What is 5G for? What replaces the smartphone? What is the next big thing? Perhaps these questions don’t have singular answers.Read More
Yesterday I played with tomorrow’s phones. Things like that happen as a futurist.
I went to see a company called Qualcomm, a name you may not know today but that will likely be as familiar as Intel in years to come. Qualcomm has been a key player in the mobile industry pretty much since its inception — certainly as far back as my direct experience goes (2000). Only now is it considering communicating more with the public, rather than with the Samsungs, and LGs of the world who rely on its technology.
The product I was looking at yesterday is the latest iteration of its mobile phone platform, the Snapdragon 800. In a single package this provides all of the major components you need to build a smartphone or tablet: networking, processing, video, and voice. Stick a screen, a camera, a case, a battery, ports and an aerial on it and you pretty much have a phone: everything else is software.
One thing that caught my attention yesterday was the chip’s ability to both capture and display 4K video. If you’re not familiar with 4K, it’s four times the resolution of HD. That means super-sharp, super-rich images with lots of depth too them, even without any 3D technology. 4K TVs are already on the market, albeit wallet-crushingly expensive.
The device can also capture and play back surround sound. Add amp and TV and you have a very high-end home cinema system, as was demonstrated with exclusive clips of Pacific Rim (prepared for the recent E3 show — yes, I was geeking out).
All very cool.
Stuff Drives Storage
Another meeting I had yesterday was with Richard Lee, president and CEO of QNAP Systems, a company I have written about here before. QNAP makes network attached storage (NAS) devices. These are small, efficient computers with lots of hard disk capacity designed to be a repository for all your digital goodies: music, photos, videos, documents etc.
Richard is delighted that the next generation of phones will have better cameras. Why? Because more megapixels means more megabytes are required to store your digital creations. The incredible amount of content that we are consumers are creating is what is driving the NAS market, predicted to grow at over 20% a year.
Richard pointed out though that it is not just the ability to capture content that is driving the desire for storage. It is the ability to share it; the availability of bandwidth needed to stream new photos and videos back to a central storage hub, access them across a variety of devices and share them with friends and family. We now have cameras with us all the time, and these cameras can — and do — instantly stream the images we capture off to our personal clouds.
And here is where it gets tricky…
Every time a new generation of mobile network comes along, everyone gets very excited about the speed of the connection between the phone and the mast. 2.5G, 3G, 4G — the conversation is always the same. Likewise with the advent of new home broadband services: you always hear about how many megabits per second your cable or DSL line will deliver from the exchange to your home.
What you never hear about is the link between the exchange and the rest of the world — what is known as ‘backhaul’. This is where it all starts to fall down.
In February this year, Tellabs — an old client of mine — sponsored some research that identified a $9.2bninvestment gap in backhaul networks. Based on a five to six times increase in the volume of data carried over mobile networks, at current investment levels, operators will be 16 petabytes short of capacity.
That’s quite a lot. Despite Tellabs clear interest in there being an investment gap (the company provides — surprise surprise — backhaul network gear), I don’t have trouble believing these figures. If anything they feel intuitively a little conservative: we know that there are bottlenecks in the UK’s mobile networks today, and data consumption is increasing incredibly fast.
Crunch, Crunch, Crunch
When demand for data outstrips supply capacity like this it has become known in the telecoms industry as a capacity crunch. Backhaul looks likely to be the latest in a series of crunches that have happened at different points throughout the communications networks over the years. There will be more: there are already rumours about a coming crunch in fixed-line (i.e. your home and business broadband) networks in the near future.
It’s not surprising that networks don’t scale evenly to meet demand. Nor that there’s sometimes a lag between demand and the supply to meet it. This is the nature of the market, particularly in an industry where demand is scaling so rapidly.
The alternative is to build based on forecast capacity. But we’ve been through that before too: it happened across the world during the dotcom boom. One client of mine bought and built itself a global network on the basis of forecasts of demand for internet services. At its height in 2000 the company was valued at $37billion. By the end of the following year it had been through Chapter 11, a form of voluntary bankruptcy, and acquired by Cable & Wireless for just $800m.
The company was called Exodus Communications. No-one wants to risk being the next Exodus.
What does this sentence mean to you?
‘This year the UK government will auction off 250MHz of radio spectrum to be used for 4G mobile phone services’.
If the answer is ‘Nothing’ then let me try to explain, because this stuff is important.
Think of a child’s drawing of the world. Land at the bottom, sky at the top, people and houses in the middle.
Now imagine that the sky is divided into 300 million very, very thin horizontal layers stacking up into space. These layers are grouped together into bands — imagine them as different colours like a rainbow. Some are thinner, some thicker.
This rainbow of colours represents the radio wave spectrum. Each colour represents a group of radio frequencies (a ‘band’) we have allocated to different uses. Some bands carry AM or FM radio, some the signals for 3G mobile phones, and some your Wi-Fi network.
Frequencies are measured in Hertz (Hz for short) which represents the number of times the wave oscillates (goes up and down) every second. The frequency has a huge impact on how useful a band may be for different things. High frequencies can carry lots of information but need lots of energy and don’t travel very far. Low frequencies aren’t so good at carrying information but will amble on endlessly over very great distances.
Most modern digital services exist in a sweet spot between 10 Megahertz (10,000 oscillations per second) and 10 Gigahertz (10 million oscillations per second). Here you get a reasonable balance between range, energy consumption and information capacity. Though there’s still a marked difference between the characteristics at the top and bottom of this range.
Each band is a different size, depending on how many layers (Hertz) the relevant governing body has decided to allocate to each use. Bands are typically allocated by the Megahertz (thousand Hertz). Some might have just a few tens of thousands (e.g. Wi-Fi), some might have a few hundred thousand (3G).
My suggested visualisation is by no means scientifically accurate but it does get across one important thing: the amount of useful spectrum available is finite — so we need to make good use of it. Just as we do any other resource.
Back to the beginning
So back to my opening: the UK government plans to auction off 250MHz of spectrum next year to be shared between up to four different operators — e.g. Vodafone, 02, EE or Three.
Now let’s do some maths.
Take your 250MHz and multiply it by the amount of data that can be carried on each single layer/Hertz. This is measured in bits per Hertz per second, and the ideal world figure for LTE — the 4G standard to be used in the UK — is around 8 bits per second per Hertz. That means that in any one place at any one time, the total data capacity of 4G is around 2 million bits per second or 2 Gigabits per second.
Let’s put that into real world terms. Netflix reckons you need the following in order to stream its content:
- 1 Megabit per second to watch standard definition TV on a laptop-sized screen
- 2 Megabits per second to watch standard definition video on a large TV
- 4 Megabits per second to watch HD video
On this basis 500 people in Oxford Circus or Exchange Square in Manchester could all watch Strictly in high definition at the same time over 4G. That doesn’t sound too bad.
But the real world and the ideal world are very different. According to studies by Ofcom and others a more realistic figure is 1.3 bits per second per Hertz. That means instead of 500 people happily streaming Strictly, you’re down to 80.
Some people are more pessimistic still. That number could be significantly smaller once you take into account the impact of handling different types of traffic and real, real world conditions. Suddenly you’re down to just fifteen people on each network being able to watch Strictly in all its HD glory.*
But I don’t care about watching Strictly in HD!?
Strictly is just a useful metaphor. 4 Megabits per second is not, in the grand scheme, a lot of bandwidth to demand.
The first portable devices with 4K or Ultra High Definition screens are already starting to appear. These will devour four times the bandwidth of full HD. Every time more bandwidth has become available, we have created services to fill it. Trust me: 4Mbps will feel like dial-up in a few years (and if you don’t know what that feels like, and you really want to experience the web the way it used to be, check this out.
So what’s the answer?
Now think: how many of the services on here could not be delivered over an internet connection? For a start, all broadcast services, and all mobile services could be. This doesn’t determine what you use to carry that internet connection (e.g. 3G), but it does mean you don’t need to allocate chunks of spectrum to specific services. Rather you just allocate spectrum to different means of delivering an internet service.
Because technology will move forward constantly you don’t want to keep reallocating the same spectrum every time a new standard comes along. But maybe break it into two or three blocks with a rolling programme of upgrades, each time a significant new milestone is reached. This way we can ensure there is always enough bandwidth for us all to watch Strictly — and do a lot more valuable (and enjoyable) things, whenever we want to and wherever we are.
*Can I just point out at this point that I am not a particular fan of Strictly Come Dancing — it just happened to be the first programme that popped into my head.
The telecommunications industry was very slow to understand the value in the data it moved, not just in its volumeRead More
Second Life is just a starting point. Imagine you could dip in and out of a virtual world at will, using your mobile device.Read More