ake78 (3D & photo) - Fotolia

The 5G wheels are spinning in APAC

Adoption of 5G across the Asia-Pacific region will be led by China, South Korea and Japan, but telcos will need to find the right pricing strategy to compete with IoT connectivity upstarts

This article can also be found in the Premium Editorial Download: CW ANZ: CW ANZ: Riding the wave of enterprise mobility

As the first 5G networks begin to appear across the Asia-Pacific (APAC) region from 2019, all eyes will be on the types of applications that will benefit from the increased bandwidth and the business opportunities that will emerge.

The world has seen similar upgrades in mobile connectivity, each time spurring new innovations and applications, from video calling and mobile gaming to over-the-top services such as video streaming platforms. The transition from 4G to 5G is no different.

In an exclusive interview with Computer Weekly, Ericsson’s APAC CTO Magnus Ewerbring explains why the move to 5G will take place sooner than ever, and gives his take on enterprise applications that will take advantage of 5G services.

How do you see the adoption of 5G playing out across the APAC region?

The majority of 5G adoption will be in four markets – US, China, South Korea and Japan. Korea Telecom, for instance, will launch its 5G network during the Winter Olympics, based on an early version of the 5G standard. This will be followed by China in 2019 and Japan in 2020, in time for the Summer Olympics in Tokyo. From a network and device perspective, we can expect to see launches in the second half of 2019. So far, there has been a lot of interest in 5G from operators in the region. In fact, I met one operator that has not launched 4G and is thinking of going with 5G instead.

These operators won’t wait too long to roll out 5G because of the value that the technology will unleash in enterprise segments such as the industrial internet of things (IoT). Compared to the consumer segment, where early 5G devices are likely to be expensive and volumes will be small, the enterprise segment is less price sensitive, making it easier for operators to justify early investments in 5G.

Besides the IoT, what other potential applications will benefit from 5G?

As with any roll-out of new mobile networks that offer faster connectivity, people will always want to know what applications will take advantage of the new technology and how much money telcos can make. The hard fact is we don’t know, but what we can do is use our wisdom and project into the future.

We see three main areas that will benefit from 5G – consumer mobile broadband, industrial IoT and fixed wireless access. I think the untapped potential is IoT bots that use machine-to-machine communications that are real-time and critical to enterprises. For example, you can remotely control machines and vehicles such as tractors on a 4K display. For that to work, you need high-bandwidth connections with no delays. This is completely different from the usual gas meter-type IoT applications. We are currently working with a mining company that is looking to drive a nearly autonomous truck in its mines located far away for reasons such as safety and efficiency, where one person can control vehicles in several mines.

Another area is healthcare. In China, we’re working with AstraZeneca to connect healthcare devices to ascertain how often they are being used, so AstraZeneca can provide predictive maintenance for those devices. The next step for the healthcare industry would be remote surgery, which some people may not be comfortable with today, but it may well be possible 10 years from now, enabling people to engage the best surgeons around the world.

Some of the 5G applications that you mentioned are already being tested with today’s 4G bandwidth, which is somewhat sufficient. What value would 5G bring then? Also, will having more and more bandwidth discourage further development of technologies such as compression that had been built to overcome bandwidth bottlenecks?

The bandwidth that we have today with 4G – up to 1Gbps – is phenomenal. But consider the launch of the iPhone 10 years ago. It was a 2G phone, but would you like to go back to the first iPhone? Although 5G networks will be launched in a few years’ time, the technology is expected to be widely adopted only in 2026 or 2027. What will the internet look like then? It’s likely that we will laugh at having just 1Gbps. It’s a bit of bootstrapping, where we on the technology side will push the boundaries as far as we can and envision what 5G can be used for. And I think it will be very wrong to say we don’t need more than 1Gbps, because that could be outdated quite soon.

“In future, telcos like Singtel could offer a ‘hospital slice’ with the bandwidth and QoS needed to conduct remote surgeries”

Magnus Ewerbring, Ericsson

What you are alluding to is with more information and data being carried by 5G networks, how can we ensure that real-time, critical applications such as remote surgeries get priority access to bandwidth? Two years ago, we introduced network slicing and we recently did a trial with NTT Docomo in Japan to slice a network based on the required quality of service (QoS).

So, in future, telcos like Singtel could offer a “hospital slice” with the bandwidth and QoS needed to conduct remote surgeries – and charge more for it. They could also offer an “electricity meter” slice that transmits measurements every two minutes. That slice would be more affordable because the QoS requirements are less strict.

Those smart meter applications sound like something that the likes of Sigfox and Lora are already trying to address with their low-power wide area networks.

In my view, they are in a very different arena. When I talk to operators, I think why on earth would they go down that avenue? The load on the network is nothing for an existing operator to add on the ability to serve gas meters. The incremental cost is nothing. It’s just additional revenue. They also have good coverage in homes, which means they can reach the gas meters too. What is often overlooked is that the 3GPP is a large international community that standardises on things like security, privacy and integrity. We can ensure a trusted model, one that can assure people that no one is tapping their phones. As for the likes of Lora, which standard are they relying on?

But those players are saying they can offer cheaper services than telcos are offering.

Sure, but if you are an operator, the cost of implementing NB-IoT (narrowband IoT) is small. Of course, people often say the telcos charge too much, but that’s the name of the game in the industry. If we sober up and look at things from a distance, the telcos will need to find the right price point. Some telcos, such as China Mobile, have done that – just look at the mountains of shared bicycles in Beijing provided by the likes of MoBike, which are using cellular connectivity to power their services.

You mentioned that China is one of the first markets where 5G is expected to take off. Some Chinese telcos are using the TDD (time division duplex) spectrum for their 4G networks. Is there a 5G roadmap for them?

Some years back, we had many discussions on that, but that’s not that important to me. What is important is that operators get lots of spectrum in good bands. These bands should be where there are a lot of operators, so they can get the attention of device manufacturers. Now, if you look at the bands that have been allocated, it has been typically FDD (frequency division duplex), though China has introduced TDD, which China Mobile is using for 4G services. Some European operators have TDD spectrum, but most of them are on FDD spectrum. For 5G, we want to get new spectrum and hundreds of megahertz of it, so we need to get on higher bands, which happen to be TDD. But whether it’s TDD or FDD doesn’t matter because we build products for both frequency bands.

In APAC, some countries have switched off 2G networks and analogue TV broadcasts, or are planning to do so, freeing up bandwidth for next-generation mobile services. What impact would that have on 5G developments?

There are two different wheels spinning. Regardless of 5G, devices are becoming more affordable every year. In markets like Indonesia, a growing percentage of users will be able to afford smartphones as users migrate upwards from 2G to 3G to 4G devices. The allocation of frequency bands will change accordingly. The 5G wheel is spinning as well. While 5G may start with new bands initially, we will also start to see 5G spectrum farmed from existing bands. This will happen earlier in markets that are changing faster, like Singapore, South Korea and Japan.

Read more on Mobile networks