EI Workshop 121: Open Spectrum for Development in the Context of the Digital Migration

Sixth Annual Meeting of the Internet Governance Forum
27 -30 September 2011
United Nations Office in Naiorbi, Nairobi, Kenya

September 29, 2011 - 16:30PM

***

The following is the output of the real-time captioning taken during the Sixth Meeting of the IGF, in Nairobi, Kenya. Although it is largely accurate, in some cases it may be incomplete or inaccurate due to inaudible passages or transcription errors. It is posted as an aid to understanding the proceedings at the session, but should not be treated as an authoritative record.

***

>> CLAIRE SIBTHORPE:  Good afternoon, everybody, my name is Claire Sibthorpe, I'll be moderating this session.  Thank you for coming.  Spectrum I think has come up quite a few different times at different sessions at the IGF so this session will be focused on spectrum management and I think some of the changes that are happening sort of drive the need to look at approaches to manage spectrum.  This session has been organize by APC who have a number of projects focused on the issue of spectrum.  One is research on spectrum regulation in Africa, Asia and Latin America, and the research and reports are available on the APC website, and a very useful paper that's been written by Steve Song on the panel introducing the issue of spectrum, there are a few copies here, you can also get it on the website.  I would certainly encourage those who are new to this area to read it.  It's very useful.

The other project APC has is looking at digital migration in West Africa.  It's in the context of the work that APC is doing in this area.  We felt it was very useful to have a session, and I think because of the interest of time, I'll just start by handing be over to our first speaker, which is Steve Song, who is the founder of the Village Telco, and he was also the lead researcher for the research done on spectrum in South Africa.

>> STEVE SONG:  Thanks.  So my remarks really are to contextualize the discussion we're having, and really this discussion arises from the importance and increasing importance of wireless technologies, in terms of reaching everybody and giving access to everybody.  Technologies that work in the industrialized world, last mile copper solutions, whether ADSL or coax are not going to be the solutions, at least not on this continent, those solutions are going to be wireless, which brings us to the issue of managing wireless technologies and indeed the issue of spectrum management, and the key reason for spectrum management is simply managing interference, making sure that signals don't interfere with each other and the historical approach to doing that was to simply separate signals far enough away from each other that they didn't interfere, and while availability exceeded demand, that was a perfectly reasonable approach, however, in the last, say, 10 to 15 years, we've gone to the other situation where demand exceeds the available supply or at least the apparent available supply, and we get into the situation of a struggle around what to do in terms of how to make more spectrum available, and it's important to understand what we're talking about, so what is spectrum, and that debate has led to a lot of sort of philosophical investigations, what is the nature of spectrum and there are people who argue that it is a common thing, it's like air or water, something that should be shared amongst everybody, it's reusable, a bit like the digital world and comparisons have been made, from open spectrum to open source approaches to sharing things.

On the other hand, you have private property approach, an argument made by Ronald Coase in the 1950s, the best approach was to treat it like private property so people would pay for it and avoid the tragedy of the commons.

The truth, as usual with complicated things, is somewhere in between, that spectrum is elastic and we are learning to pack more data into the same amount of bandwidth, and that shows no sign of stopping at the moment, yet we also know that that process is not infinite, there's what's known as the share and bound, that is, the amount of data you can pack in.  It's elastic but not infinite.  How do you manage something like that?  Well, it's quite complicated.  And it probably needs a variety of approaches, which include private property approaches and include commons approaches where appropriate.

It's important, I think, to have civil society engagement and policy engagement on this issue because the consequences of bad spectrum management are quite profound.  Bad decisions in spectrum management tend to take a generation to fix.  You know, failure to make smart decisions in spectrum allocation, because industries build up around spectrum allocation and incumbents become entrenched in those allocations, fixing them takes a very long time, so doing it right is very important.

At the same time, failing to do anything is also a high cost.   Because spectrum is so valuable these days, taking too long to make spectrum available is also a very expensive error.  But not wanting to take too long in doing it.  Things are changing very quickly in terms of the technology for the use of spectrum.  We live in an era where technology is getting more and more efficient on the order of every couple of years so that, you know, what we are able to do with the same bandwidth steadily increases.  At the same time the technology to do that is getting cheaper and cheaper, and that means we may need to think of different paradigms for using spectrum, less in the kind of monolithic sort of top‑down centrally controlled approach to rolling out spectrum and a more entrepreneurial locally driven approach to using spectrum, and all of these things have an impact on affordability of access, on rolling out affordable rural access, on the development of entrepreneurship and indeed in innovation in the use of these technologies.

So I think that's all I want to say at the present.  It's really just to try and kind of frame the discussion around spectrum and the importance of spectrum regulation and increasing access.

>> CLAIRE SIBTHORPE:  Thank you, Steve.

Next we'll hear from Willie Currie, a counselor on the Independent Communications Authority of Southern Africa, CASA.  Willie, we would like to you look at what is the context of spectrum regulation in Southern Africa and how is it changing in the light of digital migration.

>> WILLIE CURRIE:  Thank you.  Well, at the moment one of the major projects that's underway in South Africa and in Southern Africa is the digital migration of analogue TV stations to digital terrestrial television, and this process in South Africa has certainly been running almost for ten years, I would say, starting off with early investigations, digital migration committee in the early part of the century, and then progressing to a formal digital migration policy around 2006.

What I think is significant is the dependency of moving the TV broadcasters from analogue to digital to free up the digital dividend of high value spectrum for other uses and I think particularly in the African context for wireless ‑‑

(please stand by).

>> This digital migration process is a form of illumination, so the UHF band between 470 and 862 megahertz is going to be used both for digital and for maintaining the existing analogue channel foreseebly into 2015 so that then has a bearing on when spectrum may be released as this digital dividend.  We've seen that the world radio conference of 2007 at the ITU has set aside the top portion from 790 to 862 megahertz for IMTS or wireless broadband access.  There's currently a process around looking at another 100 megahertz below that in terms of WRC 2012 and leading to 2015 to release another batch of spectrum for IMTS or wireless broadband.  So in a way I think there is an opportunity because of this rather prolonged process of switching the analogue TV stations to digital of seeing whether there's an opportunity in Africa, in Southern Africa, for the development, the use of television white space technology as a form of wireless broadband provision, and I think certainly in South Africa we are about to be tasked with running an inquiry into digital dividend spectrum and one of the things the Government is asking us to look at or will be asking us to look at is possible uses of television   white spaces, so that's roughly what's happening in South Africa, in the region, through the Southern African Development Community discussions, into ministerial debates on digital migration, the association of Southern Africa regulators, the cross of the communication regulators has this issue on its agenda, so I think we're starting to see a real push to make digital migration happen after a long gestation and debates around standards with Brazil and Japan and the possibility of starting to get our heads around what uses there could be for digital dividend spectrum.  Thank you.

>> CLAIRE SIBTHORPE:  Thank you, Willie.  I should have mentioned that we are going to ask for experience from a few different regions and then open it up for discussion, so next I would like to ask Carlos Afonso, who is a member of the Brazilian Internet Steering Committee, he's also an advisor on the APC Open Spectrum Initiative to talk about the context of spectrum regulation and digital migration in Latin America.

>> CARLOS AFONSO:  Sorry to be late.  I was supposed to be in two workshops simultaneously, and I took the last 15 minutes or so trying to find the other one.  Just to say a few words and leave.  So I am here now.  Sorry.

The Latin American perspective, we have a regional forum of regulators.  I don't know if you can call it a regional regulator association in the way we would imagine an association that can frame or propose common rules, et cetera.  It's basically a forum, so the individual regulating agencies are relatively independent, depending on the country.  We analyze particularly five countries.  In two of them the regulators cannot ‑‑ in three of them, the regulators cannot be considered not even formally independent, in one case it's formally established that it's subordinated to the minister of communications.  In other cases they are independent supposedly, but not necessarily so in practice so I don't think this is a picture which is much different from the regions in the world, they are subject to political pressures and of course the economic interests of the large operators.

I know we have short time, so straightly jumping to the thing of the digital dividend, I have seen that several of the countries have adopted a similar standard which is derived from the Japanese, and embedding in the standard, the possibility of being interactive with the final user.  The main problem here, if we are counting on the digital dividend to get access to these bands for some, say, community purpose or some other application, it will be very difficult because the current owners ‑‑ frequencies, the large broadcasters ‑‑

(please stand by)

>> ‑‑ as a decision of regulator, you know, in, say, partnership, quotation mark, with the operators or whatever, decisions that preempt any initiative from our part or on the part of civil society, other groups or other interests to take hold of these frequencies.  So it's important for us to watch this practically.

The other thing that I think is important is, as we already talked about, the so‑called white spaces, I think in several countries these white spaces   are an opportunity now which we can try and find ways to start at least experimenting with them and showing, say, pilot projects and so on, this could be done, I think, in at least a few countries, and I know that in Brazil there is a strong possibility to do that, probably Venezuela, from the dialogue we have with the people who did reports as well, I'm not sure about Colombia, but there is an approach from the agency which has an advisory regulatory rule with the minister of communications ‑‑ with the Minister of Communications that is trying to look at capacity building on the spectrum   which is unique, I've never seen any other Government agency trying to do that, capacity building for civil society, for any organization that wants it, for people to understand the issues, to understand the technologies involved, the problems involved, et cetera, and it seems very open, given the opportunity for the organizations to suggest issues with discussion ‑‑

(please stand by).

>> Experience shows that if you do not have a fiber network to carry the bulk of this traffic, you won't do this only with radio in most countries, in most countries, and in many regional areas or even in big cities, in he very large cities.  So what we notice in the five countries in which we had the study developed is that all of these countries are in some way or another taking seriously the issue of developing a national backbone infrastructure based on fiber.  Brazil is already with 20,000 kilometers of fiber and is going to be 33,000 kilometers in 2014, connecting 80 or 85% of the municipalities.  Venezuela is doing the same, Colombia is doing the same, Ecuador is doing the same.  Argentina has a major set of rings, backbone rings, interconnected, covering the entire country being developed right now and with fiber connections to neighboring countries like Chile, Bolivia, Uruguay, et cetera, with the idea that you have of to have points of presence of this network in the community, at least in the municipalities, for the transit to be viable without problems of bottlenecks and so on, and is some of them are also tying this to the international connectivity, for instance, the Brazilian backbone is already building for two undersea fiber cables to Europe and United States with plans to do a third one with South Africa.  This I think is very important to follow as well, because, you know, you have to combine this, the wireless facilities and projects and networks within the communities or small towns or villages or whatever, with the possibility of to the Internet, the Internet or the international transit.  Regarding this, one important ‑‑

(please stand by).

>> This is a very key point because none of the reports show that there is a policy in this direction, except Brazil because of the particular fact that I am a boards member of the state company in charge of deploying the fiber in Brazil but the other ones we couldn't get a clear word on what will be the way that local providers of any type will get access to points of presence to run wireless networks.

>> CLAIRE SIBTHORPE:  Thank you very much.  I'm now going to turn to Muriuki Mureithi who is an independent consultant in Kenya and who did research on spectrum management in Kenya, and Muriuki, I wonder if you could sort of tell us ‑‑ to be implemented in a country like Kenya.

>> MURIUKI MUREITHI:  Thank you, Claire.

(please stand by)

>> We didn't talk about Internet those days, today we have it, and about 35% of the Kenyans are able and using Internet.  Also, we ‑‑ (please stand by)...

>> The Government or the Government agency will say yes, we are allocating this frequency for this and we are allocating you to do whatever we want to do on that frequency, so the key thing is the issue is technical parameters, is the issue of interference.  And this has been there for a long time in our country, historically it's been there and this is one of the reasons why we are seeing some of our companies buying up smaller companies that have got the spectrum.

But one of the other challenges arising from it is the case of the consumer, out there you buy frequency at the same price, you use it, it is therefore the same cost structure when you go to the rural areas and population is dispersed, it means that user will have to be penalized by that process.

Another way of looking at it which the regulatory is looking at is market.  Yes, if companies are prepared to buy off smaller companies with the spectrum, so can we be able to realize, discover an economic value, economic light and so in the process therefore.

(please stand by)

>> So long as it supplies technical criteria.  So with this, our regulator is looking at how can we put out more spectrum to solve our ever‑growing ICT system?  So they are looking at three processes.  One, the Government historically has a lot of spectrum for historical reasons, they have a lot of spectrum, so the regulator in the strategic plan current is looking at ways to work with the Government, perhaps even at some incentives, to migrate some of the Government agencies, to be able to release some of the spectrum.  The other area of course mentioned area by Willie is the digital migration, and by next year we should all be digital, there's a lot of work going on in that area, so the intention is how to release part of that spectrum to solve the ever‑growing need.

The other area also mentioned by our regulator under consideration is also setting some of the spectrum, but as I said, we still have to address this issue of the impact to have downstream, the impact to have contributing the market.

And finally, the market, the Government is looking at, is frequency allocations, and also regular frequency spectrum audits, the intention is to see tracing this resource properly.

So all these models are in discussion at various stages by the regulator.

Now, one thing that perhaps I would like to note, this is based on a study from last year, that spectrum appears a very technical issue.  The effect it has in the use, so it is not yet ‑‑ (please stand by).

>> CLAIRE SIBTHORPE:  At this point, just for any points of clarification?  Okay.  Then I'll proceed onto our next speaker.  So we've heard a bit about what's happening in different countries, and now we're going to broaden the discussion out a bit again, so next will be Sascha Meinrath, director of the numeric foundations open technology initiative, and Sascha, I was wondering how you could say how new technologies could contribute to a shared use experiment.

>> SASCHA MEINRATH:  Sure, I would be happy to tackle that.  I work in a DC think tank working on cutting edge policy but I'm also a telecom historian, so I thought I would start with a simple quote.  It is probable that many of the greatest inventions and improvements of the future will come from amateurs who by experimentation chance upon undreamed of things.

A book called the home radio, how to make and Tuesday, from 1922.  You can imagine in that era, the 1920s, probably one of the most vibrant innovative decades for electromagnetic spectrum use in human history.  It pioneered the broadcast model which has been with us now for 80‑plus years.  It led to most of the inventions and quite honestly most of the spectrum policies that are still with us today.  So I really want to focus on this notion that we have analogue regulations for a digital world right here, right now.  We have radio, we still have guard bands and guard zones, we have exclusive use, we use a broadcast business model that doesn't map onto the possibilities of today's technology.  Systematic enclosure ‑‑ (please stand by).

>> We both can operate in the same space without causing harmful interference to one another.  That exact same phenomenon is entirely possible in broadcast media.  A number of us fought for this proceeding.  Eventually the FCC decided not to rule on it, and it is now dead.  But it could be resurrected and certainly could be implemented in other countries around the globe.

Number 2 is what's pending today, television white spaces and I certainly hope Paul will spend a little time talking about this because I think it is right now a technology that is ripe for implementation, that could roll out affordable both broadband connectivity and local area networking to areas of the globe that otherwise couldn't afford the dominant business model or have needs or uses that they want to put these technologies to that aren't met by the dominant business models.  Number 3 is what I will call a proposed solution, which is the unlicensed GSM bands, something that I talked about a little bit yesterday.  But this notion that if you had unlicensed bands, you would provide excess capacity for those that need it in areas of congestion while still making it available to communities, entrepreneurs, individuals that want to use these yous spaces in places that is are under or unserved, it would cost us almost nothing, weave seen it demonstrated in the AT&T's and oranges of the world, offloading capacity to wi‑fi hot spots, Steve and I were talking between 35 and 40% of all AT&T offload traffic gets to an unlicensed band, that's great, let's do that all over the place but let's not give you the exclusive license, which is what I constantly hear everybody talking about today, give us more licensure of additional bands and somehow everything will be great.

Well, tell that to all of the people that can't afford the dominant business model.  Tell that to all the people that have been left unserved, the poor, the rural of the world that haven't yet received connectivity.  So all of this falls under, I think, a simple rubric, which is that what's before us today is the shared use of electromagnetic spectrum and I think it's the single‑most important issue facing spectrum regulators today, the single‑most important issue.  And the real question is, are we going to learn from history or are we simply going to repeat the mistakes that we've made in generations past.  Thank you.

>> CLAIRE SIBTHORPE:  Thank you very much, Sascha.  Next we'll hear from Daniel Wilson, a senior policy advisor at the BBC, I was wondering if you could comment on the state of debate around use of spectrum in the UK.

>> DANIEL WILSON:  Thank you, Claire.  First of all, I want to say what an excellent paper I think Steve's paper is and I think it speaks exactly to what Mr. Mureithi was saying before about the importance of bringing to the public domain, I think someone who was brought to the panel late in the day, it was a reminder to me of some of the history of this.

The UK soften taken as an example of market based approach through the spectrum auctions that are regulator has led, that's been seen as an example across the European Union how spectrum can be managed but I think it's important to notice that that is only the spectrum auctions are only around part of the available spectrum and they're only part of the debate and the debate in the UK and in the European Union looks not just at the economic value and maximizing the economic value but also the crucial public value of spectrum, because after all, it's a very scarce public resource, and it's important in all these decisions to factor in the social, cultural benefits that come from spectrum use.

So the BBC can certainly appreciate the value of social as well as economic value of using spectrum for mobile operators, broadband operators who after all are most likely to be the highest bidders for available spectrum that competes in the same bands as broadcasting.  And the Internet is a core platform for the BBC.  We're investing 600 million pounds until 2017 in broadband infrastructure as well as 100 million pounds a year in broadband content, so it's a crucial platform for us and definitely one which will be ever‑more important.

But it's also important to notice the disadvantages which come with allocations or with auctions one for mobile use, for Internet use.  First of all, it's a one to one technology using IP, using television is a one to many technology, it's a merit good, so the more people who use television, it comes at no extra cost, and you get the social benefits of something which is universal, free at the point of use and something which in the UK is in 19 million of the 25 million UK households, so it's something which can really deliver social value and bringing the nation together and getting out social, political, cultural, valuable content which Government has decided is something and society has decided is something that's worth funding to the whole population at no additional cost in transmission costs irrespective of how many users there are, and I think that concept of universality is very important because it's not something which is very easy to deliver in many areas.  We've heard about the problems of rural broadbands before.  It's still very much an issue in the UK striving towards universal broadband access and high quality access, but the cost of reaching the last 10%, the last 1%, is extremely high, and Governments are looking to invest in doing so.  They are needing to subsidize it.  But it's something which, you know, you can see that different technologies have their different advantages.

Just going back to the UK history of how we got to the market based approach, we had to switch over, it's due to finish next year, shortly after the Olympic games in London, and alongside the switchover there was a massive Government marketing campaign but also social help for those who might be left behind, 600 million pounds put aside to help the elderly, the disabled to switch to digital, and again, when looking at spectrum allocations, broadcasters were very keen to promote digital television.  It's far more efficient.  We can use more channels aside from all the potential use for broadband.

But you do need to think about the people who might be left behind along the side.  There's also the interference issue which is quite large in the UK debate at the moment, so the 4G spectrum which is going to be auctioned off in the UK next year hopefully, but it's been delayed for many years through litigation from the tell co‑s largely, but ‑‑ from the Telcos largely, but the regulator, think that might interfere with 3% of digital television sets, which when you think 3% of the 90 Mill yon homes is a lot of people, a lot of people again who might be the least able to adjust to this change, so again, I think the key there is factoring in that social context.  So the UK is switching next year, the European Commission hopes all countries will switch by 2012, I believe, although I'm not necessarily sure that will happen in every country across the European Union, and we're now looking at whether there are potentially second digital divisions in Europe, and I suppose to come back to my core argument, also at looking at whether in developing countries you might potentially leapfrog digital television and go direct from analogue to IP TV, you have to think about the benefits of the universal access, the free access to say digital terrestrial television that isn't pay access to contents which society has judged is valuable, for both social democratic purposes.  So absolutely I think that spectrum auctions can be a key part of the solution, but you have to bear in mind the trade‑offs involved and factor that into the decision‑making process from the very start.

>> CLAIRE SIBTHORPE:  Thank you very much, Daniel.

Now our final speaker is Paul Mitchell, who is a senior director of technology policy at Microsoft.  And I was wondering, Paul, if you could comment on the opportunities for multistakeholder collaboration around spectrum use and management.

>> PAUL MITCHELL:  Sure.  In fact, this panel right here is a good example of the multistakeholder need to address spectrum holistically.  One of the challenges that have been presented here are the need for universal access to content that's preserved by the broadcasting model.  We've heard about the difficulty of actually building out last mile broadband.  We've heard about interference problems.  When you come right down to it, though, we have a situation where in most of the world n reality, most of the spectrum is actually not used most of the time.  What we have is a set of allocations based on perceived public benefits, some of which are real and some of which are actually intangible and unknown, and allocations based on business models.  And as was earlier stated, there is a real need to sort of bring the models into the digital age, as opposed to leave them in the analogue world.  So when you think about multistakeholder, it is really something to address this, it is something that needs to involve industry, Government, the civil society in general, and needs to factor in all of the difference needs but use modern technology to come up with solutions for this century and the next century, not the last century.

As one example, we have been working with a number of partners, including the BBC, in a trial of white spaces technology in the city of Cambridge in the UK and the purpose of the demonstration which includes the regulator Ofcom, Nokia, BBC, Archiva, I'm sure you'll miss some, Samsung, Microsoft, who   actually doesn't make anything in this space, we're just trying to be a catalyst to make more broadband happen and the purpose of the trial is really to demonstrate that we can, as a collected group of participants, successfully use white space technology without interfering with anything to deliver on a range of scenarios, and those scenarios include regular sort of in‑city broadband, rural broadband, smart city applications, location to services and entertainment and media services are all use cases that are being demonstrations and trialed in the city of Cambridge.

Something else we've been doing is we have a set of spectrum observatories, which are essentially test and measurement units that every second sample all of the spectrum from zero to 6 gigahertz in this case, and they make a record of what's being used.  And we have two of these in the Puget Sound area near Microsoft's head quarters, one in Washington, D.C. and these have been up and running for a while, and in the city of Seattle we have several airports, we have several military installations, we have a very wired mobile aware society, we have every conceivable use of spectrum that you can imagine and again, when we look at the plots of the actual usage of spectrum we find that most of the spectrum in that area most of the time is actually unused.

Some of the models that we have today for allocating spectrum are roughly the equivalent of reserving spectrum for marine traffic in the Sahara desert, which doesn't actually make sense.  So in terms of the trial aspect, what we've tried to do is pull together all of the representatives of all of the associated stakeholders, not to presuppose a particular business model, not to presuppose a particular application, but to demonstrate that the technology is capable of doing amazing things, and if all the stakeholders can get together and actually recognize how these scenarios can play in society, that conversations change from protection of the status quo to how can we do more, how can we do it better, how can we help build out our broadband universe across the planet.  In this continent, using spectrum is pretty much the only way to deliver broadband to most of the continent.  The reality is that the cost of running fiber connections to rural Africa is just so staggering that it will in my lifetime certainly never happen.  On the other hand, using broadband, using wireless technologies efficiently, he is specially those that can be brought in at consumer cost curve economics, has the potential to really open up a world of opportunity, economic development, better healthcare, better access to education throughout the developing world and that's really what we're focused on, trying to bring the digital benefits of the digital age to the consumers across the planet.  So I'll stop there.

>> CLAIRE SIBTHORPE:  Thank you very much.  So now I would like to open it up for comments or questions.  Henrik, is there any remote participation?  No questions, okay.

All right.  If you could introduce yourself and where you're from.

>> Thank you.  My name is Daniel Labom, I'm from Kenya.  I will just start off with an example, the gentleman who said we have spectrum which is not used.  As I sit here, I have a computer which has an operating system, I have a phone here, the  phone has no plan, so I should give it to my neighbor to use it just because I'm not using it?  Another meeting I was in, my neighbor had an iPad, laptop and a phone.  So really, in real life, it's not possible to use things 100% of the time, but that's beside the point.

I'm just confused about what's the definition of white space?  Is it an internationally accepted definition or is it something that has been tried out in one country and therefore it's being exported, and is white space space between bands of frequency channels or is it spectrum that is going to be made available when migration occurs and we clear a sort of spectrum?  What is we need to define what white space is, because I've heard different people talking about white space and talking about totally different things that actually are confusing me.

I had the origin of white space came from the printed page, which is the space around a paper which has not been printed on, because it's not possible to print words on every single piece of paper, so even in printing, there's a reason why they leave white spaces so that you can be able to read the words printed on the paper.  Thank you.

>> CLAIRE SIBTHORPE:  Can I just check, we have a couple questions and then pass them back to the panel?  Are there other ‑‑

>> Thanks very much.  My name is Elizabeth Nuguella, I work for a company called Qualcomm.  I just have one question, and that employs to the ‑‑ that goes to the last speaker who talked about bringing these services at consumer price points.  I'm very interested in that because I would like to get a feel for what those consumer price points might be and how the technologies we are talking about today would improve on maybe some of the technologies that are today widely used by the populus in Africa.  Thank you.

>> CLAIRE SIBTHORPE:  Maybe one more question, then we'll go back to the panel?

>> Hello.  My name is Bella, I'm a member of APC.  My question goes to Sascha, I think.  You mentioned something about, I didn't get the full name, but something about low level temperature, something, I think maybe you can elaborate a bit more about this, in the U.S., maybe people something in the developing world can look at.  I think that's it.

>> CLAIRE SIBTHORPE:  Okay.  I'm going to go to the panel, then we'll open for another round of comments and questions.  So I don't know who wants to tackle what first.

>> I'll tackle the white space thing, just to be clear.  When I talk about white space and when anyone from Microsoft talks about white space, we mean the unused but allocated spectrum in a particular range of bands, so if we're talking about the TV bands, the spectrum in a given place like right here, that within that set of bands is currently not being used.  So white space in different cities in the television bands will differ, so in the United States, for example, as a simple example, in Seattle, channel 2, 3, 4, 5, 6 and 7 all exist, in Bellingham channel 2, 3, 4, 5 and 6 don't exist but 12 does, so I could use 2, 3, 4, 5 and 6 in Bellingham, so we had a demonstration until this afternoon and then the radios had to go back to the U.S.  The channels we used here, which we were demonstrating in the TV bands, were channels currently not being used by any broadcaster right here in this geography.

Does that answer your question?  And the other, you asked a question about price points, just quickly, the white space technology that our partners are working on is designed toly to the same cost curves as typical wi‑fi devices, so today a standard consumer access point is about $30, wi‑fi chip sets are essentially in every computer, every mobile phone, virtually everything, they're very very inexpensive.  They start at about 1,000 and then they went down, and the white space radios should do approximately the same thing.

>> CLAIRE SIBTHORPE:  Steve and then Sascha.

>> STEVE SONG:  I just want to speak a little bit more generally about white spaces, maybe expand the definition a little bit, in that white space technology I think represents the future of spectrum use in the sense that it is serendipitous reuse of spectrum using spectrum‑sensing techniques, so what we've gone from is spectrum regulation is based on the on early days of broadcast where the technology was relatively crude and you had to have a television transmitter that was shouting in order to reach the comparatively deaf television sets that existed back then.  In order to make that feasible in terms of having difference channel, you had to have guard bands separating those channels in order for there not to be interference, and the white space actually refers to, you know, on the old television sets when you get that noise on the television set, that sort of white snow, that's that guard band in between the television channels that was empty because of the sort of relatively deaf and shouting kind of technology that we started out with.

Where we are now, we have much more delicate equipment that can listen and sense if anybody else is talking in the room, and if somebody is talking in that room, if somebody is broadcasting, move to another room, and do it dynamically and automatically.  So rather than having to permanently allocate a space to a particular user, you can serendipitously and through technology and not through regulation or through policy reallocate and reuse empty bits of spectrum.

Right now the television bands are the real opportunity for serendipitous reuse but really this kind of technology could be applied across the entire spectrum band.

>> CLAIRE SIBTHORPE:  Sascha, did you want to ‑‑

>> SASCHA MEINRATH:  Sure, I want to touch on a couple other things, but I also wanted to touch on, spectrum is probably one of the only natural resources that no matter how much you use right now, renews itself to 100% the next millisecond.  This is important because actually, when you aren't using your phone, if your neighbor has the same phone company, he may be using that same spectrum, and in the same way that you can coordinate that use, you can actually coordinate not just with another cell phone with the same carrier, but across a number of different devices that may or may not be locked into that one carrier.  And the notion is just if you listen before you talk, if you analyze before you broadcast, which is what smart radio or cognitive radio technologies do, you actually can make far more efficient use because millisecond to millisecond you can infill those frequencies that aren't being used at that one moment in time.  And you can do so without causing any adverse effect the next millisecond.

In terms of interference temperature, so there's electromagnetic radiation all around us, right?  Light is electromagnetic radiation, and every single frequency you have signal and you have what's called noise.  It's the background radiation that's always with us.  The interference temperature is simply the level at which that noise would interfere with a signal.  So you can imagine, if you broadcast below what's called this noise floor, you won't cause interference to the main signal.  So what I was talking about earlier, how people can whisper in this room and I can talk and their whispering doesn't make it impossible for me to talk or for you to hear what I have to say, that same phenomenon exists in electromagnetic spectrum.  So what you can do is you can imagine in a television bandwidth, a 100,000 watt station broadcasting out, you could have a 10 milli watt device connecting a couple of things together with no possibility, no realistic possibility of causing harmful interference.  And this enables all sorts of other uses within the same frequency that a primary broadcaster may exist in.

I finally wanted to touch briefly on the average return peruser and business models, et cetera.  A lot of what I do, wearing my other hats in doing community wireless networking, we have found that the connectivity is so cheap, it does not make sense to measure.  That we're able to repurpose available connectivity, that we're able to provide models for sharing connectivity that drop the cost of broadband access to a level so cheap, we're talking about 1 to 3 cents per user per month, that setting up a system to actually bill people, that transaction cost is far greater than the cost of the connectivity itself and I think this is what's possible when you start thinking about other models for delivering connectivity and we just have to get there as a society in terms of thinking outside of the box of a specific kind, a specific way of delivering broadband access, et cetera.

>> CLAIRE SIBTHORPE:  Did any of the other panelists wants to comment on the questions that have come in?  Carlos?

>> CARLOS AFONSO:  ‑‑ when at MCI it was said to us that if they could go back in time, they would never charge for long distance costs on the basis of measuring the minutes because the cost of doing that with the automatic switching, et cetera, would not any more, you know, be worth it.

So even at that time, you know.

>> CLAIRE SIBTHORPE:  So I would like to open the floor to the next round of comments and questions.  We have one remote and two people here.  Maybe we'll start with the remote, and then the two gentlemen.

>> Sure.  This is Lillian from Colombia, she wants to know what could be the role of the regulator if a policy of the spectrum is accepted.

>> CLAIRE SIBTHORPE:  Gentlemen, do you want to come to the microphone?

>> Yes.  My name is Tomas, I am from Open Society Initiative with Southern Africa, my question is to do with the cost of digital migration.  What would it mean for poor countries that are battling, 2015 is the deadline for migration, it's also the deadline for MDG's, so what are the chances that most African countries are likely to be left out, and what would be the implications?  Thank you.

>> I'm Yoshi from a 3G mobile operator and also wholesale provider in Japan and also I'm director for Japan ISP association and one of the organizers of IGF Japan.

I would like to make a couple of comments based on our experience starting from a DSL operator to 3G.  Firstly, we had many discussions among operators and also had a chance to exchange information with several carriers in Europe, that basically what you see is that especially because of the use generation, which catches up in this digital world very fast, that after introduction of broadband and Internet to the community, it seems that the demand for any ‑‑ I mean, the traffic demand for any person is not too different from any country to another country.  That was from our experience when we started ahead of the British telecomm, DSL, that after two years they caught up with our demand, then we saw a similar traffic pattern, traffic pattern means per application and similar traffic pattern per hour and similar volume of traffic per person.  So I was in one of the other workshops where we were discussing IX points in Africa, and what I noticed was that still the current African Internet traffic is pretty much research and business, in other countries the peak traffic is 10:00 or 11:00 p.m. in some cases midnight, so I think there will be a quick transition in this region that you start from this business and research oriented traffic to consumer traffic where the volume is much different and the peak hours are much difference and if you reach that level it will be something around 50 to 100 kilo BPS per second subscriber.

Secondly, from our experience where we are handling the similar size of document traffic under only something around 10 megahertz of spectrum, where a document has more than ten times of their spectrum, there's a big difference between circuit switching and packet switching on, how do you say, the wireless network, so that it seems that if you move to pure packet switching on wireless networks, such as HSP or MTV you can improve the real spectrum efficiency more than ten times, meaning that you have to consider how you handle voice under the current spectrum usage because there's much, much difference in the usage, and from our experience, especially, you know, that the most challenging is the dense population in Tokyo, that under such configuration where you have many buildings which have like 20 stories, 30 stories building, and when you come to really the technology limits of about 100‑meter radius for base station, which is very very dense, that you can only handle something around a million or 2 million users per 5 megahertz pair.  So of course the population density varies from one country to another country, but once you have these high story buildings and then you come to the technology limit of base station coverage, then you'll see similar size because of the technology and also the demand, which is about 50 to 100 kilo BPS per second per user.  And secondly, at that limit, although these 700 to 900 megahertz traditionally looks like a pretty good spectrum, at that short radius, something like 2 gigahertz will also work in a similar manner than the lowest spectrum.

But which means that there is a need for different policies and different maybe business model in the rural areas and in the city, and from our experience, especially in Japan, where the number of basically mobile users and even the number of mobile Internet users are nearly saturated, the customer demand is of course flat rate because they don't want to pay per volume, and at this stage what happens is that whatever you improve, then there will be no more income coming in for an operator, so at this stage the thing that will survive is a switch of the business model, that is why a lot of Japanese operators are moving to wi‑fi offload, then there's a challenging issue whether you can offload to a public or private access line which does not belong to a private operator, so there are many complicated legal and business issues at this level, which I expect that even in this region, looking at the development of large cities, that in several years it will come to this stage very quickly, so if you don't expect that this transition, very quick transition in a small number of years to happen, then at some point what you'll see is saturation of the supply or maybe even customer complaints or have your troubles coming in very quickly what you see on some of the cities where the traffic on the road is so bad.

>> CLAIRE SIBTHORPE:  Thank you very much.  I think that time is going run out, so if anybody has any very quick question, very quick comment, then I'll pass back to the panel to sort of go and make their final comments.

>> My comment is related to the comment that we probably don't need the ‑‑ to have access to content making these white spaces, but generally in rural areas, we don't have content, contents resides elsewhere, so I think we still need to build the backfills, especially to a school, you need fiber or high capacity back holds.  Secondly, I'm involved in migration in Kenya, and a survey was done, and out of the 5 million TV sets, 53% were black and white, therefore, those nice devices that we're being told, they sound like cake but we just want bread.  The time will come where we can have the cake, but for us in Africa we're grappling the basic case of first migrating use the current screens, current broadcasting technology, then after that I think we can have successive levels of migration.  Thank you.

>> CLAIRE SIBTHORPE:  I'll ask each of the panelists to make any final comments and then maybe go around the circle, starting with Willie.

>> WILLIE CURRIE:  Well, just on the costs of digital migration, yes, there's the cost of the migration itself but there's all the opportunity cost of ‑‑ there's also the opportunity cost of not migrating in the next five years which is not then having access to the digital dividend in terms of mobile communications, buts on the other hand, strategic conduction of white space technology might counteract that.

>> I want to address the question from Colombia, what is the role of the regulator.  I think the role of the regulator is on the one hand to bring about public engagement from industry, from Government, and civil society on these issues to make the decisions that are in the best interests of the country and I think the current discourse that tends to go on at the regulator happens mostly between industry and Government and there needs to be a stronger component of civil society involved in that.  But on the other hand, the role of the regulator is also to ‑‑ is basically to fail safely because what we've seen in the history of spectrum management and technology, you know, these facts about the 80% or more of the spectrum being available, the reason that is true is because either we misunderstood what technology was going to involve or we misunderstood the person that we gave ‑‑ or company that we gave the spectrum to, so we look at those complex spectrum allocation charts, they're filled with technologies that don't exist anymore, and, you know, I'm always chasing, reminding myself of events like this, you know, that not ten years ago, it is word mobile phone wouldn't have featured in this discussion five years ago, it would have been Y maximum, two years ago nobody would have brought up the issues of tablets at this discussion, so I think we have to be very humble and therefore design policies that are very adaptive and that recover from the mistakes that we'll inevitably make.  Thanks.

>> CLAIRE SIBTHORPE:  Muriuki any comments?

>> MURIUKI MUREITHI:  Very quick, it is not feasible that all of the spectrum can be out of the commons, what we are having and perhaps have for a long time because of various uses will be the combination of three models, so perhaps we need to start thinking about moving away from calling them allegra.  Thank you.

>> We're going down the line, I guess.

I just want to plus one that comments, I think, you know, what we're really pushing for is a rich ecosystem of different spectrum regulatory frameworks, so, you know, we have what I work on is actually seven different options and actually if you Google seven key options for spectrum, you'll pop up some of my work as your first hit and the notion is you want all of these different elements, you want licensed, you want unlicensed, you want light licensed, you want interference temperature, you want all of this in the space.  That's what gives you the most diversity, which enables you to meet the most different needs, which are also equally diverse.

When I think about the different tech and business models, it's important to keep in mind how spectrum regulation has come about, and for better and for worse, most of it starts in the United States and it usually starts in a major conversation at the federal communications Commission between a very few civil society folks and a huge number of incredibly wealthy, incredibly active, massive U.S.‑based telecommunications groups who are trying to figure out how do I maximize my profit, and then the rest of the world kind of harmonizes around the decisions that the FCC makes.  For most of the world the decision model those decisions are based on doesn't fit your reality, so you have to ask, like, would the regulations that were made, the decisions that were made based on that business model being pushed at the FCC map onto my on the ground reality?  And if your answer is my community isn't like suburban Washington, D.C., you may want to think a bit more about what regulations, what key decisions better map onto my country.

>> PAUL MITCHELL:  I guess it would be the plus two on that.  I would just add that a key role for the regulators right now is to facilitate innovation and discovery and discussion and not get in the way of trying out these new technologies but actually at that to sill at a time ways to expand the understanding and expand what can be done, and as Sascha said, if your reality is need for educational facilities in a rural region that doesn't have broadband, what's the best way to get broadband there, if your reality is a lot more television, more television.  But really think about the society needs and from the perspective of where we are today in time and where we're going, not where we've been.

>> DANIEL WILSON:  Thank you. I think one of the interest things that happens in the UK switchover was the emergence of groups which had initially been overlooked by the regulators, so particular use of wireless microphones, not be just by broadcasters and cynically you might say the reason that enough notice was taken of this community because the broadcasters made their views very clear and added some muscle to the community groups who raised this issues, but lots and lots of theaters, small community groups, who use wireless technologies and might have been overlooked in the process.  So in the sort of cost/benefit analysis that you do of migration to digital, you have to factor that in as well as it is broader economic benefits but also social costs of those technology moves, whether it's buying new equipment, but also the loss that previously was, but all sorts of potential for better use of spectrum and I think one thing that certainly I hope is more widespread than just European brought casters, is a duty to use them efficiently, so we're always trying to come up with new standards, try out while space, like we are in Cambridge, also look at standards like DBD2 which is a standard which has allowed us to launch more television channels but also spectrum which could have been used for other uses, too.

>> CARLOS AFONSO:  I think I can compliment a bit what Steve and the speaker from Kenya and Sascha said, regarding a question, I think another thing is that we have to make sure that he we do not lose what we have already concurred, which is the various parts of the spectrum which we can use as no license or unlicensed or to a certain extent open spectrum.  As an example, in the wi‑fi world in some of our countries you have a limit of transmission power of like 400 milliwatts.  Also you cannot transmit beyond your premises, your legal premises, so you cannot transmit beyond your house or your whatever, your school.

And then if the community association wants to use this technology, it doesn't pay any license to cover the entire community using just 400 milliwatt radio, they can do that, but they will have to pay a license, so we lose the open spectrum characteristic already that we have.  So we have to fight for a community that can use what is open, unlicensed, not to have to pay license, to operate a small network in their communities.  There are some cases in Brazil for which people are getting arrested because of that, they are not charging anything, operating 400 milliwatt radios, but the signal is going beyond the premises and someone else is using it, with their knowledge.  They cannot do it.

So this I think we have to preserve, besides trying conquer more terrain for the open space.  That's my point.

>> CLAIRE SIBTHORPE:  Thank you, everybody.  I guess it's up to me now to just do a few kind of summary comments.  What I sort of picked up in this I think very interesting discussion was that it's a time of change, technical change, and a time of opportunity and that the regulation of spectrum is an issue that regulators are currently really grappling with, and I think not one where there's any finding any easy answers, also that more people need to be involved in the discussions about spectrum so that their interests are being heard.

But that it's a time of opportunity where with changes in the way that we manage spectrum, you know, it offers opportunities to get those connected who aren't connect and had to use the spectrum more effectively and that a key dimension to this is flexibility in the way that spectrum is managed.  Hopefully that sort of sums up some of the key points.  I just want to thank very much all the panelists for their time and for Henrik who did the remote and Pablo who pulled together the speakers, I hope everybody found it useful.  Thank you.