Friday, September 01, 2006

Future of Wireless Comms Book now available

West Sussex
The book to which I contributed a chapter is to appear in January 2007 and is now available .
Simon

Friday, August 18, 2006

UK Wi-Max: Spectrally Challenged?

from Warnham

Wi-Max is a mixed blessing. It's a fantastic technology, with great potential, but plenty of questions hang over the size of its potential success.

I've previously written about the generality of the technology: Wi-Max is not magic - it's subject to exactly the same trade-offs between range, capacity and quality which any system deals with, so that mobile Wi-Max is likely to achieve much the same range and practical throughput as 3.5G, despite higher advertised peak rates.

Nevertheless, Wi-Max (and similar technologies, particularly Wi-Bro) is showing great uptake outside of the UK, such as the recent announcement by Sprint Nextel in the US that they are intending to spend $3 billion to support 100 million WiMax users by the end of 2008.

In the UK, however, there is a limit on deployments due to the lack of available spectrum. The 5.8GHz band (just above 802.11a Wi-Fi) is available, but those frequencies don't propagate very far and don't penetrate into buildings very well. Additionally, the 'light licensing' regime in that band means that there is no certainty of protection against interference from other users.

Following some auctions and subsequent aggregation Pipex and UKBroadband have some decent spectrum at around 3.5GHz. UK broadband is operating using other technologies, but Pipex is certainly pressing ahead with WiMax plans via trials and have marked up the estimated value of their licence from zero in 2003 to £5m in 2004. There is reason to suspect this is a substantial underestimate now...

The really 'tasty' spectrum for Wi-Max would be the set of bands around 2.6 GHz. It propagates well, there's a decent amount of bandwidth available and equipment should be readily available as these bands (or nearby) are widely available in other major markets. Ofcom has announced its intention to licence these as part of its burgeoning spectrum awards programme, but plenty of roadblocks remain, including the fact that these bands are equally attractive for extra 3G capacity, and are currently designated for 3G alone at European level.

It could easily take until the end of 2007 before the relevant auctions take place, and there are many unanswered questions about the way the spectrum will be packaged and managed. All this adds up to considerable uncertainty for Wi-Max in the UK.

I've been asked several times how much this spectrum might be auctioned for. My crystal ball is far too cloudy to put any precision to that, but for an interesting data point we can look to the current auctions by the FCC in the US for spectrum for Advanced Wireless Services - basically 3G and related services. At the time of writing the price was standing at over $12 thousand million (being careful not to confuse transatlantic billions).

Crudely, if we weight that figure by the ratio between the UK population and the US population we might expect the UK auction to yield at least £1,300 million. In practice I think there are at least two factors which might increase the UK pricing. UK population density is nearly ten times that in the US, increasing the return for operators relative to the network investment. Additionallym the US already has spectrum suitable for WiMax, so we may see operators supporting competing technologies in a bidding war.

It's not quite at the same levels as the heady days of the 3G auctions in 199 (which produced around £23billion), but if Ofcom can move quickly to clear the remianing obstacles,it promises to be an interesting auction. On the other hand, if they don't, WiMax could miss its mass-market opportunity in the UK altogether.

Friday, August 11, 2006

The Mesh Conundrum...

from Warnham

Mesh Wi-Fi is a hot topic, with commercial products, standardisation activities making good progress and much excitment about the potential for wireless cities using this technology. Onthe other hand, reserach commissioned by Ofcom suggests that the some of the claimed spectral efficiency and capacity advantages of meshes, particularly at lower frequencies and in mobile scenarios, may be false.
Why the disparity? Is the Ofcom study not relevant to wireless city scenarios?Or are manufacturers overlooking some basics in the rush to produce product.? I'd be interested in views either way..

Thursday, August 10, 2006

UK Communications Market Statistics

Ofcom have just published their annual market report, packed with interesting stats and free of charge:
http://www.ofcom.org.uk/research/cm

Some examples (my emphasis):

  • Total retail revenue in the UK communications sector reached £50bn in 2005, a 5% increase compared to 2004. The biggest annual rise (£2bn) came from the telecoms sector
  • Consumption of many communications services is rising. On average in 2005, mobile subscribers made more calls and sent more texts than they did in 2001, internet users spent almost 20 minutes more time online per week.
  • There is also evidence of a significant difference in communications usage patterns between young adults and the general population: for example, 16-24 year olds spend on average 21 minutes more time online per week, send 42 more SMS text messages, but spend over seven hours less time watching television.
  • Availability of several new digital services increased during 2005; 3G mobile services are now available to over 90% of the population and the proportion of unbundled exchanges is up ten percentage points on 2004.
  • Take-up of digital services continued to increase over the year: by March 2006 18.3 million UK homes took digital TV services, 11 million homes and small businesses had broadband connections, and there were around five million 3G subscribers.
  • Mobile telecoms comprised £13.1 billion (34%) of total retail telecoms revenue – up from £12.0 billion in 2004
  • Some 31% of UK call volumes in 2005 were from mobiles – a rise of three percentage points from 2004, and 11 percentage points higher than in 2001.

Consumer Push Email to drive mobile data volumes

From Warnham

One of the predictions I have made in my chapter for the forthcoming book "Future of Wireless Communications" is that push email services will be taken up extensively by consumers - particularly young ones - and this will be one of the factors which heavily drive the growth in data volumes for 3G (and other) networks.
It seems that RIM think so too, as it is rumoured that they are soon to launch a consumer version of the Blackberry.
Another of my predictions is that networks will start to offer instant messaging with free updating of online contact lists to drive traffic. Still waiting for some evidence of this one...

Sunday, June 25, 2006

So just what is wireless?

From Home

In this part of the world, the last few years have seen the word 'wireless' become synonymous with Wi-Fi.

For several years more, in North America, wireless has usually been taken to mean what Europeans call cellular mobile.

When I say wireless, I mean the totality of wireless communications, including all the frequency bands, technologies and application areas. It even includes non-radio technologies, such as infra-red data and line-of-sight optical communications. This definition is reasonably consistent with the definition currently in Wikipedia.

Wireless in this broad sense is more than a century old.

Why does it matter? Because the danger is that a poor choice for mapping an application to a technology and frequency band could lead users to conclude that any wireless technology will produce the same result. They may then be reticent to put any more time, effort or money into getting it right, and that would limit their capabilities to communnicate and gain benefits, as well as slowing the industry.

Reclaim wireless! The campaign starts here...

T-Mobile Announces HSDPA Service

From home

The following data rates, offered via HSDPA mobile and over the wide area (i.e. not Wi-Fi) would have sounded like science-fiction just a few years ago:
  • Initially (Q3/4 this year) with a download speed of 1.8Mbps everywhere in the cell, compared to just 384kbps (up and downlink) for standard 3G.
  • Next year 3.6Mbps download nearest the base station, degrading as the subscriber moves out to the edge of the cell,
  • By the end of next year, 7.2Mbps download out to the cell edge, together with 1Mbps uplink using HSUPA.
I'd caution that an over-the-air rate offered everywhere does not translate to the same user rate everywhere, just as 11 or 54Mbps advertised on your Wi-Fi client does not translate to anything like that if the channel is busy or if you are suffering interference. Nevertheless, these are impressive numbers. How long until that kind of rate is widely available via Mobile WiMax?

See here for the full article (you'll need to register with ComputerWire).

T-Mobile Announces HSDPA Service

From home

The following data rates, offered via HSDPA mobile and over the wide area (i.e. not Wi-Fi) would have sounded like science-fiction just a few years ago:
  • Initially (Q3/4 this year) with a download speed of 1.8Mbps everywhere in the cell, compared to just 384kbps (up and downlink) for standard 3G.
  • Next year 3.6Mbps download nearest the base station, degrading as the subscriber moves out to the edge of the cell,
  • By the end of next year, 7.2Mbps download out to the cell edge, together with 1Mbps uplink using HSUPA.
I'd caution that an over-the-air rate offered everywhere does not translate to the same user rate everywhere, just as 11 or 54Mbps advertised on your Wi-Fi client does not translate to anything like that if the channel is busy or if you are suffering interference. Nevertheless, these are impressive numbers. How long until that kind of rate is widely available via Mobile WiMax?

See here for the full article (you'll need to register with ComputerWire).

DVB-H vs DMB for Mobile TV: Whither 3G?

There is currently a debate raging as to whether the best long-term technology for broadcasting mobile TV, is DVB-H or DMB, with DVB-H looking like the front runner.

But this is odd. 3G was created as a universal standard for wide area mobile access to high-rate data applications. When it was being standardised, no-one was entirely sure what those applications might be, but they pressed on anyway and lots of marketing folks spent time around whiteboards discussing what the 'killer application' might be. There now looks to be a strong possibility that Mobile TV is an example of such an application: high bit-rate, mass appeal, and the real prospect that users might actually pay more for the service, thereby increasing ARPU.

And just as operators find a killer application, they decide that 3G might not be the technology for offering it! I find this hard to believe. Instead, I think operators will find ways to make 3G do the job. Let me explain why.

One justification quoted for moving to DVB-H/DMB is that 3G gets heavily loaded by high-rate applications, causing cell breathing and hence gaps in the network coverage, particularly indoors. That's certainly true, but technologies like HSDPA and 3G LTE (so-called 'Super 3G') will enhance the data rate capabilities of 3G, and new base station form factors and distribution technologies will provide an economical way to roll-out indoor coverage.

Another is that 3G is not fundamentally a broadcast technology, requiring the same programme to be sent individually to each user (unicast), worsening the loading issue. True, but MBMS will deal with this by providing multicast options, and the availability of large amounts of flash or hard disk storage on phones means that they will download delay insensitive programming, which represents most TV, when the networks are not fully loaded.

Also, UK 3G handset penetration has now reached about 15% (source?), growing to 68% by 2010, there aren't any DVB-H/DMB handsets on the open market yet. I think it will take about ten years longer for such technologies to reach 75% penetration than 3G. Operators can't wait for that and will do what they can with 3G, which is currently fairly under-utilised. Operators such as Vodafone are already doing this.

Lastly, but perhaps most importantly, the general discussion is that wholesale operators such as BTWholesale or Arqiva will own and run the networks over which Mobile TV is provided. There is some logic in working that way, but clearly they won't do this for free and this will erode the ARPU advantage of Mobile TV. Again, I think the mobile operators will not let this happen too easily. These networks will themselves need huge investment and will take many years to provide extensive coverage, by which time 3G network coverage will have expanded hugely in extent and depth. Currently even the relevant spectrum is not available, although Ofcom is working on that, with auctions expected next year.

Overall, then, I think the relevant battle is not between DVB-H and DMB, but between these technologies and 3G, where the issues are more finely balanced. Operators might just be making the right noises about the broadcasting technologies while figuring out all the ways in which they can avoid the day when they have admit that 3G is not fit for the very purpose it was designed for.


Saturday, June 24, 2006

How Many HotSpots doth a HotCountry make?

From home

In my last post (on the BT HomeHub), I commented briefly on the range available from Wi-Fi hotspots compared with a wide-area cellular mobile operator. But what's the real situation?

Wi-Fi access points might have a typical range of about 30m (depends on lots of complicated radiowave propagation issues and the quality of your access point and device and lots of other things, but bear with me).

According to BT’s own list of hotspots (5.52 MB), there are 9362 of these in the UK, including those provided by other service providers (e.g. The Cloud and T-Mobile) who BT has a roaming agreement with.

By this simple arithmetic, these hotspots currently cover some 25.5 million square metres.

The UK land area is 244,820 square kilometres. That means only 1/9249 of the country’s land area is currently covered. It would take 86.5 million hotspots to cover everywhere. It seems that they are a long way from being able to provide a service which could compete with that provided by mobile operators.

But hold on. According to government statistics, “Over 90 per cent of our population lives in urban areas covering just eight per cent of the land area.” This area could be covered by ‘just’ 7.8M hotspots. Imagine that it costs £1000 to install each of these (including hardware). They’d then cover 90% of the population for 8 thousand million pounds.

Or just 30% more than Vodafone paid for their 3G spectrum licence in 1999. That was before they had even started building the network, which even today does not cover 90% of the population (Vodafone currently say their 3G network covers 73% of the population, compared with 99.7% for their 2G network. See map). It gets better: BT will even get you to pay for the hardware (£200) if you want to help them by adding another hotspot to their network.

Discuss.


BT Home Hub launch: Just a big Ipod, or a point of control over your activities at home ?

BT have just launched their 'Total Broadband' home service (see bt.com/broadband). Based around an ipod-styled 'home hub', it provides a lot more than vanilla broadband, including :
  • Home Wi-Fi access
  • Videocalls
  • Unlimited local and national voice calls
  • Wi-Fi access at BT (and roaming partner) hotspots and at their forthcoming 'wireless cities' via BT's 'Openzone' service.

They are also offering a Wi-Fi internet radio and it looks from the press as though they are intending to offer home security (burglar alarm) functions, gaming and ultimately video on demand (about time too - they've been talking about this for at least 20 years).

The home hub itself is hugely significant. Although in essence it's just a customised router and access point, it acts as a point of control in delivering the bundle of services listed above, and thereby reduces customer churn for BT. If you like the set of services you are paying for, it's unlikely you'll find quite that set from any other provider (and the 12/18 month minimum contract term should help to hold onto you too). It also provides BT with an opportunity to upsell you on all sorts of add-on services in the future, too (I gather the software on the hub can be updated remotely by BT). Most importantly for BT, it lifts their service from the cut-throat, low-margin world of access-only broadband to being exactly where they want to be: a provider of differentiated, high-margin, 'sticky' services.

The inclusion of the Openzone minutes is also interesting. Until now the service has been very much targetted (and priced) for corporate users. Once suitable devices are made available, this will enable BT to offer you the same range of services anywhere you go within range of a hotspot. But, you might say, finding a hotspot is really hard work and they are hard to find compared with mobile phone networks coverage, which is available almost everywhere. Watch this space...


Future of Wireless Communications Book

I am pleased to say that I was invited to contribute a chapter to the book of this name, edited by Prof. William Webb, Head of R&D for Ofcom. I have just finished working on it and am looking forward to seeing the result in print later in the year.

This is the second edition of the book- the first edition is now five years old.

I am amongst prestigious company, and can only hope that my chapter lives up to this in some small way. My co-authors are:
  • Editor: Professor William Webb: Head of R&D and Senior Technologist for Ofcom
  • Professor Peter Cochrane. Previously CTO at BT and now involvedwith a wide range of ventures, start-ups and innovative consultancies.
  • Professor Dennis Roberson. Previously CTO of Motorola, now at the Illinois Institute
  • Tomi Ahonen. Author of a wide range of books focussing around 3Gservices, applications and marketing.
  • Stephen Temple. Senior strategist at Vodafone and one of the keydrivers behind the emergence of GSM.
  • Padmasree Warrior. CTO at Motorola.

In my contribution, I have first divided the next twenty years into three fairly arbitrary periods of five years, five years and ten years. For reasons which will become clear if you read the book (to be published towards the end of the year), I have given these names as follows:

  • 2007-2011: The age of wireless proliferation
  • 2012-2016: The age of wireless similarity
  • 2017-2026: The age of wireless mundanity

In each age I have divided my thoughts regarding the future of wireless into the following general areas:

  • Services & Applications: what people are doing with wireless;
  • Devices: what people are using to access their wireless services: the user’s means of access to the wireless world;
  • Infrastructure: the changing equipment and system topologies for providing users with wireless access and interconnection to other systems and data sources;
  • Air interfaces: the 'over the air' language via which devices and networks intercommunicate;
  • Spectrum: the electromagnetic medium which permits wireless to work in the first place.

The first four of these follow a fairly conventional communications hierarchy. The last deserves a special mention, however. Spectrum, in the context of wireless communications, acts as the 'layer zero': the very stuff of which wireless consists. As result, I have paid special attention to its characteristics and impacts on the development of future systems and technologies.

I am confident that many of my predictions will be proven true and equal confidence that I will be utterly wrong in some important respects. Indeed, in several cases I have included predictions which may seem extreme in order to provoke discussion and to illustrate some important points. I look forward to discovering which predictions become realities over the coming years. In the meantime, I welcome hearing of your own opinions on how wireless is likely to develop over the next 20 years.