Questions and answers about
the economy.

Can the UK’s digital infrastructure continue to cope with rising demand?

Home-working and remote social interactions during the pandemic have relied on high-quality infrastructure for digital communication – and despite a huge increase in demand, the network has largely delivered. The next challenge is to ensure universal access at moderate cost.

Throughout the Covid-19 crisis, we have relied on digital platforms for everything from home-schooling and remote working to entertainment and contact with loved ones. While very few positives have emerged from the pandemic, the way the UK’s digital communications infrastructure has performed – in the face of rapidly increasing demand – has been one of them.

The rise in demand has been significant – use of the internet doubled in 2020 and adults in the UK are now spending more than 25% of their waking day online.

As our reliance on the internet and our mobile phones has grown, the effective functioning of our digital infrastructure has become ever more vital. While the official digital economy – made up of information technology and communications firms – accounts for only a small fraction of the total, almost all businesses now depend on digital for their operation.

Overall, in spite of increased demand, the infrastructure has maintained a good level of service and has been mostly resilient to failures. Many services have significantly improved in recent years – for example, 69% of residential broadband lines are now superfast, an increase of 10 percentage points since 2018.

While stories of the odd breakdown or Radio 4’s chronic inability to set up good lines may persist – the efforts of a lot of engineers have mostly worked. I have not used cash for months or seen TV news in decades, in spite of more time than ever on a screen – digital money carried on the internet and digital content have taken over. A third of us now watch online video content more than traditional TV, so it is perhaps unsurprising that subscription services accounted for more than half of online video revenue in 2019, at £1.7 billion.

Digital infrastructure does not just enable us to work remotely or provide us with entertainment: it has facilitated changes across our daily lives – ordering milk, checking that birdsong or identifying a flower can all be done through our devices and take seconds.

Keeping us connected like this takes serious infrastructure. Often, this is invisible and under-appreciated, but it is arguably the most important infrastructure we have. We need many different types of infrastructure to live and work, such as utilities and roads, but most deliver much the same service as they have for many years. But communications infrastructure delivers literally millions of times the data it did 30 years ago – it is almost impossible to imagine what roads that did this would look like.

According to the latest survey by Ofcom (the UK’s regulator of the communications sector), communications infrastructure delivers more than 300 GBytes per month per user (this is what is actually delivered, not what could be). This is around 20 times higher than the corresponding value for 2011.

These personalised signals now probably add up to more data than is delivered all the TVs in the UK from traditional broadcasts, even counting each TV separately (even though they are showing the same signal – so Netflix is not so big a change as you imagine).

Of these data, 99% is delivered by the fixed line network and only about 1% by mobile. While this may seem surprising, it is because most of the data on all handsets comes via WiFi from the fixed network, even if to a mobile phone. Both mobile and fixed data volumes are rising by orders of magnitude each decade and this will continue for some time.

What does the network actually look like?

The fixed core network that connects everything (including WiFi and mobile base stations) is overwhelmingly made of silica optical fibre. The ‘access’ network, which is the last kilometre or so to your home, is likely still to be a copper wire pair, originally installed even up to a hundred years ago for landlines, or analogue voice telephony. This has been effective far beyond what was originally envisaged, but these copper wires are now near their fundamental limits and will be replaced by fibre within a decade.

Fibre is currently being used at a tiny fraction of its ultimate capacity – perhaps a millionth or less. This means that this infrastructure will most likely last another hundred years or so. Further, fibre is generally more reliable, for example, it does not corrode when wet.

The last link in the network will often be wireless and this is also improving and getting shorter. Many things from WiFi6 to 5/6G and new technologies that direct signals to each user will help here, and incidentally reduce energy usage. It is notable that communications technology has not only delivered far more capacity for about the same energy, but it has even reduced the energy used. Fibre optics, for example, is much less energy consumptive than copper, which is partly why data centres use it internally.

Operators tend to point out that the telecommunications industry has by and large seen little benefit from the vastly increased volume of products it carries, let alone from its increasing significance. Revenues have not risen with goods carried – but this is a worldwide problem without an easy answer. The suppliers of end terminals, such as smartphones, have done much better.

How is the network changing?

The network consists of much besides the basic transport fibre and wireless, and exciting developments are taking place in this area. The data routing and switching functions that are vital to all communications need very fast and expensive kit.

Many years ago, this changed from simple switches that connected the ends to ‘packet’ switching where all data travelled in small packets with an address label at the front – making all digital services rather similar, regardless of the content (whether video, voice, internet and so on). This was a lot more efficient, but the next stage is to put the functionality into generic standardised hardware like super-PCs with the clever stuff in software, instead of expensive single-purpose devices. This is much harder than it sounds but may enable equipment from many suppliers to be mixed and matched in ‘open disaggregated networks’, reducing costs and increasing resilience.

And the actual physical transport that carries the signal – these days normally a modulated light wave – is still changing. New ‘hollow core’ fibres guide light in a central hole ‘nudged’ into place with 100 nanometer (1/1,000th of a human hair) membranes. This largely gets rid of the limits imposed by the solid glass core normally used today, allowing more capacity, higher speeds (that is lower latency) and probably lower loss (and thus longer range). Radio antennas are evolving too, so that signals can be ‘pointed’ at the intended receiver rather than just sprayed about – improving quality as well as reducing power consumption.

What will this be used for? Experience suggests that predictions here are always wrong. For example, no one anticipated that mobile phones would be much used for video, whereas this is now is the dominant type of data carried on them. But once the capacity is there, people will use it.

And it is not hard to suggest things that could be useful to communicate that currently cannot be done easily. For example, just connecting the huge number of car and bicycle dash cams to each other, which could be useful in seeing further ahead, would saturate any envisaged network.

What should we care about in a future network?

It is important that capacity isn’t the only focus when considering the future of our digital infrastructure. There are existing ways of improving this.

We should care about coverage – many people do not even realise that things they use all the time, like speech recognition and navigation or the need to call for help (often now automatically and using digital techniques) – need communications technology. We are increasingly stuck if we are even briefly without it, so it needs to be ubiquitous.

And digital technology helps in less obvious ways too. For example, the vaccines being developed for Covid-19 are not made using the actual virus; they based on the viral DNA sequence transmitted over the world-wide network.

While connection speeds are getting faster, 13% of households are not online, which may have had serious implications for children in these households who needed to join lessons remotely this year.

Further, we need low latency (no delay) for many things, including anything that uses remote processing (like speech), not just games. And all this needs to be seamless from a user perspective.

Can we do all this for moderate cost? Well yes, we can – and it will transform our lives even more than it already has been. But adequate funding to ensure universal access will be crucial.

Where can I find out more?

Who are experts on this question?

Author: Will Stewart
Photo by Mika Baumeister on Unsplash
Recent Questions
View all articles
Do you have a question surrounding any of these topics? Or are you an economist and have an answer?
Ask a Question
Submit Evidence