Fünf neue Perspektiven auf 5G und seine Möglichkeiten: Aus dem All, von der Tiefsee und hohen Frequenzen, den Bremsern des Lichts – und schließlich dem erfindungsreichen Nutzer.
- Ungeahnte Anwendungs-Szenarien
Autonomes Fahren, mobile Downloads großer Videos in Sekundenschnelle oder Chirurgie per Fernsteuerung: Es gibt vieles, was wir mit 5G tun wollen. Wofür der neue Mobilfunkstandard tatsächlich genutzt werden wird, sieht aber garantiert noch anders aus – da ist sich Gunnar Menzel sicher. Lesen Sie hier, warum mit überraschenden Use Cases unbedingt zu rechnen ist.
- 5G per Satellit im Hinterland
Alle Welt ist ständig online? Nein – denn nicht einmal die Hälfte hat überhaupt Zugang zum Internet. Wie 5G aus dem All sehr bald Breitband-Internet überall verfügbar machen könnte, erfahren Sie im zweiten Blogartikel von Gunnar Menzel.
- Datenübertragung in Lichtgeschwindigkeit?
Dem steht selbst bei 5G einiges im Weg – ganz ähnlich wie schon im 19. Jahrhundert beim optischen Telegrafen-Netz. Wo weiterhin die Bremsklötze liegen, zeigt Artikel Nummer drei.
- Hohe Frequenzen für hohe Device-Dichte
Zehntausende mit Smartphone im Stadion oder zig mobile Devices pro Stockwerk in der Stadt: Die Kapazitätsgrenze von 4G ist immer schneller überschritten. Wie lange uns der neue Mobilfunkstandard ausreichen wird und was die dichte Verteilung der kleinen 5G-Antennen lokal noch so alles möglich macht, sehen Sie im vierten Artikel von Gunnar Menzel.
- Die Telefonschnur aller Handys
Der interkontinentale Datenaustausch fließt noch immer durch Transatlantik-Kabel auf dem Meeresgrund. Wie dieser Internet-Backbone die Datenübertragungs-Geschwindigkeit von 5G drosseln kann – und was die Tech-Giganten tun, um das Schnellste für sich rauszuholen, erfahren Sie in Artikel Nummer fünf.
1. The element of surprise
So you finally decided to strip that horrific wallpaper off the spare bedroom walls, or build that deck you’ve been promising yourself, or do any one of the other things around the house you have plenty of time to imagine doing but no time to actually do.
You’ve watched the how-to videos, seen it done on home improvement shows, and all seems clear in your head. But in the hardware store, faced with 37 varieties of cross-headed screw, you realize you should have made a shopping list slightly more sophisticated than: “wood, screws, some kind of varnish.”
The element of surprise
In August 2019, Google revealed what it called “surprising” internet behavior (it takes a lot to surprise Google about internet behavior). In a survey of 24,000 people in 10 countries, more than half reported watching online video in stores as they were shopping. Fifty-five percent of people weren’t relying on scribbled shopping lists or memory at all, they were re-watching the how-to video on their mobile and picking what they needed off the shelf.
“I’m not a list guy,” said one respondent, “I’ll re-watch the video in the plumbing or electrical aisle to make sure I’ve got everything I need.” And, ten thousand product-placement marketers began trembling with excitement.
No one knows how long people have been doing this, but it can only have become possible as mobile connectivity improved. If they had to wait minutes to connect and initiate a video, people probably wouldn’t bother. What’s fascinating is that this use case for online video was invented by consumers – nobody thought to make videos specifically to be watched in store.
Expect the unexpected
One of the great expectations for 5G is that its low latency (sub 10 ms) will enable a series of applications that are currently roadblocked by 4G’s limitations. Crystal ball-gazing articles on the subject tend to agree on the list of upcoming miracles.
True self-driving vehicles will finally become a reality because they will be able to network with each other over low latency 5G, say the prophets. It will certainly help, but the real problem with self-driving is that AI just isn’t smart enough yet. We’ll be able to download 4K movies in seconds. Fine, but so what, you can’t watch a 4K movie in seconds. Remote surgery will be possible. Maybe, but it’s hardly going to be a mass market, and 5G doesn’t speed up the long-distance internet backbone anyway.
All these things may or may not come true, but what’s really exciting about 5G latency are the applications we haven’t thought of yet. The kind of use cases that surprise Google. One thing you can be sure of is that people will find wonderful, and horrible, ways to exploit 5G’s capabilities.
Fundamentally, 5G will dramatically increase the flow of data and, therefore, the potential value that can be derived from that data. In this environment, the need to be flexible, agile, and open to new ideas will be more critical to business success than ever.
2. In space, no one can hear your latency
There’s a gentle wind in the trees, a breathtaking view across the water, and your legs are satisfyingly tired from a long hike. A rare moment of peace in a natural setting. Out of curiosity, you check if your phone has a signal. Zero bars. No chance of a distracting email or SMS. You relax a little more.
We’re so used to always-on connectivity that an entire industry exists to help people spend less time online. You can’t scroll social media for more than a minute without seeing an article about switching off your phone and finding a new purpose in life. It’s easy to forget that for most of the world, the kind of always-on connectivity we dream of escaping is itself a distant dream.
A disconnected world
The UN expects half the global population to be connected to the internet by the end of 2019. They’ve been working hard to help developing countries reach this goal, but even if it is achieved (the same target was missed in 2017), that still leaves 3.8 billion people not connected. Worse, the UN defines “being connected” as using the internet once in a three-month period – hardly the kind of access most in the developed world are used to.
Current technologies are most effective at providing connectivity in high-density urban environments, but about half the world’s population don’t live in urban areas. Several enterprises are working on revolutionizing broadband connectivity by putting a core network into space.
Watch this space
In May 2019, SpaceX launched 60 satellites on one rocket as the first stage of its Starlink space-based broadband network. With characteristic Musk understatement, they marched across the night sky in a startling “satellite train”. announcing the arrival of a bold new enterprise.
SpaceX plans to launch up to 12,000 more Starlink satellites by the mid 2020s, creating a global core network that can deliver 5G anywhere. Other companies, including OneWeb and Amazon, are planning their own mega satellite constellations with the same aim.
Unlike traditional satellite communications networks, which rely on a few large satellites in high orbits (about 35,000 km), the new approach is to deploy a very large number of small satellites in low orbits (340–1,150 km). The main reason for this is to minimize latency. Even at light speed, it takes noticeable time for a signal to travel 35,000 km into space and back. In contrast the Starlink network is aiming for latency of 10–20ms.
Unlike its predecessors, 5G is being designed to work with space-based networks. The global 5G standards coalition, 3rd Generation Partnership Project (3GPP), is incorporating interoperability with satellite technology to ensure it is flexible enough to take advantage of the network of networks expected to development in global telecoms infrastructure.
If you think the internet is big now…
Right now, a lot less than half the world is genuinely online. The arrival of Starlink and similar networks will improve the options for connecting the rest of the world, but it’s not quite as simple as that.
These space-based networks are essentially an alternative to the existing internet backbone. Like the ground-based backbone, Starlink won’t be able to connect with edge devices directly, local nodes will still be required and SpaceX plans to build up to a million ground stations in this role. None of this will be cheap, so it remains to be seen how those costs mesh with the idea of connecting the world’s unconnected, who are also the world’s poorest.
Assuming Starlink and its competitors do bring significantly more people online, what will that mean? The value of the internet has always been in its reach. Commercially, the more people can access your services, the greater your potential revenue. With millions, and perhaps billions, more people online in a relatively short period of time (Starlink is due to be fully operational by the close of the 2020s) the potential for growth and diversification is staggering.
3. The speed trap
Financiers François and Louis Blanc stood trial for manipulating France’s stock markets by piggybacking data on a superfast, classified comms network built by the French government. Don’t recall the story? Perhaps because it happened in 1837.
The hot technology of the time was the optical telegraph – a cross-country chain of big towers within line of sight that used moving signal arms to pass coded messages. The system could transmit information across the country in minutes – thousands of times faster than messengers on horseback. It was a latency revolution.
The Blanc brothers’ hack was the oldest in the book – social engineering. They bribed telegraph operators to add coded information about stock movements in Paris to the messages they were sending. Receiving this information in Bordeaux days ahead of competitors, they cleaned up.
And yes, they totally got away with it. There was no actual law against using the system for personal communications, so the case was dismissed.
Moving fast and breaking things in Napoleonic France
The networks of optical telegraph towers that stretched from Moscow to London in the early 19th century sound absurd to us today, but they were the culmination of an ancient tech tree. Beacon fires on hilltops had been transmitting messages across empires and kingdoms since prehistory.
Invented in Napoleonic France in the 1790s, optical telegraphy improved on beacons in two ways. First, operators watched neighboring towers through telescopes rather than with the naked eye, greatly increasing the viable distance between the network’s nodes. Second, the moving signal arms on top of each tower were able to encode arbitrary text, unlike beacon fires which are simply on or off.
There is almost no cultural memory of optical telegraphy today, despite the size and complexity of the networks built during its heyday, perhaps because the technology was completely eclipsed by the electric telegraph within a few decades. Fast-moving disruptive tech had a long history when Silicon Valley was still three sheds and a donkey.
Wait and c
The strength of optical telegraphy, and beacons, is that the signal moves at light speed across the majority of the network’s geographical reach. The routing and repeater nodes (the towers operated by humans) are what slow everything down.
It’s a problem that persists today. The global internet backbone operates at light speed, as do radio signals between local nodes and end devices, but when you click a link it doesn’t always feel like the data bounces back at 300,000,000 m/s. It’s the nodes that route and convert signals from electrical to radio to optical and back again that slow things down. And it’s here in “last-mile” latency that 5G makes a difference.
But if last-mile latency dramatically improves, the speed of the backbone itself can become an issue. Light speed is extremely fast, but it’s not infinite. Even in ideal conditions (light speed in a vacuum) the round trip from New York to Sydney takes over 100 ms. Add in real-world annoyances, like the reduced speed of light in optical fiber and the fact that no network operates in a straight line, and actual latency on this roundtrip is probably between 200 and 300 ms. That’s a noticeable delay when last-mile latency is 10 ms or less.
On the edge
5G means that, for the first time, last-mile latency will often be less than backbone latency. This is a problem for any organization serving a globally distributed customer base. If your data center is a long way from lots of your customers, your quality of service will be poorer (i.e. noticeably slower) than competitors with physically closer data centers.
The potential answer to this problem has been around for a while – edge and fog computing. These may finally come into their own as last-mile latency drops and the sheer volume of data from the IoT skyrockets.
Fog can be thought of as a fabric that connects edge devices to the public cloud. But it’s not a passive conduit. Woven into this fabric is sufficient processing power to determine which parts of the tsunami of IoT data needs to be sent all the way to the cloud, and which parts can stay local. In very simple terms, it means data that is only likely to be needed locally stays local, where it can be accessed much more quickly.
No matter how delightful it is at first, consumers will quickly become used to the speed of 5G, and demand that all services keep up. At that point, the major challenge of building an effective fog network will become an urgent one for many global organizations.
If you’re wondering what happened to the Blanc brothers, they went into the gambling business and eventually opened a casino in a little-known holiday resort called Monte Carlo, figuring it made sense to take their services close to where people wanted them.
4. Know your place
The final whistle blows and the roar of the crowd fades. Around the stadium people are shifting in their seats, collecting bags and jackets. Some are already heading for the exits. You pull out your phone to check train times, or to fire up a ride-hail app.
Thousands of people around you are doing the same thing, and thousands more in the office buildings across the road, and in the train station around the corner, and the local 4G network simply can’t cope with that many concurrent connections.
4G can handle 4,000 individual connected devices per square kilometer, which sounds like a lot but quickly gets stretched to the limit when that square kilometer is an urban area with hundreds of households or offices, each with multiple connected devices, plus thousand more people passing through.
Two’s company, 28.5 billion is a crowd
5G will be able to handle a million concurrent connections per square kilometer. Not many events bring that many people together in one place, and even the world’s most densely populated cities (Manila is the current record holder at 41,515 people per km2) don’t come close.
Of course, the real problem is that one person doesn’t equal one device. Today, a third of households in the US have 10 or more connected devices. And the IoT revolution has barely begun. Cisco expects there to be 28.5 billion networked devices by 2022. That’s three-and-a-half devices per person on Earth. It’s also a staggering increase of around nine million devices per day, every day, between now and 2022.
Without the dramatic improvements in concurrency and bandwidth that 5G promises, that explosion of IoT devices simply won’t be possible. But one of the technical features of 5G that will allow these improvements will have another, often overlooked, impact.
Location, location, location
Almost nobody talks about 5G’s potential to revolutionize location sensing, which is surprising given how many online services are location sensitive. Ride-hailing, navigation, local search, delivery services – many apps we use day-to-day rely on knowing where your device is on the surface of the Earth. Most of these use Global Navigation Satellite Systems (GNSS), such as GPS – a technology that hasn’t fundamentally changed in 40 years.
5G’s pumped-up speed and bandwidth are partly due to the use of much higher radio frequencies. The downside is that 5G signals will have a much shorter range than lower-frequency 4G signals, which will mean many more cellular towers will be needed for the same coverage.
These will be much smaller than the cell towers we are used to – more like small pizza boxes – but there is no escaping the need for many more of them, dotted on the sides of buildings, telegraph poles, roofs, etcetera. These towers will also be more sophisticated, able to detect the direction of incoming signals and to directionally “beam” outgoing signals.
More cell towers mean your device will be much closer to the nearest one and often in line of sight of several. Measuring the direction and time taken for signals to travel between devices and cell towers will enable very accurate geo-spatial positioning. And none of this will involve consulting venerable satellites in high Earth orbit, or the latency of a 20,000+ km signal travel distance that involves.
The 5G standards group, 3GPP, has agreed a large number of location and context-awareness KPIs, including target accuracies, for use cases ranging from emergency response to push advertising.
Know your place
Location sensing was always possible with 4G, but not very accurate. A 4G network knows where your device is to within about a mile. Because this was useful for emergency services, additional technologies have evolved to tighten this radius, but GNSS remain the only really accurate geo-spatial location option.
GNSS has some major limitations though, the most notorious being that it doesn’t work well, or sometimes at all, if you can’t see the sky. That’s a bit like a messaging app that only works if you can see the person you’re messaging. GNSS apps are also very power hungry. 5G promises to overcome these weaknesses, delivering always-on, extremely accurate location sensing with trivial latency.
Back in your seat at the stadium, not only will you get a superfast connection at the same time as thousands of other people, your ride-hailing service will be able to tell where in the stadium you are, how long it will take you to get to the nearest exit, and maybe even suggest a ride share with the guy two rows back who’s going in the same direction.
Read more about the value 5G can bring to industry.
5. A sea change
In the earliest days of mobile phones, when they were like bricks and everyone still had a landline, people joked that any phone could be mobile if you had a long enough cord. It’s more than a little ironic that today everyone has a mobile, but global communications depend on cords thousands of kilometers long slung across ocean floors.
Submarine communications cables carry over 90 percent of data traffic between continents. They are the core of the internet backbone. Even if 5G meets its most optimistic latency and bandwidth goals, it will mean nothing if the backbone capacity isn’t there to move data globally. You can’t receive a 5G signal across the Atlantic no matter how much you wave your phone in the air. This is why there has been a surge in submarine comms cable projects over the past decade, one that has gone almost unnoticed amid the 5G buzz.
The Apollo program of the 19th century
Installing a physical cable across an ocean sounds like something that could only be possible with cutting-edge materials and engineering. In fact, it was first achieved 160 years ago with seven strands of copper wire wrapped in hardened tree sap.
In 1858 a British and a US warship met in the mid-Atlantic, spliced together two ends of the hundreds of tons of cable they each had on board and set off in opposite directions for home, paying the cable out behind them. Both ships made landfall a few weeks later and a 3,200 km-long telegraph line between Europe and the North America was open.
The cable failed within a month, but it was a world-shaking technological achievement and proved a point. In 1866 a second attempt using improved cable (but still essentially copper wire wrapped in tree sap!) established a permanent connection.
You know how annoying it is when an ethernet cable slips down the back of your desk and you have to fumble around on the floor to find it? Now imagine you drop a cable in a 4 km-deep ocean and you have to retrieve it with a grappling hook from the deck of a pitching steam ship. That’s what engineers did in 1866 to recover a snapped cable and complete the second trans-Atlantic link.
Grow a backbone
Modern submarine cables consist of optical fibers sheathed in multiple protective layers, usually no more than 25 mm in diameter. The longest, the SEA-ME-WE 3, stretches over 39,000 km from Germany to Korea. In some places they lie on the ocean floor under 8 km of water.
Unsurprisingly, these cables are expensive to make and install. For much of their history they were funded by governments and consortia of telecoms companies. In the last few years, however, this model has changed radically, with hyperscale cloud companies and large content providers now driving the biggest projects.
Google, Facebook, Microsoft, and Amazon have made major investments in submarine cable in the last five years, with much more planned. Google, for example, is currently building two across the Atlantic, another between the US and Chile, and a fourth between Hong Kong and Guam.
The technology that pushes data through the fiber is also bounding ahead. Google’s Dunant cable, stretching 6,400 km from North Virginia to the French coast, expects to achieve transmit speeds of 250 Terabits per second when it comes online in 2020. Contrast this with the 16 hours it took to send a 98-word telegram from Queen Victoria to US President James Buchannan over the 1858 cable.
When you have BigTech money, building your own internet backbone makes a lot of sense. Rather than rent capacity on connections built by telcos, these companies can route cables to minimize distance (and latency) between their own data centers. Competition from currently embryonic space-based networks, like SpaceX’s Starlink project, may make milliseconds critical.
It’s not what you know, it’s when you know
Considering the implications of 5G without thinking about the backbone network is like trying to understand cars without thinking about roads – one is meaningless without the other. Unfortunately, this doesn’t mean it’s uncommon.
The effective ban on US companies using 5G kit made by Chinese company Huawei is one example. Whether the security concerns are justified or not, the idea of a US-only 5G network shining on a hilltop is a fantasy. Huawei is a massive investor in submarine cables. It may be banned from building or operating these cables from US soil, but data from US users will inevitably travel over Huawei-owned cables (and other networks) elsewhere in the world.
The speed and bandwidth of 5G on the edge of these networks will drive speed and capacity in the backbone (developments in fog computing not withstanding). Like flowing water, data will quickly find a way around political considerations.
The 1858 submarine cable failed in less than a month, but not before the British government had used it to send a message that saved £50,000 (a huge sum in the mid-19th century). At the outbreak of the Indian Mutiny the previous year, it had taken 40 days to receive a message in London requesting troops. When the uprising ended, it took a few hours for the British to send a trans-Atlantic telegram countermanding the very expensive embarkation of more reinforcements stationed in Canada. The value of data that enables individuals, enterprises, and governments to respond to distant events has driven technology for generations, and 5G will only sharpen that imperative.
Read more about the value 5G will bring to industry.