Everyone wants 5G mobile technology, says IDC associate vice-president Hugh Ujhazy. It means you’ll be able to watch 4K videos while on the move. Yet, he says, while there is a focus on 5G's offer to consumers, the real business case for building new networks lies with enterprise customers. Bill Bennett reports
Spark managing director Simon Moutter plans to build a 5G network in time for the 2020 America’s Cup. Many of the top carriers around the world expect to have live networks at around the same time.
Which means we are two years away from a worldwide launch of next generation mobile.
That’s not long. In some respects, the technology is already here. Spark tested 5G outdoors in Wellington earlier this year. More recently, the company held indoor trials at its Auckland headquarters. While Vodafone has yet to announce its timetable, it has also run a successful pilot.
IDC’s Ujhazy says that over the long haul, carriers will spend vast sums of money building new 5G networks. This won’t happen all at once. He says the first stage will see towers in key locations upgraded to the newer technology. The real cost comes later when they add lots of small sites and upgrade the backhaul (links between core and smaller networks) to cope with more data.
He says: “If you follow the money trail, you can see that carriers won’t get all that build-cost back from consumers. Users may be willing to pay a small premium for faster data, but not enough to give carriers a return on their investment. That return is going to come from selling business-class services over the same network.”
Ujhazy says carriers need business users and enterprise customers to make 5G pay. Two things matter to carriers: the number of customers and ARPU, the average revenue per user. ARPUs for enterprise users are far higher than for consumers. That’s where carriers can find a healthy profit margin.
The good news for carriers is that enterprises will pay a premium for key 5G characteristics. He names low latency, high performance and beamforming (a radio frequency filtering technique) as useful. But they won't attract high margins on their own.
In contrast, network slicing has the potential to be a real money spinner.
Network slicing plays a similar role to software defined networking in fixed networks. It allows carriers to offer customers private, dedicated virtual networks. There is no need for extra infrastructure, it depends on software.
Carriers can optimise network slices for specific functionality. If they want it, customers can buy guaranteed uncontested bandwidth. Ujhazy says this can even mean giving certain customers their own spectrum.
He says: “Enterprise customers will be most interested in the carriers’ ability to use network slicing to create segregated networks within a 5G network. These segregated networks will be far more secure and reliable than the public network.”
New uses for mobile networks
This opens up new ways of using mobile networks. Take, say, autonomous car companies. They might buy network slices that give cars a way to communicate with each other, and with the road infrastructure, without using the public network, which can get congested. Ujhazy says: “This is all about safety. You wouldn’t want to be screaming down the road in a fast autonomous car at the same moment everyone switches on to watch a streaming high-definition rugby test match.”
Likewise, a railway company could have a network to control its driverless trains. The application might only have low data throughput, but it needs low latency and high reliability. Ujhazy says the economic value of, say, being able to run trains two minutes apart, instead of four minutes apart, through a crowded central city train network, would justify the cost of buying a network slice.
Away from transport, Ujhazy says remote health will be a key application. He also expects the finance industry to want secure segregated networks. Putting point-of-sale transactions on a private network is worth a premium.
Another key target market is those organisations with remote, high-value physical assets they need to manage. A civil engineering firm could use a secure network to operate construction vehicles and specialist equipment.
Seeing enterprise as the main 5G business case turns industry convention on its head. Ujhazy says until now carriers always built and optimised networks for consumers. They talked of putting towers where people lived, worked and played.
With 5G, carriers will prioritise enterprise customers’ needs. He says it means different ways of deciding when and where to upgrade towers. You can expect to see an emphasis on business districts and industrial zones, less on suburban networks.
One consumer application of 5G could be to deliver fixed-wireless broadband. Ujhazy doesn’t think this will pose a threat to established fibre networks. He says it will play an important role in places where there isn’t a lot fibre now or likely to be soon. He also sees this as significant for those industries in remote places, like mining.
“Enterprise customers will be most interested in the carriers' ability to use network slicing to create segregated networked within a 5G network. These segregated networks will be far more secure and reliable than the public network”
Making IoT really hum
Many of the enterprise applications using 5G, and the business case for the upgrade, boil down to the Internet of Things (IoT). It may seem strange that we need a new network for IoT applications when there are already many IoT networks.
Ujhazy says: “5G will become the standard. It unifies the market. It brings together high-speed applications, the mid-market and the low-power providers under one common platform. Until now, IoT delivery has been at the experimental stage. Over time, all the networks will merge under 5G.
“It’s very attractive if I’m an operator or a device manufacturer. I’m able to give you a low-power, long-term solution that can start off giving you little trickles of data but copes when things jump up to high throughput.
“Say you have a surveillance application. You might start off collecting a tiny bit of data every 10 minutes. Then something happens and you can switch to 4K video. If I can have one modem that does all this, then it makes it cheaper and easier to deploy the technology.”
One reason for the urgency of rolling out 5G is that, after years of hype, the IoT is now hitting its stride. Ujhazy says: “Look at all the parts available now. We’re seeing an increasing roll out of IoT platforms.
“There’s Amazon’s Lambda and Greengrass [for IoT services and devices]. Cisco has announced a new platform with greater capability. Meanwhile, you’ve got carriers offering their own platforms. The accessibility to the end-point data is all there. A successful IoT solution is something you don’t see.
“We’re getting more and more connected. When you buy a new car, it’s connected. Home lighting and heating are connected. People are putting beacons in shopping malls. There’s V2X [Vehicle to Everything communications] in transport networks, and the IoT is everywhere through the supply chain. It’s not a big bang, it’s a gradual thing, where more and more devices are connected.”
New Zealand’s 4G network is only five years old. If 5G arrives as promised in 2020, carriers will have only had seven years to recover their 4G investment. This is less of a problem than it might appear. Ujhazy says those 4G networks are not going anywhere, they will live on alongside 5G for many years to come. More to the point, the business case for 4G mobile is distinct from the case for 5G.
Sydney-based Hugh Ujhazy is associate vice-president, IoT and Telecommunications for IDC Asia-Pacific. He leads teams researching the Internet of Things, fixed and mobile networks in the region.