of this year, but it will be very small scale, and tenuous if it can be called 5G.

“We’re expecting handsets to become available in the early part of next year. Whether those are large or small scale, what technology we’ll be using – whether we’re talking about fixed-wireless access (FWA) or 5G as a pure mobile network – is debatable.”

Speaking at a recent Westminster eForum event on 5G priorities for the UK, Graham Louth, partner at Aetha Consulting, said 5G was a lot further on today than people thought it would be this time last year.

“3GPP has accelerated the standardisation of 5G, and as of December 2017, it had frozen the first standard for 5G New Radio (NR). The standard that has been frozen is for non-standalone 5G, which means it only works if you have 4G in the same location, it essentially relies on existing 4G and adds 5G for more capacity,” says Louth.

“All the other parts of Release 15 – the first full 5G standard – are scheduled for release by the middle of this year. The focus of that standard is mainly on enhanced mobile broadband – the more advanced features are going to have to wait for release 16.”

Upcoming release

Louth predicts that with the 5G NR standard frozen, manufacturers will begin to release product in the coming months, ahead of standardised deployments beginning in 2019 in the most advanced markets.

John Lenns, vice-president of product management at Oracle Communications, says that while there is still lots of work to do to complete the first phase of 5G specifications, the NR standard is enough for trials to begin with 5G radio talking to an evolved packet core (EPC), something that is already happening in South Korea and the US.

“Following these non-standalone 5G NR trials leveraging EPC, towards 2019 we are going to see some of the first trials of 5G cores built on the new service-based architecture,” he says.

However, Caroline Gabriel, senior researcher at Analysys Mason, says that while standards and some early use cases for commercial implementation of 5G technology have indeed come together quicker than many had expected, things won’t properly get moving much beyond this until after 2020.

According to Analysys Mason, a lot of mobile network operators (MNOs) are waiting to see what Release 16 (due next year) brings in terms of enabling new investment models through, for example, unlocking the potential of unlicensed spectrum, so that 5G becomes more than “just another 4G network with a faster radio”, says Gabriel.

“Unfortunately, Release 15 is to some extent just that, and the majority of operators don’t see a business case yet,” she says. “Operator research that we’re doing indicates that’s not really going to get going until past 2020, and until then, 5G will mostly be used to add capacity to 4G where required.”

Howard Jones, network communications manager at EE, the UK’s largest MNO, tells Computer Weekly this is exactly where his organisation is focusing its efforts for now.

“3GPP did a great job in accelerating the non-standalone aspects of 5G, and that’s what we’re putting into practice; non-standalone NR running on an evolved 4G core. I don’t think anybody in our market is going to be deploying a full 5G core at launch,” says Jones.

For EE, the most realistic early use case for 5G will be enhancing the mobile broadband experience for consumers through hugely increased capacity.

For the next few years, as 4G evolves towards 5G, there won’t be many disruptive changes in consumer behaviour, although, as Jones points out, we will continue to see huge growth in data usage, video streaming and more mainstream applications for Augmented and virtual reality (AR and VR). It is only a few years down the line that more advanced use cases will become truly practical.

Oracle’s Lenns reckons that it will take until 2023 for the majority of operators to be comfortable with 5G standalone cores. “We will start to see 4G traffic and core functionality moving to a 5G next-generation core,” he says. “5G roaming should also be implemented by then, along with the introduction of new network slices enabling new services and use cases.”

The Swiss Army knife of networks

Ultimately the predicted diversity of use cases will indeed be critical to the world of 5G mobile networking, and anybody who positions the technology merely as an enhancement to consumer mobile broadband services is missing the entire point of what 5G will eventually become once technologies such as network slicing hit their stride.

Louth at Aetha compares 5G to a Swiss Army knife for wireless communications, with multi-functional tools to address many different uses.

So while consumer mobile broadband is important, there will also be attachments for 3D video, agriculture, augmented and virtual reality, automotive, drones, energy, factories, transport, smart homes, and voice over NR.

“So perhaps,” agrees Ryan Ding, executive director and president of Huawei’s carrier business, “the traditional consumer business is not so interesting”.

The Swiss Army knife concept also lends itself to the increasingly widespread idea that there will be many more 5G networks than there are 4G networks. In the future, many CIOs will run their own, private 5G networks, particularly those in sectors going all-in on the IoT, such as heavy industry, agriculture and transport.

“Whilst the consumer applications we are looking at rely on a wide area network (WAN) as is provided by, for example, an MNO, some of these use cases would probably be better served by a private network,” says Aetha’s Louth.

“If what you’re doing is using 5G to monitor the performance of machine tools in a factory, you don’t necessarily need to do that by means of a public network – it may be more efficient to do that with a private network,” he says. “The objective with 5G is to allow a common technology that enables all these applications to be delivered.”

Barriers to 5G roll-out

Huawei’s Ding says it will be up to the government to reinvent how networks are built, and warns it will not be possible to lean on MNOs to roll out 5G in the same way as they did with 4G.

“As fibre-to-the-home (FTTH or FTTP) and 5G have more coverage the investment in passive network facilities is increasing, so the government’s role will be to help carriers reduce that investment,” says Ding.

“For example, the Irish government encouraged carriers to work with electricity companies, which led to a huge cost reduction in optical fibre deployment, and in many other countries, for example in Spain, the government encourages public-private partnerships to reduce the investment.”

Ding also believes the cost of radio spectrum will need to be addressed. He believes these costs are still too high across Europe, and says local regulators can learn from the examples of China and Japan, where spectrum is cheaper, therefore freeing up cash for the actual network.

“I have talked to European carriers about 4G deployments and they told us a large part of investment goes to spectrum; this is not good for enhanced mobile broadband development in Europe,” he says. “In this area, I think government has a very positive role to play.”

William Webb, CEO of Weightless SIG, the organisation developing and promoting the Weightless W low-power wide area network (LPWAN) standard for the IoT, agrees MNOs will not solve the roll-out problem alone simply because they are being squeezed financially, making them an increasingly unattractive target for investors. “It is unlikely coverage will be improved by MNOs going out and spending to improve it,” he says.

National coverage a challenge

T-Systems CTO and UK security head Scott Cairns argues that there will nevertheless be a need for the government to hold the MNOs accountable for coverage. The two previous 3G and 4G standards came with a stipulation that a certain proportion of the UK population must be covered, which had the effect of leaving huge geographical areas of the country out of range.

“5G is going to be used in a very different way to previous technology, and will be relied upon for sensitive and critical data feeds. Whilst redundancy and buffering can be built into the applications employed, we need a full coverage network across our country to allow new technologies like autonomous cars to operate seamlessly,” he says.

Weightless SIG’s Webb agrees: “We do need better rural coverage and that’s best done with a single network with mandated roaming. The best way to do that is to buy that as a coverage service – in a way we’re doing that already by requiring EE to provide ESN [the new Emergency Services Network].”

EE’s Jones says that the barriers to establishing nationwide mobile coverage in the UK – around areas such as planning policy, access to power, local topography, and so on – are well understood and will be largely similar for 5G as they were for 4G.

However, he rejects the idea of national roaming on a single network-of-networks “We believe if there were to be more obligations on coverage it should be across operators,” he says. Moreover, there would have to be fundamental changes to government policy with regard to how nationwide coverage would be funded, and a more “give and take” attitude from Westminster.

More fibre urgently needed

Jonathan Freeman, product and technology director for telecoms and M2M at Arqiva – which has been running tests of FWA-based 5G infrastructure in London, says that the way in which 5G will be deployed will be fundamentally different to 3G and 4G, but he identifies other challenges besides ensuring adequate coverage.

For Freeman, the biggest of these is backhaul access. Because 4G networks were built by adding new technology to existing masts and sites, the number of operator points-of-presence (PoPs) did not fundamentally change.

However, with the Swiss Army network pitched as a driver to the industrial economy, not merely as a means to enhance consumer experiences, the number of nodes will increase exponentially.

“Small cells will dominate, and there’s a need for everything to be fibre-fed,” says Freeman. “5G and how it is deployed will be radically different. If we don’t crack economic fibre access we are not going to see a 5G network that looks like 5G.

“4G was make-do-and-mend, but now we have to crack the challenge of getting economic dark fibre, and we have to move beyond the concept of passing by a node and do it proactively.”

Freeman also believes the industry needs to advocate for greater diversity of assets, pressing more street furniture into use, for example.

Government must be bolder in fibre ambitions

Rajiv Datto, COO at network infrastructure and datacentre services supplier Colt, agrees that the government should be even bolder in its ambitions around fibre.

“The need for micro or small cells to support 5G make it a fibre intensive requirement, yet challenges in deploying fibre currently come in many flavours – whether it’s duct availability in terms of Openreach, or the number of traffic management approvals needed to lay fibre around the City of London,” he says.

“The ability to deploy fibre is the key, and making it faster, easier, and less expensive should be one of the primary regulatory prerogatives. If I was to ask one thing of [digital minister] Matt Hancock, it would be to align and streamline the approval processes to deploy fibre to get to that last mile. This would be a big step in speeding up 5G in the UK.”

In the second part of our look at the world of 5G mobile networking, we will look beyond current standards and roll-out to find out how UK plc is innovating and positioning itself to lead in 5G, and find out what might be lying in store at Mobile World Congress.