As 2018 was close to dying out, the first commercial 5G mobile networks were given life. Verizon and AT&T switched on 5G networks in the US before the turn of the year, while all three Korean operators went live with initial, though admittedly limited, 5G deployments simultaneously.
Though deployments currently live with a commercial offering can be counted on your hands, this is likely to change rapidly, as more and more mobile operators turn on their 5G offering. In fact, at Mobile World Congress, in Barcelona at the end of February, you may even see your first true 5G mobile phone.
Despite this, it is still early days, with most 5G rollouts focussed initially on fixed wireless access (FWA), according to the GSMA – the industry body that represents mobile operators worldwide.
According to David Hutton, head of networks at GSMA Technology, those 5G deployments will come in two phases: the first, which has already began in a number of markets, sees operators plugging in a 5G New Radio (5GNR) into an existing 4G network, running parallel to the existing 4G network. “That means the roaming and interconnect models between different operators remains the same for the time being,” he explains.
According to analysis from the GSMA working groups, including the GSMA wholesale agreements and solutions group, that means “wholesale that we have today are able to support the initial 5G rollouts”.
That’s just for now, however, with the so-called “phase two” rollouts just around the corner. According to Hutton, standards for this second phase of 5G deployments were agreed upon at the end of 2018, with testing likely to be carried out this year. Deploy-ments are likely to begin in 2020, he says.
So what is Phase 2? It is “real” 5G core networks and RAN – ones that aren’t reliant on 4G infrastructure. A network that can stand on its own.
“When we start looking at later deployments of 5G, we will see a new 5G core network deployed as well,” he adds. “There’s still some analysis to be done around that on the impact between operators. However the viewpoint at this time is that the current model of roaming and interconnect will look similar. It may be extended to support 5G.”
Hutton has been involved in mobile networks for a number of years. A veteran of Nortel, where he worked on GSM/UMTS product planning and standards, and Vodafone, where he was senior standards strategist for six years, he has been at the GSMA for over seven years.
During his time in the industry, he explains, Hutton has seen several generations of mobile technology roll out. The first commercial deployments of 3G were at the beginning of the millennium. 4G followed almost a decade later, with 5G deployments beginning towards the end of 2018.
Hutton claims to have never “seen this level of interest around a mobile technology” as he has around 5G.Why? He says the difference is the diversity of use cases that can be made for the fifth generation of mobile connectivity.
Use cases range from high bandwidth high throughput in mobile broadband through to massive connectivity for internet of things (IoT) deployments, again “very different to some of the low latency use cases” that the GSMA is seeing.
“It is not one size fits all. Because of this, it is targeted towards a lot of different industries,” he adds. “We’ve hit the right time globally and in society in that everything is becoming more digitalised. Industries are transforming as well. This requires they fit very well with what 5G can provide. So there is a huge industry interest and demand for this type of technology – 5G satisfies that.”
Challenges
One of the drivers for this is industry, but another key aspect to the acceleration of 5G deployments has been the support and determination of national governments to be seen as leaders in mobile technology. The industry is under pressure to deliver a strong 5G deployment in a number of marks.
Hutton explains: “Geopolitically, a lot of governments are interested in being seen as embracing 5G and positioning themselves as leaders in this technology. So you have consumers who are interested, operators who are looking at it as a way to improve capacity in their existing infrastructures as well as the new opportunities it provides to engage with industries and new verticals, and also the governments who are looking to drive forward and transform their societies and economies as well – embracing the whole digital societies.”
However, government interest doesn’t always mean the same as government help or support. In fact, a common message from the GSMA over recent years has been on mobile spectrum or, more specifically, the shortage of it. Connecting more people requires more bandwidth, but in most markets regulators or government bodies dictate the amount of spectrum that is made available to each industry, mobile, satellite, broadcast or other.
Governments often opt for auctions, selling spectrum to the highest bidders, and this costs the industry a small fortune – money that could be directed to deploying more masts or improving the backhaul infrastructure needed to support faster connectivity.
In November, the GSMA called on regulators to make available 80-100MHz of contiguous spectrum per operator in prime 5G mid-bands (such as 3.5GHz) and around 1GHz per operator in millimetre wave bands (above 24 GHz).
“Within the GSMA, we have a spectrum team who works very hard to do a lot of advocacy with governments and regulators to make sure there is enough spectrum allocated to operators worldwide, and making sure the methods for allocation are the right ones. When we talk about potential performance of 5G it is related to ensuring that the relevant amount of spectrum is allocated and is utilised to provide that performance. If we don’t get that then you won’t get the performance that 5G promises. We are advocating for more spectrum not just for 5G but also for previous generations.”
The GSMA sees this as important because “spectrum is the oxygen which the mobile network breathes and lives”, explains Hutton. In terms of alternatives to more spectrum, carrier aggregation, which combines different bandwidths into different spectrum bands and even across different generations, could boost both capacity and performance.
“With early 5G deployments (5G NR in 4G networks) you have dual connectivity, so the device is connected to both 4G and 5G at the same time,” he adds. “This means you can utilise the spectrum on both 4G and 5G at the same time. You can also use carrier aggregation across different bands.”
As we speak, Hutton is joined on the call by his colleague Michele Zarri, who is a technical director for the GSMA specialising in this area, and Mona Mustapha, a technical specialist with expertise in the field of IoT.
On spectrum, Zarri adds: “There is an opportunity for refarming. If you look at the efficiency for GSM to transfer 1MB it takes a couple of minutes but with 4G it takes a fraction of a second. So if you can use the spectrum allocated to 2G for 4G and 5G it can improve the capacity quite a lot. Unfortunately in some parts of the world this flexibility is not there yet, so that is something the GSMA is working on.
“Some 5G will be able to be deployed in the same band that 4G is deployed in so that will make more efficient 5G spectrum available in the last few years.”
This greater need for spectrum is even more apparent when you consider the potential impact of the internet of things.
GSMA intelligence figures shows that there are around 1 billion cellular IoT devices in use already. Ericsson predicts 29 billion connected devices will be in use by 2022 – around 18 billion of which will be IoT.
With so many devices on top of more cell phones than ever, spectrum will become even more of a commodity. According to Mustapha, the variety of IoT solutions means the impact will vary greatly. She explains: “In terms of capacity, IoT devices have many different use cases and some of them may only need a small amount of data and sometimes infrequently. So even though the number [of devices] seems quite huge, the actual bandwidth needed isn’t going to be all that great for IoT devices.
“We’re looking at technologies such as NB-IoT and LTE-M that will be able to help support all of these devices.”
Another consideration is backhaul and fronthaul. Changing from a distributed RAN model to a cloud RAN creates the need for another high capacity link between a virtualised baseband unit and the actual antennas. Zarri explains: “As part of our future networks programme we have one activity which is looking at backhaul optimisation, but there is also a major problem with fronthaul. We estimate a normal 5G site will need a 10Gbps connection which can be achieved with microwave links and E-band, but most likely we require fibre to overcome this.”
Fibre, of course, isn’t always cheap or easy to deploy, especially in urban areas. To resolve this, the GSMA has begun to look at some of the available alternatives.
“We are looking at how some of the technologies that have been started can mitigate the requirement of having fibre everywhere. We expect fibre deployments to continue to grow, but the deployment of FWA make us think that there will ultimately solutions that rely on 5G itself to backhaul in situations where 5G is too expensive.
“Other solutions include using more edge computing – moving more applications to the edge of the network – which mitigates the backhaul problem but that’s very much a case-by-case study. It is problem but we are taking the right steps to address it, but there is no doubt you need big pipes to connect this much network.
