Article - Issue 41, December 2009

Wireless Communications - to 2020 and beyond

Professor William Webb FREng

Download the article (353 KB)

Professor William Webb FREng, Head of Research at Ofcom, made a series of predictions 10 years ago for the future of wireless communications. Now, as we enter a new decade, many of the developments he forecast have come about but others have been delayed or not taken up at all. He tells Ingenia why second-guessing the possible innovations in this area is important and what trends we can expect to see in the next 10 years and beyond.

Launched in autumn 2009, this digital radio streams WiFi. It has a touchscreen keyboard and social networking plug-in applications including Twitter and Facebook. It screens slide shows from radio stations, and can provide newsfeeds and weather reports to view © PURE

Launched in autumn 2009, this digital radio streams WiFi. It has a touchscreen keyboard and social networking plug-in applications including Twitter and Facebook. It screens slide shows from radio stations, and can provide newsfeeds and weather reports to view © PURE

Predictions are essential in many industries; they are the basis on which investment decisions are made and business cases built. They influence the formation of new companies and the directions of major ones. This is particularly the case in areas seen as fast-moving such as telecommunications, where making the right call about a future trend could be the difference between a company succeeding or failing.

Predictions in the telecommunications industry are seen as particularly difficult to make because the technology, and the industry itself, appear to change so rapidly. In the year 2000, in an effort to show that clear and accurate predictions were possible, I authored a book entitled The Future of Wireless Communications.

After a detailed look at the key drivers, the book made specific predictions on how the world of wireless communications would look in 2005, 2010, 2015 and 2020. The predictions in this book and a second one titled Wireless Communications: The Future (published in 2006) were made taking a range of factors into account, including network modelling to understand the capacity and costs of underlying communications networks, an understanding of factors such as Moore’s Law (see panel opposite), a look back of up to 20 years to understand what had actually happened and where predictions had been in error, assessment of standards currently being developed, input from six ‘gurus’ as to their views and predictions from experts in specific areas such as screen technology. As we enter a new decade it is timely to review how accurate my predictions have been so far and to ask whether we should refine our vision for wireless communications over the next decade, to 2020 and beyond.

Predictions FROM 2000

Back in 2000, I predicted that by 2010 homes would become wireless hotspots, with dedicated home networks likely to be using Bluetooth technology. Most homes would be connected via broadband sufficiently fast to deliver high definition TV. Mobile phones, which I suggested might be better termed ‘communicator devices’ by 2010, would have a wide range of functionality and would work with these home networks. Allowing phones to work in the home and office would bring a need for work-life differentiation which would be solved by redirection functions in the network that reformatted, forwarded or sent incoming messages to message boxes according to circumstances and preferences. I thought back then that by now video communications would be making up around 2% of calls.

I predicted that people would communicate more readily with machines using speech recognition and machines would increasingly communicate with each other. Networks would have packet-based core networks and public wireless internet hot spots would become commonplace.

Most of these ideas have come about. Homes have deployed wireless nodes, albeit WiFi rather than Bluetooth. Mobile phone handsets can work on in-home WiFi networks – although this functionality is not yet widely deployed, more because it has not been in the commercial interests of cellular operators to see this happen rather than due to any underlying technical reason. Broadband connections are available to most homes and are close to providing high definition TV capabilities; where not available, recent Government initiatives suggest that public funding will be used to address this ‘digital divide’ over the next five years or so. Public W-LAN (Wireless Local Area Network) hotspots are now ubiquitous in city buildings.

Some of my predictions proved over-optimistic. ‘Redirection functions’, where messages would be intelligently re-routed, have not occurred because they have proven overly complex to build; instead users have made use of different communications mechanisms to handle different message priorities. The prediction that speech recognition would be perfected by 2010 and would become the principal way for people to interact with machines, rather than using keyboards, appeared entirely plausible in 2000 given the rapid progress that had been made in the late 1990s. But this progress slowed and it proved very difficult to get accuracy up from around the 98% mark to the near 100% level needed to make it useful. Speech recognition is still making slow progress and is finding more applications, especially where the vocabulary it needs to recognise can be limited – for example when checking a train timetable – but it may be another
10 years before it comes into widespread usage.

Video communications are more widely used, from the basic (but free) Skype services to the advanced video conferencing suites pioneered by Cisco Systems. However, it has proven to be of very little interest to mobile phone users. Nevertheless, the basic premise that video communications would grow but would still be a minor part of overall communications was about right.

While these specific predictions do not seem far off, and certainly suggest that predictions 10 years into the future are possible, more telling was the day-in-the-life vignette I created to illustrate what new technology applications might be available for the modern executive on his or her way to work. Their electronic communicator would control the home heating and hot water and would download journey details to a car satnav. Personalised news would be delivered to a car audio and emails would be automatically read as they drove – none of these developments is widely available yet although other predictions were more successful, such as the widespread use of wireless hotspots in airports and hotels and having limited email facilities when flying.

More than guesswork

So what lessons can we learn from the past? The first point of note from comparing prediction with reality is that the technology and networks that underlie the world of wireless communications change much more slowly than is widely assumed. Where predictions were in error it was always due to predicting that change would happen more quickly than it transpired.

The second point of note is that the technological predictions were generally the most accurate. Predictions of data rates and device capabilities were rarely in error. The spread of technology, in terms of the number of devices in homes, chips in handsets and so on were also generally accurate. With many laws of progression such as the well-known Moore’s Law and laws specific to telecommunications such as Cooper’s Law (see panel below) of voice capacity as well as observations I made on consistent progressing in areas such as storage capabilities this area of prediction is relatively easy.

My predictions of services or functionality available to end-users were almost always too optimistic. Home automation of any sort appears stubbornly slow, although perhaps energy efficiency concerns might drive some changes in the future. Even more in error were my predictions of ‘personalised’ services where mobile devices made intelligent decisions on behalf of their owners, covering features from message processing to automatic re-routing of travel plans. Further understanding of the social difficulties involved in the acceptance of such ideas now suggests these products may never emerge, with users instead finding specific solutions to specific needs, such as applications or widgets that perform functions like accessing train timetables. The safest bet when predicting how people will take up new wireless technologies, particularly those which involve the devices exhibiting greater intelligence, is to expect little.

Future Implications

As we enter a new decade, there are many exciting developments that surround the area of wireless communications. Here are a few that I suggested back in 2006 and some additional ideas that should have a good chance of becoming reality.

On a larger scale, I still believe that telephone networks will gradually be upgraded to fibre to the home enabling almost unlimited applications with much of the deployment being completed by 2021. I see little change in wide-area cellular networks, except that smaller cells will be used over most of the UK by 2016, with a consequent increase in capacity and small improvement in coverage. I do foresee a huge increase in WiFi and indoor wireless coverage between 2011 and 2016, with wireless increasingly embedded in many standard appliances from electricity meters to home security systems and washing machines. This would then unlock a rapid growth in home automation and networking by 2021.

However, as we saw above, just because an application is possible does not mean that it will be taken up by the market. It is clear that most predictions that involve changes in behaviour are overly optimistic as to the length of time it takes to introduce new ideas. Personalised services such as those that automatically generate travel plans when there are delays or cancellations now look as if they may never appear and instead users will persist in using a range of separate applications that they select to suit their needs. This is a social phenomenon, not a technical issue.

Potential developments

Even given these limitations on the power of prediction, it is highly likely that the end user’s experience of wireless will change significantly over the coming years. For example, within the next 10 years, we could see the emergence of the sort of people locator featured in Harry Potter: a device that sits on a worktop and shows the location of different members of a community – perhaps a family or school class – at a glance showing locations such as work, school, home or travelling. This could be readily implemented today with location information derived either directly from a phone handset using its GPS or using WiFi-based positioning functionality and sent via standard data networks to a small display.

With location functionality widely available on standard devices, getting lost accidentally will also become a thing of the past Not only will devices provide directions but they could provide much more information. Handsets will show a picture of what the user can see in any direction but superimposed with information such as the name of buildings or the location of coffee shops. Specific services could be provided to the partially sighted such as voice commands or phone vibrations in certain patterns to indicate direction changes needed.

In the world of healthcare, wireless communications have the potential to become one of our most valuable technologies to support our ageing population with home networks monitoring the wellbeing of the elderly or infirm and informing the health service or carers if there are problems. Wireless networks can dispense medication and note if it is not taken correctly. Sensors, perhaps embedded in a bathroom mirror, could enable real time remote health monitoring for the unwell. Entertainment devices such as the Nintendo Wii can both entertain and simultaneously monitor cognitive functions. They could even help those with memory problems by monitoring behaviour and prompting people as to what they were doing and guiding them home if they get lost.

In the area of transport, wireless could allow cars to drive closer together, automatically braking vehicles to avoid collisions and calling the emergency services should an accident happen. Wireless devices could be used to tag travel baggage so that an owner would know exactly where it was in the system or allow real-time CCTV monitoring of passengers in train carriages.

Wireless finance could facilitate payment using a mobile phone – either by touching the phone on a sensor or by a mechanism akin to a premium text message. Indeed, early versions of this are already being trialled in a number of cities worldwide. Wireless can navigate us through the world, not just providing directions but showing us what points of interest are nearby and providing further information or helping with specific functions such as graphically showing all homes for sale in the vicinity along with the price and contact information.

Lessons from the past 10 years suggest that although technology trends were reasonably well predicted, the time taken for new ideas that involved social changes to lift off was often greatly underestimated. The question for the next decade is whether this phenomenon will continue. The changes needed in social behaviour to exploit new technologies will always be slower than the technological development itself? Or is it simply the case that, until now, the rate of social change has lagged behind technology for reasons of unfamiliarity or social convention?

If that is the case, we could expect that as new ways of using wireless permeate society and a new generation adapt ever more quickly to new ideas, the take up of new technologies could accelerate and even outpace our capacity to predict and deliver the engineering behind the idea.

[Top of the page]