Notes Blogged during the Wireless Utopias 05 event
Dana Centre, London Science Museum, 26 May 2005, 19.00-21.00
these notes have not been checked against the event video and may contain errors
Open Spectrum Uk is an advocacy project convened by JohnWilson
It was formed in the context of the Ofcom Spectrum Framework Review.
We have had over 60 years of radio spectrum management.
Spectrum Management concerns Rights of Access to radio spectrum.
Three models are currently in play:
- command and control
- market based approaches
- license exemption
There is a continued deregulatory thrust that started in the 1980's.
In my (JohnWilson
) opinion this is part of a neo liberal agenda.
The UK was a pace setter for spectrum management.
The 6 - 7% of spectrum currently proposed in the framework review is not enough for an unlicensed allocation.
The question is: how can we move on from legacy/property-based models of spectrum management?
Internationally, the UK is seen as exceptional in the extent to which it is focused on Market Mechanisms.
Spectrum management must take into account the Economic, Technical and Public Interest.
In the current process there has been a failure to engage the public interest agenda.
There is no proven-perfect model for technology or regulation.
I suggest a Darwinian model of competition between regulatory techniques.
There has been international advocacy on this agenda, especially in the US, but not in the UK. We hope to provide a forum to fill this space.
William Webb is author of the Spectrum Framework Review and head of R&D at Ofcom.
Responses to the Spectrum Framework Review are now in.
It was first published in November 2004.
There are 3 types of Spectrum management considered:
- Command and Control
- the plan is to restrict this to a few frequencies where there are no options - 21%
- Market Forces
- Planned to be used for 72% spectrum
- Spectrum trading = trading of licenses
- Allowing the ability to alter geographies and equipment profiles
- Allow users to change use, e.g. from broadcast to cellular telephony
- They believe there is enough unlicensed spectrum out there
- But they have argued for a small increase from 6% to 7%
Ofcom is keen to see if there are other ways of dividing up spectrum.
- For instance, using new technologies.
Ofcom also keen to see new technologies develop
- Either by sponsoring R&D or allowing new technlogy in existing (licensed) bands
- Wimax, mesh, adaptive antennas, software defined radio
Introducing new technology in license exempt bands
- Generally doesn't require regulatory decision
- Can be deployed as soon as available
Introducing new license exempt technology in existing licensed bands
- Dynamic spectrum aceess
- Ultra Wide Band
- with UWB Ofcom set out to implement the US model but are now going through European model
- UWB regs still under review
To free spectrum there should be
- No barriers to use
- Simple and transparent process for changing use
- Rights well defined to protect investment
The Framework Review attracted 100 responses
- Large number from radio amateurs (40)
- Very few respondents said our approach was fundamentally wrong
- Most comments overwhelmingly positive
- Revised Spectrum Framework Review policy statement due in July 2005
- Ongoing studies
- We expect to allow higher power unlicensed in rural areas
- Looking at cognitive radios
We now move on to the far harder task of Implementation which is thankfully the job of the next speaker.
Peter Bury is head of spectrum licensing at Ofcom.
My Job is to implement the vision that William outlined.
This will deliver serious economic benefits to Europe and the UK.
Liberalisation of the spectrum to make it tradeable will net billions of pounds.
The market is better placed than the regulator to decide how to allocate spectrum.
We will liberalise so that the spectrum can be used as appropriate by the market.
Our top priority is providing spectrum
- Auctions are the appropriate way to move spectrum into the market.
- Most efficient way to evaluate the economic benefit of the spectrum.
- Need to see what it can be used for.
- Acquire defensible property rights for spectrum.
- Defining rights for spectrum users.
Expectations are that:
- Early in the next year to 18 months, new capacity in the 410 Mhz band.
- DECT guard bands auctions.
- 2 chunks of radio spectrum will become available.
- 3G expansion bands (240 Mhz).
- A strong lobby to make that a technologically useful band.
- Analogue TV switchoff will make spectrum available.
- Details of the implementation plan will be published in a consultation document.
- More responses are needed and welcomed.
Michael Marcus joined the Federal Communications Commission in 1979 where he held a variety of positions. He proposed and was responsible for the 1985 spread spectrum decision that established the unlicensed bands for spread spectrum, and played a key role in the Spectrum Policy Task Force.
There has been a creative feedback loop between US and UK regulators.
There exists mutual admiration between US and UK regulators even though the UK regulator keeps changing its name.
Classic spectrum management developed early in 20th Century.
- Originally there was no spectrum management - anyone could do anything.
- This changed after sinking of the Titanic.
- Spectrum became something that government felt it should regulate.
In the 1920's:
- Upper limit of usable frequencies was 20-30 Mhz
- This was considered (and is still called) the High Frequency band
- Governments chose who could use spectrum.
- Devices were for long distance communication.
- Few technologies available.
- Modulation options were simple: Morse Code or AM.
Today the situation is very different:
- Tons of technical details: definition of GSM is a book inches thick
- Wireless getting up into higher reaches of the spectrum
- Frequencies around 60 GHz are absorbed by Oxygen, making the band inherently short range
- 77 GHz now in use commercially.
- 95 GHz devices in development.
- thus well-suited for unlicensed use.
- RA's predecessor proposed unlicensing 60 Ghz.
- "This was a Eureka moment for me."
- But now Ofcom is about to license 60 Ghz - the very band most suitable for unlicensed! Why?
Telecoms is an economic underpinner.
But too many choices prevent capital from investing.
Overregulation also discourages investment.
3G is based on Qualcomm technology.
Qualcom is an American Company - why?
- Deregulatory atmosphere in the US.
- Ideas look possible to investors.
- Many skillful innovators.
The problem of spectrum management is 'whitespace' - holes in spectrum utilisation.
'Worst case demand' has to be provided for in spectrum allocation.
- You have to choose between 'haves' and 'have nots'.
With new technologies, you can exploit whitespace.
There is a parallel with electricity generation:
- In the US it is cheaper to buy "interruptible" electricity
Why not have interruptible spectrum rights, cheaper than dedicated channels?
Dewayne Hendricks is chairman of the Dandin Group and has been responsible for investigating spectrum sovereignity for five Indian nations and building wireless networks in Tonga.
FCC does not have jurisdiction in the five indian nations' territory...
Unlike the previous speakers, I actually build wireless networks.
Before telegraph, the fastest communication was foot speed or sailing speed
Telegraph moved us from foot speed to light speed.
Marconi began offering commercial service by 1903, within 2 years of his first demonstration of practical wireless communication.
At that time spectrum was a commons.
- Anyone could transmit or receive
The sinking of the Titanic changed that and interference became an issue.
A regulatory path started in the US which has now become the FCC.
Spectrum went from a Commons to Property.
Now we face the transition from Light Speed to Warp speed.
Wireless makes points in space stand on top of each other.
2 years ago - Municipal Wireless Infrastructures weren't on the agenda.
- Now most cities in the US are announcing Wifi clouds.
- These are emergent systems - new things coming unpredictably all the time.
- Lots of stuff happening "under the radar"
Municipal wireless uses equipment designed for 300-foot ranges - and makes it go miles.
- Doesn't work perfectly, but works 'Good Enough'
Cheap, disposable equipment - user owned.
This first happened at the beginning of 20th Century
- Now we are doing that again, but at the same time we're in new territory.
- You are the experiment.
- You are the Marconis.
Ofcom uses words like "marketability" - that's business stuff.
I'm talking to people who are building networks, who are just are building stuff that works.
- The power of "existence proofs"
- Lots of cities are now coming online.
is documenting that revolution.
Wireless networks can be built in ways where the whole is greater than the sum of the parts.
You can be sitting in Manhattan and "see" 24 different networks.
- Shouldn't that create a tragedy of the commons?
- NO EVIDENCE THAT IT HAS.
Innovation is at the edge, esp. in user-owned and -provided networks.
Spectrum commons work - all it takes is a few simple rules in cheap devices.
Wireless in laptops stepping up to ethernet speeds.
NSFnet was turned off in 2005.
- Was the Internet's backbone, but its loss wasn't even noticed.
It's ten years since the start of the commecial internet.
- Where we will we be a year from now?
- What is your radar missing right now?
We're heading toward 'same day service in a nanosecond world'
Who has heard of the darknet? (Show of five hands)
Peer-to-peer traffic is 60-80% of global internet.
UK is innovating and ahead of the US in darknet use.
Bit Torrent now accounts for 50% of all traffic.
Regulators aren't talking about that.
It's much more than just illegal content.
Dale Hatfield (former spectrum chief for US Govt) once said if you want to know where technology is heading, watch what people do illegally then legalise it.
"Good enough" will dominate user deployments.
Moore's law means devices will continue getting cheaper and faster.
We can look forward to 1 gbit radios at WiFi
Best practices are needed to make it work.
Innovation is at the edge and you are in control.
Onno is a researcher and network builder from Indonesia responsible for getting literally millions of Indonesians online.
Excuse me if I'm not too coherent - this is 2am in Indonesia!
I help provide low cost access to the internet.
For many years I steal frequencies: in Indonesia 2.4 Ghz wasn't de-licensed until this year.
When I started 12 years ago, it cost $800 per month for 64kb, $300 per month for dialup.
- Now $15-30 per month for broadband.
In 1993 there were:
- Few users
- No cyber cafes
- No ICT magazines
12 years later:
- 10 million internet users
- 15,000 Wifi networks
- 2000 cyber cafes
- 4000 schools on internet
- 1000 community radios
- 20 ICT magazines
We have done this without Worldbank or IMF help.
- It's my kind of network Jihad!
How have we achieved this?
- 12+ years education process
- User financed
- Distribute CDROMs with technical info.
- Free download of books from web sites.
50,000 Indonesians now participate in wireless/ICT mailing lists.
- email@example.com is most active
- I get 600 emails per day - I'm dead!
We run wireless workshops and charge entrance fees of $3-5.
- Publish books on technical topics for $2-3.
The media and press help us nationally and internationally, keep us from getting arrested.
- Helps spread knowledge/awareness beyond Indonesia.
We have changed the way we do business.
In the old model we lobby government to provide service, they debate it, raise funding and then deliver us the service or change the law.
The process is too long and expensive, so we make a new process:
- We provide knowledge to the community.
- The community grows.
- They build the infrastructure.
- They train the next generation of radio rebels (picture of schoolkids building a radio tower).
- They build the towers.
- We become strong and useful, then can pressure the government.
This year the government finally changed the law to make 2.4 Ghz operation legal in Indonesia (applause)
We still have a lot of work to do.
- A thousand new users every week is too few - look how big Indonesia's population is!
In 2.4 ghz band, we still need to get rid of business licenses.
And delicense the 5.2 Ghz and 5.8 Ghz bands.
And create more public awareness.
And make our networks sustainable.
Gordon Agee has spent the past three years building a pioneering Wireless Broadband project in Buckfastleigh, Devon, and now is working on Broadband4Devon advising small businesses on how to get broadband access.
It is very inspiring but difficult to follow Onno.
I feel like a CountryLife
We organise grants to small businesses in Devon to help get them online.
This has given me an insight into the lives of the 1.1 Million people who live in Devon.
- Mainly employed by small businesses.
I've been through 2.4GHz 3.5GHz 5.8GHz - I've played with all the gear.
I think we are entering the "age of ruts" in the UK.
- Technological rut
- Commercial rut
- Market structural rut
The govt isn't bothered with changing the ruts. They are influenced by:
- Wirehead DSL friends dominated by BT
- 3G crew dominated by 4 major mobile players
On the other hand there is:
- The wireless crew
- Lots of citizens etc.
These are treated like naughty children.
The first 2 groups (and the government) don't have any interest in letting the naughty children play.
The future is high bandwidth wireless, but there doesn't seem to be any long term strategy.
this section awaiting notes from PeterCochrane
Peter is ex CTO of British Telecom and a distinguished engineer. I normally leave it to Peter to wrap up a session and today is no exception.
I was paid to be a naughty child in BT.
I was pretty upset about 3G which bankrupted UKPLC for 5-10 years and I have been critical of 3G auctions.
What went wrong with 3G?
- The auction process was a rape of industry
- Commercial dictats and tech igrnorance
- No demand for services offered
- Failed WAP services
What went right with wifi?
- Free spectrum
- Open standards
- Fulfilled real needs
We are moving from a world of 'Do It For Me' to a world of 'Do It Yourself'.
People look at things through a microscope instead of a telescope.
- It often pays to look in a different direction.
- It is very difficult to see what will happen in the next 20 years.
Networks are getting dumber as intelligence moves to the periphery.
My home in 2005:
- 3 terabyte server
- my son reasoned that 'I can afford one' and 'we need one'
- 2 optical fibers (we put the fiber in ourselves)
- 10 Mbit internet
- 1 Gbit lan
My home in 2010:
- 1 peta-byte server
- 2 optical fibers
- 1 Gbit internet
- 100 Gbit lan
Everything is accelerating at triple digit speed.
We already have the technology to deliver unlimited bandwidth.
Paradoxically, more wireless means more fiber.
Computational power solves many of the problems of the past.
- Allows us to process signals to get around interference.
The fundamental challenge is that we can't predict what spectrum can do.
The single-use-per-band model no longer works.
We will have to dynamically allocate spectrum.
- Users can manage local spectrum themselves:
- If your microwave oven interferes, get another one.
- If the phone doesn't work, switch channels.
The dream: seamless access over all networks all the time.
We need to escape 'firewall' restrictions.
But where to put all the antennas?
- We need smart antennas that can use multiple frequencies simultaneously.
- And multichannel adaptable chipsets for radio front-ends.
Audience Question and Answer session
I like the idea that innovation happens at the edge of the network.
But why is that model in play today when previously innovation came from the centre?
Aren't we going for the or
rather than the and
Shouldn't we be encouraging central research and innovation as well as self-innovation?
The reason innovation has gone to the edge is computational power.
Innovative stuff is happening in people's houses (viz the games industry).
Go to 405themovie.com and download it - its a good example.
There is a complete StarWars
movie on the web made by amateurs.
RIAA is getting upset over a few downloads, but by 2015 the 'super iPod' will contain all music ever recorded.
- Will you buy it or steal it?
The model where people need to spend lots of money for spectrum access doesn't benefit the public.
Opening the spectrum to market forces will just let the Rupert Murdochs of the world buy all the spectrum.
Remember that in the 3.4 Ghz auction 1 company bought all the licenses and then didn't use them.
I chair the Welsh broadband stakeholders group.
Here we are in an urban area. Where we operate in Wales, it is really rural. We have different issues.
- Power levels are very important to us.
James tells me that he can get 100 GBP per month per megabit over 5.8 Ghz in London
- Can't find anything like that in rural areas.
I suggest you can relax some of the rules around existing technologies without leading to interference
- (to stimulate rural uptake?)
Why are we talking here?
The government has made up its mind that spectrum auctioning will be the way forward.
I am also appalled at the techno-utopianism demonstrated here.
The technology is not the answer - it is a social project, awareness is important.
I give away knowledge.
If the people accept the knowledge, they can create the technology.
Community uptake of the technology is driven by education.
I think its interesting that in the Ofcom presentation they focused on market forces, though privately they say that they want a mixed economy.
If Ofcom seriously believes in mixed economy they should provide spectrum access in a variety of uses.
We need to escape from the 80s and allow different social groups and forms of organisation to compete rather than using the market as an approach on everything.
A range of academics point towards a mix of spectrum management approaches.
Something that might benefit a commercial service may not benefit a noncommercial service. Is what's good for an operator always good for society?
It's hard not to be struck by the gap between the vision that Peter, Dewayne and Onno describe, and the direction that Ofcom is taking. Looking to the future, will there be dissatisfied users - dissatisfied regulators - or a course correction?
There's is no benefit in this discussion for users.
They will just use what the market offers.
There is no benefit in the public becoming involved in spectrum management issues.
They just want to have technology that works.
You can't buy some of the technologies talked about here.
At what point does interference become an issue?
In Electronic Warfare we have a concept called 'burn-through'. People think that if you keep adding more and more users, the world suddenly comes to an end. This isn't true - you just have to rework the geometries a little. The world does not come to an end when you add the last user.
We're going to look at people who are trying to alter geometries.
As someone who is running a wireless network in London right now, technology is never the problem. The problem is how do we get a critical mass of people to say 'let us do what we want to do with our global commons. Please.'
In Ofcom's investigations into the auction model, what is the revenue forecast for terrestrial television? I do believe that these things are commodities and the current models are too entrenched to change. Could you purchase licenses as a resource for the community and give them away, for instance?
It would be amazing if communities could do this, so yes. The model of community broadcasting we have now is heavily regulated and underused at present.
How much would 401 Mhz cost?
How much will your community-owned house cost? The market will decide. There is an assumption that spectrum auctions are there to raise revenue. It's not true. It is illegal for us to raise revenue from spectrum auctions.
The BBC archive was put under Creative Commons. We are moving towards [situations where public goods are made available in the public domain.]
A question for Peter Bury. I noticed in your discussion of the implementation plan that you said the DECT GSM sideband is earmarked for spectrum auctioning. But in an economic survey commisioned by Ofcom from NERA during the public consultation process on the future of that band, the conclusion was that the optimum economic use of the band would be achieved by making it license exempt. Why has Ofcom gone against the advice given to it by its own commissioned report?
Further technical evaluation of the band found that GSM is not a polite protocol and that the proposal to license exempt the band was not technically feasible. A number of licenses will be auctioned concurrently in the band.
To wrap up I'd like to ask the panelists "What is your wireless Utopia?"
I'd like to quote Dale Hatfield again - look at what people are doing illegally and legalise it.
I want bandwidth wherever I go.
Would you know a utopia if you saw it?
We'd like to be in a world where we don't have to regulate - where the market can decide.
Let's have a 25 year roadmap on spectrum.
An anecdote: in the planning of the first mobile network, we scoped it for 100,000 users. There are now 65 million GSM users in the UK.