Welcome to the United Nations | Department of Economic and Social Affairs

FINISHED COPY

 

NINTH ANNUAL MEETING OF THE

INTERNET GOVERNANCE FORUM 2014

ISTANBUL, TURKEY

"CONNECTING CONTINENTS FOR ENHANCED

MULTI-STAKEHOLDER INTERNET GOVERNANCE"

 

WS51

CONNECTING THE CONTINENTS THROUGH FIBER OPTIC

04 SEPTEMBER 2014

11:00

 

 

 

*  *  *  *  *

This is the output of the realtime captioning taken during the IGF Istanbul, Turkey, meetings.  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. 

*  *  *  *  *

 

>> MODERATOR: It is always interesting with arrows and techs on microphones.  Welcome to this workshop.  About fiber build out and advisory in the world, specifically focusing on the Persian region.  We will also, as you will see, cover fiber build out both domestically and internationally in other regions as well. 

My name is Patrik Fältström, Head of Research and Development for Net Nod in Sweden and I'm happy to moderate this session.  So the panelists we have here are from our right, is Michael Kende from Internet society.  Alireza from Iran.  We have Kursai (spelling) from Kuwait.  We have Colin from Fiber to the Home Council, Europe.  And we have Tracy Hackshaw from Trinidad and Tobago.

And each one of these people will start by giving an overview of their situation and what they are working with and their connection to fiber build out.  And after that, we are moving into a discussion on similarities between the regions, differences between the regions, what arguments are to be used when you want to build out, what problems there might be with build out, and I hope to get interactive session that also involves you and the audience.  I would like to start by giving the word to Karin, Chair of the Fiber to the home Council of Europe, please give an interaction.

>> K. AHL:  Thank you for inviting me to this panel.  I'm the Chair of the Board for Fiber to the Home Council Europe.  But that doesn't mean I'm only focusing on Europe.  We also have a global council which have representatives all over the globe.  In this specific region, the Europe Council covers it and also the Amina (sp) council covers it.  We do it together. 

I'm looking forward to the discussion later today because I heard rumors yesterday that some made a statement about is fiber dead?  No, fiber is not dead.  I don't think we would be in this room if fiber was dead.  And actually, we've never seen so many operators in so many countries and so many network owners investing in fiber.  Maybe not fiber to the homes specifically, but fiber networks, fiber backbones and fiber between the cities, fiber between the offices and so forth.

I'm representing the organisation which gathers fiber optic manufacturers and component manufacturers.  Of course we have a very profound technology background but we also cover policy implications, regulatory issues and investments.  Investment focus is on something that really is coming up now as one of the big target questions for us.  So I'm happy to dive into that.

Last week there was a report launched in the U.S. stating the U.S. is much further advanced in fiber to the home than Europe and other regions.  I think that is very interesting to see.  We do this in different stages in different regions.  I think that is interesting to discuss.  Europe is not lagging behind as everybody says, but we do it in different stages and we use different technologies to reach the end goal.  For that, that is difficulty fiber to the home.

So I am happy to be here today.  I will definitely dive into the other presentations and discussions.  And don't hesitate to raise a hand and say a question.

>> MODERATOR: Thank you very much.  Yes, I also saw the report coming out and of course for a person in Europe, and specifically Sweden, that the report points out should be one of the best countries in the world.  But in that report, we were not at all.  So that is something that we should talk about a little bit more.  Over to you, Qusai Al‑Shatti.  Can you talk about your region

>> Q. SHATTI: My name is Qusai Al‑Shatti from Kuwait Information Technology Society , an NGO based in Kuwait with advocacy on IT, on the Internet.  I just have a map.  And I'll be talking about our experience in the Gulf and Middle East regarding marine cables. 

I have a map here, it's a bit possibly not as clear.  If I take the case of Kuwait, we are connected to 4 million cables.  One we call it the fog and one we call it the GCC falcon and the other one is called the flag cable.  And that is one direct cable between us and Iran.  It's a cable that connects Kuwait and Iran.  The importance of the marine cables for us is that they played a key role in lowering the access cost for Internet.

In the 90s, for a dial-up service, we used to pay monthly 200 dollars for 56K.  Today, we pay $600 a year for 10MB  and actually on the mobile, it is $600 for 20MB.  So that is what the marine cable does.  If we don't have access to the Internet, there is no need to talk about Internet Governance and so on.  Access is the basic thing.

Now, regarding the marine cables, when the Gulf countries did the first marine cable, the fog, they did for a capacity of 5 Gig in the late 90s and they did for voice, not for data at that time.  So it was useless for us and it didn't help.  But then came the private sector and they invested in doing what we call the flag and the pocket.  The flag is around 1.25 Tera.  And the falcon about 5 Tera.  And then we needed to make a regional connection with countries across to us like Kuwait and Iran.  And that cable has the capacity of 500 gig today, almost.

And instead of going marine, land line with Iraq, actually we went with Iraq through the marine cable, which is the falcon.  So we have a redundancy with Iraq through the falcon marine cable.  Not only that, as strategically for that, we need a land line cable.  This is where we did the last cable between the Gulf countries which was called the GCCX and it was done originally because the Gulf countries created a power grid, utility electricity power grid system and they needed to exchange data through that so they needed to do a land line cable.  And strategically enough, it not one only connected through the Gulf but also to the red sea so we can have that connectivity to Africa.  It has a landing point, one in Sudan and one in Djibouti.

Now for countries like Kuwait, Iraq and Iran, where these cables does not connect us with the whole world.  The maximum it can connect us with India and to Egypt and possibly one of them actually, the falcon, connects us with Italy.  The cable itself.  That means we need to depend on other international marine cables like the simway 3 and simway 4 and soon‑to‑be simway 5, to connect us globally to the Internet, which makes us vulnerable.  These cables, when a ship hook hits these cables and takes us out of the net.

In 2008, for example, there is a famous or well‑known story about the simway 4 and the simway 3 where a ship hook did not only cut off simway 3, it cut off simway 4 too.  And for 36 hours Kuwait was out of Internet connectivity. 

It means that these cables are not the sovereign right of the nation that it passes through if it is going through sovereign waters.  It is also the right of other countries, which they don't have because of their geographical location, only to connect to these cables. 

So what we need to do here is, how we can create a framework, a public policy, where these cables, which is layer one, the basic thing, without there is no Internet, no connectivity and no data, where all of these ‑‑ if something happened to that cable or the operation of that cable, the accountability is towards all the countries, that is these cables are serving.

And this is an issue that we need to address.  When a cable is cut off and the sovereign -- Egypt, it is not only Egypt's problem, it is our problem too.  We will suffer more.  So, this is an issue that we need to be addressed at least from our point.  And this is as a security issue more important than after security in my opinion.  Thank you.

>> MODERATOR: Thank you very much.  So you're talking about Kuwait being disconnected from the Internet.  Sweden was disconnected in the Internet in Kuwait as well.  So I think it is important, myself, to remember that what we are talking about is that what happens is that a cable cut impacts the ability for the two parties on both sides of the cut to interact with each other.  So, next person, what do you have to say about what is going on.

>> AUDIENCE MEMBER:  My name is Andrew, and I'm from Iran.  I'm just want to talk on ‑‑ I'm not working at the infrastructure but because I'm living in Iran, also I have some information regarding the current situation.  I have in Iran regarding the accessibility of the people and how they can connect to the Internet.  So technically, we have ‑‑ I want to tell people about the situation that we happen how people actually in a major city connects to the Internet in Iran.  Most of the users using the ADSL up to 8 Mb per second are connected to the Internet.  They are using the copper.  They are ADSL over copper and then it goes to the TeleCos and then the TeleCosconnect to the Internet that mainly relies to the government and to the fibers and the fibers then goes out and connects us to the Internet.

Previously, we had one 3G operator in Iran that can connect to the Internet by 3G and now we have two others joining them and now we are seeing that the full operator is coming and they are going to build the fibers as a pilot project for 8 million Iranians that connects into the Internet through fiber and the LTE over the GSM.

So this is the current situation of that accessibility in Iran and also I'm not the only problem we have according to the basic rules of Iran, all the infrastructure has to belong to the government.  So the private sector investment actually, most of them are belonging to the government.  So this is a thing that is happening in Iran and a thing that if you want to see some kind of a change in the infrastructure, this has to be a little bit changed.

There are some companies that is running the fibers.  One of them is the Ministry of TeleCo that does the fiber and also the Ministry of Energy and the rail base system.  All has their own fibers and they connect to each other.  And actually, our entry points are from Dubai and the other to Turkey.  So this is the current situation and if there is some more information, you want, I can help you and take questions and answers.

>> MODERATOR: So you say is that the private sector companies that compete with each other build services and top all the government‑owned passive infrastructure. 

>> Exactly.

>> MODERATOR: But on top of that, you have multiple entities that provide this passive infrastructure even though all of them are government controlled.  So to some degree there is also competition between those providers of fiber, right?

>> Exactly.  And sometimes they are just renting each other.  For example, if they want to increase the capacity connecting to the Internet, if the Ministry of TeleCo wants to increase the capacity through the telco, and through the fibers, then they derive dark fibers from the railway system and run their own system of things. 

And I think if I'm not wrong, currently there are three or four organisations that they have permission to have access directly to the Internet.  One of them the Ministry of TeleCo and the other is the TV broadcasting company and the other is IPM which is the parent organisation of the IR and I forgot the fourth one but there are four that connects to the Internet directly.

>> MODERATOR: Thank you.  So, the next person, Michael Kende from the Internet Society will talk about Africa.

>> M. KENDE: Thank you for being here.  So I want to talk about the experiences and the work we have been doing in Africa.  I started working on some of these issues before joining the Internet Society and have continued and we have been looking at barriers to connectivity within the region and setting up IXPs to make traffic exchange more efficient. 

But what I wanted to talk here is just go a little through the value chain of how traffic gets into Africa starting with submarine cables.  So on the submarine cable side, I think the story is pretty encouraging.  A lot of this information is from Steve song has a website called, "Many possibilities" that has great data on the submarine cables.  And he put together this history you can click through whether the submarine cables came. 

So it starts with 1999 there is nothing, no cables, all satellite.  Then, 2000, the first cable but just the north.  And will 2001, sat 3 on the west coast, 340 gigabits along the west coast and then nothing.  No new investment until 2009.  Now there has been an explosion of new cables and capacity.  So in almost every coastal country, they have a landing point or at least one, possibly two or three. 

So an amazing story and even better than that, a lot of these cables have open access conditions so that they are not given to the current incumbent who can effectively not pass on the efficiencies.  But I think I was working in the Gambia and five ISPs have licenses now and own capacity on the cable and anyone else who comes in is guaranteed to be able to buy capacity.  So an amazing turn around in 5 years. 

One thing that is not happening, and this is a theme, there is not much measurement of the results, at least formalized.  So I had a great story that you always want to hear.  When I was in the Gambia, they turned on the ace cable in 2012.  They turned on the ace cable for two days and turned it off and I just met with someone and said, you work on the Internet.  Maybe you can explain why I always Skype my family in Germany and usually it is hard to get a voice connection but for two days I could use video.  And then it went away.  And hopefully now it is back. 

But I thought, that is a great story.  They invested 700 million dollars and there is a great return in the speeds but since then, looked around and no one has formerly given before and after what happened to the speeds, the people are getting.  What happened to the prices.  How is this amazing investment paying off?  So I think that is a theme with all of these things that it is great to put it in but also good to measure to make sure that the results are showing.

The next bit is terrestrial.  And again, Steve song on his website has this after‑fiber project that shows, and you can see all the capacity moving in from the coast, some of it directly connected to the sea com or cable companies but moving in to the coast, liquid TeleCom just completed a fiber ring in east Africa so it is redundant and resilient. 

There is still some challenges.  When we did some work on this.  We found that there is still a lot of challenges with rights of way access.  You have to go city by city and get permission and that could add a lot of time and some cities view that as a great opportunity to tax so they'll basically double the price of deployment just by asking for licensing fees.

There is not much sharing.  If the electric wires or the roads or other rights of way don't get shared as much as they might be, and the problem with this is that even if you have the submarine cables or especially because the submarine cables, the cost of international is going down and people may still use it for tromboning if the cost of getting to the IXS still too high because of the domestic road blocks.  So, that really needs to be harmonized in a way of sharing the rights of way by the government, that would be helpful. 

One significant issue that we found when we did this work was cross‑border.  There is 16 landlocked countries in Africa that of course will never have a landing station.  Others it would be good for them to have access to multiple landing stations for competition.  And it is still hard with we did the study, we found a lot of road blocks getting across borders.  You have to coordinate up to one side and make sure you coordinated up to the other.  And that is a significant challenge. 

And there is still not enough capacity necessarily across the continent.  And someone did do a study, these ripe atlas probes that people are putting in and someone just programmed them all to ping other probes across Africa and showed that there is still some improvements.  There is still double satellite hops to gets from one side of Africa to the other with almost a second of latency when you go up and down twice.  So the measurements are very important from all of this.  There is still some room for more fiber across the countries. 

On the metro side, we haven't really studied this specifically.  I know Google has this project link in Uganda where they just chose this country to lay down some fiber, get a license, lay down some fiber and become a wholesaler to all of the mobile ISPs and the other companies.  And the point of this, they offer unlimited link.  You pay a monthly fee and you get an unlimited capacity or unlimited within the fiber that has been deployed.  And the technology.  I haven't seen any results of that but that seems to be a promising approach that a private sector company can do that.

On the fiber to the home side, I hope I'm wrong, the only place I seen a little bit is in South Africa just for obvious reasons it's not going to be a priority.

So what I would say is that there are investments taking place.  Governments should think about the road blocks that they are imposing within the country and at the borders to bring down the cost and the road blocks to get more capacity.  And then people should be thinking about measuring privately and whoever is doing these investments so that when there is new policies, when there is new investment, we can also all seat results of that and identify places like these double satellite hops where investments could still take place.

>> MODERATOR: Thank you.  So let's see that I understand you correctly.  The sea cables you see that the fiber is open and relatively easy to get access to.  Was that was also true for the fibers on land?  Because I think you said that the actual right‑of‑way was difficult and exclusive.  Is that also for the fiber itself?

>> M. KENDE: So submarine cables, through the investment consortia, they are not just incumbents anymore.  Although, if the prices are coming down, at least I haven't seen any studies that show it's been effective in lowering the cost of transit.

>> MODERATOR: But in land, is there a difference between exclusivity for the right‑of‑way and the fiber itself?

>> M. KENDE: There is not much sharing so some companies are building for their own use and not wholesaling it.  Liquid is a wholesaler so they are making it available to others.  So it's country by country whether or not it's exclusive or open for everyone.

>> MODERATOR: In general, the right‑of‑way is the problem.  Karin, I saw you wanted to comment but I wanted to give the microphone to Tracy first so we go over all speakers first and then let you in to speak.  So Tracy, please.

>> TRACY Hackshaw:  I'm from Tobago from the Caribbean region and I am asked to speak on a specific region as well.  So for all the islands that have challenges with Internet access.  In the Caribbean, we are quite lucky, actually.  As you may know, we are centered between North America and South America.  So, by luck of geographic location the Caribbean is very well connected because many of the international operators have to land in one of the islands to connect to other jurisdictions. 

Some people say the Caribbean may be one of the most connected places on earth.  And Puerto Rico, for example, they have nine land cables and Trinidad, there are five and British Virgin Islands, there are seven cables.  The challenge in the Islands, including the Pacific, is not the fact that connectivity is available via fiber.  It's how do you get from the ‑‑ I think it was raised with the African situation just now, but the challenge in the Caribbean is moving from the connectivity at the landing stations into the homes, into the offices, into the environments.  And Governments and regulators have significant challenges in asking operators to move beyond what is considered to be commercially viable locations.

You find in all the islands the cities and towns tend to be very well connected in some cases, with reasonable access.  But as you move out of the cities and towns, the access diminishes quite significantly and even wireless access attempts generally fail because many of the islands are volcanic and very mountain us and very challenging topographies. 

So while you may look at a map and see a lot of land nodes and a lot of connections, the actual access bricking that into the islands into the Pacific, very, very challenging.  So the discussion that I would like to contribute to is, how can we, as the Internet Governance Forum put small developing states in the Caribbean and Pacific and African and Asian region, to bring connectivity beyond the downs and villages that have commercially viable propositions?

In a large case of the Caribbean, most of the landed cables are owned by foreign interests because they are on the way from North America to South America and beyond.  And in some cases to Europe.  So there is no vested interest of those operators in those fiber owners in doing anything beyond landing the cable.  And it is left to the private sector within the small islands to do that.  Without government intervention, without regulation, that is going to support moving connectivity beyond the urban centers, how do we then provide that access to the next half?  So in many countries, you have 25‑30% connectivity through fiber and in some cases through wireless.  The question, how do you get other 70% connected?  And that question has remained with us quite a few years.

And universal service funds notwithstanding, projects of Governments done with access centers notwithstanding, it still remains a challenge to bring fiber to the home, fiber to any particular location beyond the urban centers.  So that is how I wanted to frame my discussion today.  Thank you.

>> MODERATOR: Thank you very much.  So Karin, please.

>> K. AHL:  I just wanted to get back to the points of rights-of-way.  If we look at different regions today, the two points that are standing out mostly that hampers access to fiber is right‑of‑ways and GIS data accessibility.  And that is something that we need to focus on, not only for Africa but for all regions and Europe and the U.S

I mean this is an instrument for all land owners to really hamper and slow down the deployment, definitely here in this region.  So, that is the point for us to focus on.  We have done so for many years and my assistant in Africa is focusing a lot on this topic.

>> MODERATOR: Michael?

>> M. KENDE: It is an issue everywhere.  I think 10 years ago I did a study for tech group in the U.S. called Tech Net on the rights-of-way.  And comparing or ranking all 50 states on the rights-of-way policies and there were huge differences.  And the model was a state that took control of their municipalities and said, there is one forum you can use and one maximum price you can charge and one deadline in which to answer.  So that people could go online, know this they are going to get an answer, build into their business plan how much it was going to cost per municipality.  And then I think we really have to get in, especially for the metro and long distance fiber into the sharing issue that you shouldn't be building a road without putting a duct in at least, and everyone should know to some degree when others are building in common areas so they can share those costs.

>> MODERATOR: Thank you.  Qusai?

>> Q. SHATTI: First let me reiterate the importance of the land line cables also.  They are equally important and sometimes most strategically critical than marine cables, especially for landlocked countries.  Second, the business of cables, fiber optic cables, should be privately led. 

Our experience is that when the private sector do these cables, they would better more efficient and with more capacity, while in our region, while the cables from the private sector was for the capacity of 4.25 tera and 5 terabytes, the cables that were done by governments like the 4 cable was only 5 gig, and our cable that links us to Iran, for example, because it is done between two ministries of communication was only for 500 gig.

So this should be also a policy for the privately‑led initiative and we should adopt policy that is allows that.

>> MODERATOR: So what I hear you talking about, has to do with the right‑of‑way for the fiber itself, sharing the ducts when the right‑of‑way actually exists, sharing the fiber when that is actually put into place, and having these private sector‑led.

That's what I heard so far.  But what about sort of the connectivity that we heard Tracy talk about from the sea cables or land cables to the end users?  There is a gap there as well.  Are specific things have to be done there as well?  Is it the same kind of principles for that gap as well?  Or is it more unknown.

>> K. AHL: I don't know if I have a straight answer to it but we have done studies on it and what we see is a driver, at least, when you have a country connected and then want to get out of the end consumer and customer in the end.  We see a big progress in countries where you have multiple operators acting at the same time. 

And then you could have a privately‑owned or multiple privately owned network owners and government investments competing about the end consumer and then we see it happening and then we see a new access method and possibilities to access and consumer arising, which is also very competitive and interesting to that consumer

That is what we have seen in some markets.  In other markets, there are other incentives for getting connected.  So it's not like a clear answer, but we can see a huge step forward in those countries where you have competing operators who wants to get out to the end consumer and what is interesting is, for example, in Europe, you can see the TeleCom in Germany acting in one way but when they are acting outside Germany they have a different agenda doing it differently.  Which is really interesting.  Then it really moves fast.

>> MODERATOR: So it moves slower on the home market, right?

>> K. AHL: More or less, yes.

>> MODERATOR: Quasi?

>> Q. SHATTI: Thank you for the good question.  Let's take the example of Kuwait.  We have an ambitious project five years ago initiated to have fiber to home connectivity.  But the minister went to bureaucratic red tape, let's say obstacles, and some infrastructure related issues where they decided to say, for the new urban areas we'll have fiber to home, and then we will think about connecting the, removing the old cable network and install a new fiber.  Well, what happened is that the new urban areas, from the ISP business, they can offer high band width capacities because they have fibers.  They are connected ‑‑ and switches.

For areas where I live, where we have a cable network where the best performance cannot give me more than two meg, these areas became the market for the wireless mobile operators or the ISPs who operate on the last mile, let's say the last mile, they have a wireless technology.  So, in a way, that created a kind of menace.

>> MODERATOR: So it is imbalanced but at the same time there is competition between the verses access mechanisms?

>> Q. SHATTI: We have three mobile operators and they all offer Internet connectivity.  So there is competition.  But this competition, because of infra product structure reasons, there is a kind of imbalance.

>> MODERATOR: So I understand, the competition is on national level but in reality, depending on the infrastructure, you always have one of those being dominant if there is fiber, have you the fiber operators.  If there is not fiber, you use other mechanisms.

>> Q. SHATTI: The regular ISP can connect.  So I go wireless and I have higher.

>> MODERATOR: Alireza, yes.

>> A. SALEH: Regarding fiber access in Iran.  We have, as I said, we have operator that signed a contract and ran it as a pilot project for 8 million Iranians in two major cities.  But there are two things.  One of them is that because the eCommerce in Iran most is based on the old methods and eCommerce is not ‑‑ it is developing very fast but actually not the majority of the business in eCommerce so that is why I mean it is going very slowly and also, so the size of business is not that big and because the infrastructure is owned by the government, this kind of project should be owned by the private sectors.  But technically, because it belongs to the government it is going too slowly.

The other thing I think, the fiber connections to the homes are not very fast in Iran because most of the users are using mobiles to access the Internet or go shopping.  Because there are not enough higher speed required contents in the country.  So I think that it is more motivation to the mobile operators and wireless that give the connectives to the wireless devices as my friend said.

>> MODERATOR: Thank you.  Tracy?

>> T. HACKSHAW:  So one of the challenges we have again is similar to Kuwait, so bringing fiber to the islands and getting wireless to come up you still need to bring something to the points that will provide the access in the wireless areas.  Again, because of topography and commercial nonviability, you tend to find even the wireless, even if there is connectivity, the private operators are not encouraged or motivated to bring up even wireless connectivity in the Caribbean.  You tend to find duopoly operations where they are both entirely fiber and they run mobile networks as well. 

But even when that happens, you still do not find the underserved areas that have been identified which are not commercially viable being addressed in any great sense.  The solution has been, in many of the islands, to have government intervention.  So Governments have to find a way to incentivize those private operators. 

There must be some level of public‑private partnership where either they are using service funds or they demand or mandate that you connect these communities using some project and trying to use the access centers.   When, in order to connect a community of that certain speed, you have to run fiber.  You have to.  There is no two ways about it and wireless will not cut it. 

So, in the island we believe there is a government rules at play or regulatory rules at play to demand, if you want to use that word or insist or more or less persuade operators to move into the non‑commercially viable areas.  And in some cases when they do move, the operators themselves are surprised that they are not commercially viable before according to their calculations, but as soon as they go into the area, we tend to find it is one operator moving into the area and then they have monopolies in those areas because once one moves in by themselves, they are able to connect up an entire 1000 households and made that connection very viable.  There is creative ways to look at these issues in small states and small islands to have that situation looked at through public/private partnerships.

>> MODERATOR: So Michael, you mentioned for this fiber that was up for two days, a person could sort of use Skype with video as a positive thing.  It sounds a lot ‑‑ when I talk to people about fiber just like here, everyone locally wants to use the fiber to connect to something somewhere else.  And Qusai even mentioned that when the fiber was cut, no one could use the Internet.  Of course they could use it locally. 

So I have a question about providers of services.  I presume that they are hosted a broad even though the provider of the service is local.  Is that changing when a fiber is connected?  So when you get highest bandwidth dot services move locally or the other way around?  Or there is no relationship?  Or don't you know?  You're shaking your head, Michael.

>> M. KENDE: We are doing a study on this in Rwanda.  There seems to be no relationship.  They have good fiber connectivity through Tanzania, through Uganda or Kenya, but there doesn't seem to be any relationship because for the content providers it is always cheaper to host a broad and they'll host a broad until they start realizing the implications that they are causing on the ISPs and when price comes down on the fiber, it costs less to bring it back.  But it still imposes latency and a significant cost on ISPs.

>> MODERATOR: Karen?

>> K. AHL: I agree.  What we see is they want to centralize it because you have also the possibility to maintain it in a different way.  If you put out locally, you need persons to take care of it and added costs surrounding this and they want to centere this to be able to maintain it more professionally.

>> MODERATOR: Also the local businesses choose to have it a broad.  Yes, because they are attracting customers all over the world of course.

>> K. AHL: That's the new business plan of today.

>> MODERATOR: Quasi?

>> Q. SHATTI:  Of course we always want to localize the traffic.  As much as we reduce the upper stream traffic, we are also reducing the cost of access because we are saving on bandwidth.  But it is a challenge for us to do that. 

Not only that, ISPs are still for us, a challenging issue.  We may not necessarily ‑‑ we have ISPs but they are not properly set up or it is based like an initiative between a group of ISPs but not yet a well-established organisation, let's say.

And in the region for example, in the Middle East, we don't have an ISP tier 2 network.  We are all tier 3.  So these are all challenges that addresses the question.  We hope one day.

>> MODERATOR: So you say that even that local traffic exchange between ISPs is difficult.  That to some degree what you talked about as well, Michael, that even though you have good traffic from an ISP in the country in Africa to whatever globally but to neighboring ISP, specifically like in neighboring country might be as bad as before even though we might have fiber in.

>> M. KENDE: I think it's getting better, but it relates a lot to this point.  When people start to talk about local content, or content hosted locally, it is usually interpreted very literally that every country has to host everything.  And I think that if you think about regionalization in trying to build up some scale and having a regional hub the way Amsterdam or London is in Europe, then you can start to build up a scale and get the price down for the hosting.  But then you have to have really harmonized, just like in the EU or other regions, harmonization, so you can cross the borders cheaply and buildup some scale.

>> MODERATOR: Tracy, that problem with non‑localized content and services, that if I understand correctly, that has been the history in South America and Latin America Region all the time, where all transmission is sort of from Miami which means that Miami has more or less been the best location, best within quotes.  But I understand that is getting better with connectivity and also cross connect activities between Latin America region?

>> T. HACKSHAW:  So in Latin America and Caribbean region we have an ISP project that has been on the way for the last about five years.  So we do have tower‑shaped or whatever the direction is.  Sort of a download culture environment in the region.  All content is being drawn from the U.S.‑type servers. 

With the ISP proliferation, we have begun see some modification.  However, until the local content begins to emerge in many of the islands in this case, in my case, I can't speak for Latin America fully.  We are still going to have that challenge with less caching service abroad into the territories.  You have to still see heavy traffic out of Miami, in our case, and I guess in Australia and Asia, in the case of Asia‑Pacific.

Another challenge we have, obviously, is within its ISP, whether the operators themselves are going to collaborate entirely.  In some cases, two operators are connected and that is not helpful in those situations.  So again, there are challenges and they have to deal with control of those situations until they are rectified.

>> MODERATOR: Local content.  I was waiting for someone to mention that.  Specifically when we are here on initiative by people from the far speaking region of the world where the local content is interesting, specifically because you also use different scripts across the region.  We have two questions from the audience.  Let's start in the back.  Your microphone is not on.

Can someone help the gentleman with the microphone?

>> AUDIENCE MEMBER: My name is -- from Nigeria and I'm attending this event as a ambassador.  I just want to ask a question.  What is your experience in relation to how your ISPs and your local content gives the international content ‑‑ for instance, Google, cache servers around most countries and against pulling the content directly from the U.S.

For example, in my country, Nigeria, we pay by data size.    We pay for 5 gig you pay a certain amount.  And you get charged for the same amount of data you pull locally compared to what you pull internationally.  So I'm think figure you have local content, they should not actually be removing the data from what you bought already.  So do you have an experience in your country?  Is your ISPs giving preference to accessing local content free without charging for the Internet?

>> Michael?  You want to tackle that first.

>> M. KENDE: So we did a study of the impact of the Google caches and we looked at it in Nigeria, it wasn't just on Google but we were looking at the impact of ISPs in Kenya and Nigeria and did see the Google cache had this impact.  Significant amount of new traffic because it was local and it was quicker to download resulting in a lot of new revenue for the ISPs and lower costs that weren't directly passed on because there was one data charge.  So just on that question of the data charge, I think it has been tried in South Africa and Chile that if you access local content t doesn't go against your data plan but that was on DSL plans and I think they have always the challenge of how do you think know did is local when you're downloading and how do you know it is going to be free?  That has been tried in two countries.  I haven't seen any studies to see the impact.

>> MODERATOR: Thank you.  Qusai?

>> Q. SHATTI: Well, let's take for example how the ISPs in Kuwait they created their own IXP.  Because they are lucky enough to be in the same building.  So they created a mesh between themselves.  So each IXP has a cable from a switch to the other and they localize the content inside Kuwait.  So when I used to access my online banking or access a local newspaper where the traffic goes to the States and come back, now through this mesh, which is not a proper setup but it is working, right?  It is now totally local.  And when cable was corrupt, this was operational.  So it's not a proper thing.

But today the local content we are talking about and I will make a point to my colleague, is not the newspaper on my online backing or day trading, it is the Google.  It's the social networks.

That's the global.  I want to have them added based in my region where the traffic can be localized for them because this is my highest.  I want WhatsApp which is peer-to-peer to have a regional hub .  That's a localization of content.  This is the mainstream upstream traffic

>> MODERATOR: So you want Google and others to move into the same building where you're bank and newspaper is, right?

>> Of course.

[Laughter]

>> MODERATOR: See we have a question in the front and then another in the middle.

>> AUDIENCE MEMBER: I'm here from Sweden.  I'm from the regulator of Sweden.  And thank you for the discussion of this very hot topic for this section. 

I guess completely I agree with my brother as marine cable goes to another that hub.  And then that of fixes the cost of the service definitely in our countries.  For example, in Sudan, we have many of the fibers of marine cables, flag, easy sat 1 and 2 connecting us to Saudi Arabia.  The price is very high.  For the universities it costs 100 U.S. dollar per megabit per second.  For organisation, it is 500 for one megabit per second.  For years the Internet for user is much better in term of prices.  It costs maybe $1 per week because of the sharing is very high and now we have 10 million subscribers out of 30 million.  10 million from 30 million.  But still it is questionable. 

So my question is, we need to ‑‑ we are different from the case of Iran where all the submarine cables and landing stations are owned by private companies, which is operators.  So as a country, we cannot manage this very well.  When the government owns the gateways and the infrastructure. 

Accordingly, I think we need to work around to find a way to lower the cost for the customers.  And this needs to be done globally because the cost is now affected the quality of the service that we have and affected our broadband plan.  So what do you think about this?

>> MODERATOR: So I am trying to understand your question.  You said the price for the end users is still high but you see the wholesale prices are low for whatever services the ISPs need.  So, why is the cost high in your country you think?  And I'm asking you because you said you were from the regulator.  You must have investigated this.

>> AUDIENCE MEMBER: Because trying to get connection from our country, for example, to Saudi Arabia and then from the Internet to Europe, definitely and the cost of that submarine cable and point to point link is very high for them.  So costs have become very high.

>> MODERATOR: But that means the wholesale price of that specific piece of the product they are building is expensive?

>> AUDIENCE MEMBER: Exactly.  And whenever cannot manage this as regulator, cannot manage this to lower the price because it is completely private.

>> MODERATOR: Any comment?  Anyone that knows about the situation in Sudan?  Qusai?

>> Q. SHATTI: That is the issue of a privately‑led or privately‑invested marine cables.  When the private sector -- or when the private sector invested in marine cables like Tarter, Lucent and Altell in our region.  They Marketed their marine cables well so they brought more international carriers to our region where now I can enjoy a better competition and better pricing schemes.  Competitive pricing scheme for my data.  And data services. 

As a result, three years ago, the price of the international service has been reduced by 20%, two months ago, the ministry reduced again the cost of international circuits by 20%.  But this is depending again on the skill of negotiation with the international carriers.  And the volume of the traffic and what goes on.

But, for example, when the falcon or the GCC bridge falcon cable, which is now 5 terabytes, the price of the international circuit on this cable is now by far less than the floor cable which is 5 gigabyte bandwidth and nobody wants to use it.  So which is better?  The one done by the government over the one by the private sector?

>> MODERATOR: What you're talking about, yes, you're comparing private and public sector.  But what I think you really talk about is the need for competition on the wholesale side, right?

>> Q. SHATTI: On the private sector because they want to have investment, they are keen to bring more international carriers to that region and that creates competence schemes.

>> MODERATOR: In your case, it seems to be the case that what I hear you saying is that for that specific high price wholesale sort of link, there is no competition?  You're a localized piece and cannot buy transmission in any other path.  Correct?  Thank you.  Next question in the middle.

>> AUDIENCE MEMBER: Good morning.  My name is -- TeleCom operation in Brazil. (Indiscernible)

>> MODERATOR: Can you repeat that?  I didn't understand.

>> PRESENTER: The question is, why operators of submarine fiber in these regions, especially from Brazilian perspective, do not optimize the traffic points of exchange data?  That is the main question.  Because he is an operator in Brazil and he is thinking about why you're talking a lot about fiber?  And why you do not talk about exchange points of traffic?  That's the question.

>> MODERATOR: We have a person in the audience that wanted to comment on that.

>> AUDIENCE MEMBER: I am from Argentina and I run the IXPs there.  And I understand the question, trying to say that South America is similar to your region.  We have small cables or small quantity of cables.  But the problem is in the whole market, as Patrick said, is quantity.  And they need to get the investment on the cable.  And their position of the man from Brazil is that the amount of traffic that you can exchange locally as an example of the building, you said, but the problem is how you deal with the caches and other content to be hosted locally? 

Because in our region, for example, local content is no more than the 10% that our users are using.  The Google cache, for example in our country, it is more or less 5% of the total content that we have.

So our issue is to try to get more content locally based on caches or other CDNs, trying to get our Internet exchange points, most of the traffic that we can.  Instead, going abroad to try to find the traffic.  I guess this is your point.

>> MODERATOR: You can hand over the microphone to the gentleman in front of you.

>> I'm Christian Kauffman from Akamai so I am one of these content people with the caches.

>> MODERATOR: Yes.

>> AUDIENCE MEMBER: In general, and I think this is true for Google as well, in general as we are in the business of bringing content closer to the end user, which is a global statement.  That is not just for the western hemisphere.  We try to bring them in every country.  And I think especially like in Africa, Latin America, we made progress in the last years.

For the countries which do not have caches today, the reasons are either from a legal perspective because we have got a check.  From a legal perspective if we are liable for the content which we transport, which in our case is not true because we are not owning it.  Or if there are any other issues we have to take care of.  So it takes some time.

The other point is, depending on where we connect, if you put caches in a network, which is easier, or if you go to an IX and now comes the funny point, if you go to an IX, we have to fill the caches as well.  The content which is somewhere else in the world has to come to the cache. 

So at the moment where we start going to an IX and we buy transit, whatever price you have, like this 100, 200 dollar, 1000 dollar, now we have to pay as well.  Which destroys every business case in going to an IX.  So we talked to a lot of IXs in the world, and also in emerging markets, and tried to find ways around that.

I guess in some cases, we were quite successful, South Africa is one of the points and others.  But it takes some time to talk about it and find a way.  It is not a straightforward approach like somewhere in the U.S. or in Europe where you just apply it and it is more like a commodity thing.  You have to find a way around the pricing and the issues and the particular countries.

>> MODERATOR: So it sound a little bit like a catch‑22 problem.  The prices are high as long as there is not more traffic.  But on the other hand, the prices don't go down unless you have more local traffic, which you don't get unless you have the caches there that don't want to pay the high prices.  So, nothing happens.

>> AUDIENCE MEMBER: It is a catch‑22 or 21.  That is true.

>> MODERATOR: Port 21.

>> AUDIENCE MEMBER: But there are ways around like when you start to deploy caches in the networks itself because then it helps.  It gets faster for the particular network but it is also getting cheaper for the country

>> MODERATOR: And also for the peers of that specific network in the cases that is allowed and happens?

>> AUDIENCE MEMBER: This is now one of the points, just very quickly.  We figured out we want to put caches in a network.  Every network is more or less happy about it because it helps them a lot.  In the moment where we ask if they want to share the benefit with their competitors to help the country and the renal on, they are not so motivated anymore.  So, some do, most of them not.

>> Anyone in the panel that would like to say something on this topic?  Michael?

>> M. KENDE: I heard somebody say that 10% of the traffic is local and 35% is the Google cache.  But in some sense, wouldn't you consider that to be local?  I mean, from the point of view of the ISP, from the point of view of saving the cost and it's locally hosted or locally relevant but not created locally. 

Because, I mean, and then I think the argument has to be for hosting these caches and I know Akamai is now really starting to move around these regions as well that the savings are enormous.  Whatever the cost of the transit, it has to be 3‑1.  You bring it in once and you save quite a bit of money.  Now bring it in over and over.

But again the competitive issues get in the way.

>> MODERATOR: But there is also who pays for those bits, right?  Right.  Please.

>> AUDIENCE MEMBER: First I would like to thank our panel efforts for interesting presentations.  I have two closely interrelated questions.  I wonder actually you have already reflected upon those but still I wonder how active your regulators have been in ensuring competitive environments for private interests as our panelist from Kuwait mentioned.  To expand infrastructure in terms of ensuring effective rights-of-way conditions, infrastructure sharing, joint construction, the situation with Iran is more or less understandable where the state still controls the bulk or the whole of the infrastructure. 

But in other countries like African region, the Caribbean, Kuwait, and as our panelist from Caribbean region mentioned about universal service fund, I wonder whether the establishment of such a fund has been successful experience for boosting the deployment of broadband infrastructure?

>> Can you present yourself please?

>> I'm from Armenia and I represent Civil Society organisation and the same time, I'm a ICT expert working on internationally funded projects.

>> MODERATOR: Thank you.  Anyone that would like to respond on that?  Tracy.

>> T. HACKSHAW:  I can respond to the last point.  Again, the results are varied.  In the islands you find that there are creative ways of trying to deal with it.  So most universal service fund projects tend to be bringing access to underserved areas through schools, community centers, and access centers.  However, there hasn't been a direct focus on bringing broadband to those areas.  So it is access.  Whatever kind of access is it, wireless and so on. 

I think there is a creative approach that some countries have taken.  My country has done that.  And that you have to insist that the access in those areas, the schools, the centers and so on must be broadband and must be fiber in some cases.  And that way as you seen in my country where the network expands by mandate, so the broadband network only reached X and Y can now go to Z we haven't seen a lot of that in the island and I think that has to change.  Government has a role to play because the limit has been to bring access.  And simply bring access not high speed access but broadband access. 

So that is slowly adjusting, slowly and changing and I think the universal fund can happen as well.  Because you're not going to find most operators bringing new access, new types of access in those areas to bring what they have.  So copper or wireless at its lowest levels.

So you wouldn't see the LTE or the HSPA that is exploding out.  So it has to be a creative way and I think it has to define so when the Governments or regulators say bring access, they need to qualify that by saying what kind of access to bring.  And it is now changing but because you have reached the areas already with some of the funds, it is very difficult to go back to the operators now and say, you have reached them, now you need to double up on that and to bring a different tine of access.  So, it is a double‑edged sword in that regard with some success.

>> MODERATOR: Thank you.  So far back there.

>> AUDIENCE MEMBER: My name is -- I'm from the University of Hannover and I have a background in IT law so I'm a white elephant here and my apologies if my question is trivial. 

But what I would like to know about is how you will see the future relation of competitiveness of satellite Internet versus Hyper‑‑fiber optics.  I see the issue of latency of geostationary satellite but there is still the low‑orbit approach which I think has been tried a couple of years ago.  It didn't really work out I think for commercial reasons. 

But, what I do see if I look at especially the African situation, we now have this fiber optic ring around Africa and address, and still have lots of issues to get into the core land of Africa whereas I would think from my non‑expert view that satellite Internet might be a very good way to directly address end users all over the continent.  Would you agree or is that a bit too naive in my thought?  Thank you.

>> MODERATOR: Karin and Qusai.

>> K. AHL: I wouldn't agree to that.

>> MODERATOR: I'm not surprised.

I'm pretty sure you get the question all the time.

>> K. AHL: We get the question all the time and in most panels I sit in, I always get the question about mobile versus fiber or satellite versus fiber.  I don't think there is a competition between them.  I think we need complementary infrastructure and since fiber will be a basic infrastructure for off loading and in most cases I would hope into businesses and ISPs to get their services out there. 

Of course, both mobile and satellite will be complementary.  We have places up in Sweden where you see a lot of fiber which you can't reach.  Satellite an option, of course.  I think we need to be aware of the need of capacity and redundancy and limitations we see with other infrastructures.

>> MODERATOR: Qusai?

>> Q. SHATTI: I disagree.  When I just mentioned in my first intervention that we used to pay $200 per month for a dial up, that is because our connection at that time was satellite based.  If it wasn't for the marine cables, we wouldn't pay 600‑700 dollars today for a 10MB. 

But, definitely as my colleague said, it can complement and it can be a backup, emergency, a way for the remote areas of geographic call obstacles that may lead the fiber cannot be connected.  But again, even today, we have the private satellite telephone which covers our region.  But it didn't pick up that much.  It is still for remote areas.  So, they were not competing.  Because satellite as of today was loose coastwise but it can be a backup, emergency, useful in case of disasters as a quick set up and so on.

>> MODERATOR: And there are places in northern Sweden and north of Sweden where satellites doesn't work really well and I have been helping one product, a ship going north of Russia, and of course all of the new operating systems from Apple and Microsoft requires connections to licensing servicers to be able to work, which is not so good up there because you don't even have satellite Internet.  So they downgraded some of their computers to Windows XP and other version that doesn't require Internet connection and that is kind of interesting as well.  So no orbital satellites?  Maybe good for people that are as short as me but not for tall ones.

>> I just want to bring another issue regarding prices.  Because the prices are still in Iran for the Internet transition is high because the protocol they are using is old ones because they are coming from communication background that people are working at the infrastructure and they are using the kind of synchronized protocols.  So actually they are not utilizing the fibers the way we are. 

So I mean there is one suggestion that changing the protocol and better utilize the fibers that they are using, so I mean this sudden who of the issues that I think we currently have regarding high process.  So I don't know very much about ‑‑ I'm in a situation in other creases that how they utilize the fibers.  This is one of the main issues.

>> MODERATOR: When you and I talked, I still understood that deployment of fiber between and connecting the actual sort of other upstream and/or the copper cables, DSL, helped the actual access for them even though we used DSL for the access network, right?

>> Yes.  I mean, the connectivity for the end user is all over the copper.  But for example, if you want to buy some bandwidth from the Ministry of TeleCo, it calculates based on the -- and you can see in the news the Ministry of TeleCo says we establish 3 to 10 gig Internet.  That is the problem.  So because the infrastructure that is being used in Iran, I think it is based on the old transition.

>> MODERATOR: Thank you.  We have another question in the back, please.

>> AUDIENCE MEMBER: I'm Mike from the Association of Progressive Communications.  I have been involved in a number of submarine and terrestrial fiber, technical and feasibility studies.  And I just wanted to agree with the suggestion that satellite and other non‑fiber technologies are seen as an interim solution as opposed to a permanent alternative.  The roll out satellite in remote areas before the fiber gets there. 

We have seen this across the continent and Africa and now fiber is reaching to the home and is being implemented in South Africa but Tanzania, Kenya, and a number of countries in north Africa.  So, I think we will see the growths of fiber to the home inevitably at some point.  It's just a question of when.  I think some of the key issues around this also that we are now seeing gigabit Wi‑Fi commercially available this year, and next year they will be 10 gigabit Wi‑Fi

This is another interim solution.  So you don't have to wait for the fiber to get there to get a high bandwidth connection.  You can bridge your last mile and do your back up with significant power and microwave technologies these days.  In terms of licensing and regulatory environment, one of the f factors that is key is dig once policies.  We know that deployment of fiber terrestrially up to 70 or even 90% of the cost is in the civil works.  So if this is done once in the operator for their fiber must lay other ducts so other operators can come in and lay their own ducts.

Similarly, the plans must be published and the example in Nigeria where this a law now, they also required this they run their fiber within one kilometer of a public institution, they must lay dark fiber to that institution.  So that is an interesting example.  The universal service fund there is also planning on subsidizing the extension of the fiber networks into more rural areas. 

By the same token, we also have countries in Africa where they have received funding for a open access submarine fiber landing station where all of the localized ISPs have ownership and then the decision was reversed and then the submarine cable landing station was nationalized and ISPs now have to buy at a set price that the government decides how much revenue it wants to earn. 

Final point, in connecting continent, I think we really have to be aware of those difficult cases in between.  I think what we're going to see is a real divide with the very small island states with low populations out in remote areas which don't have the possibility to get connect to fiber very cost effectively.  And if there are any opportunities where ships are passing nearby or cables can be rerouted to ensure these small island states are connected, I think it is in important point to bear in mind.  Thank you.

>> MODERATOR: Thank you very much.  So, I was thinking now we have another 11 minutes to give the panelists starting from the left to make it easy for me, what would be your recommendation to someone to make it more, to make the Internet access, or make the situation better? 

Let me be unspecific.  You can decide who you're giving the recommendation to, what you mean by being eager and it can be local in your country, in the area or in the world.  What would your recommendation be?  Tracy, let's start with you.

>> T. HACKSHAW:  So that's provocative.  I strongly believe that the government has a significant role to play in that.  Governments do need to work closely with the private sector in doing this.  At this stage of Internet deployments, it can't all be about commercial viability. It has to be about the public good and what is in the public interest. 

So I think a recommendation going forward is for Governments to work closely, not only their own and to work closely together in all countries, especially in small islands to ensure that their regulators and their Internet agencies, and the private sector and the operators and so on, work close together in some sort of task force being set up to establish exactly what the requirements are, where the deficiencies are and what at a can do together, working together in a multistakeholder model, to bring the access to everybody that that the stage of deployments.

I think we are talking about satellites and so on, all these things are drops in the bucket.  Little point solutions to try to solve small problems.  We need to solve big problems with big solutions.  So we need to work together, bring the regulations up to speed.  Don't work in silos and ensure that all stakeholders are reached that way.  Thank you.

>> MODERATOR: Thank you.  Karin?

>> K. AHL: I have to focus my recommendation as to Europe.  We are also here in Istanbul today, I think it is to see on the rankings, it sudden who of the most spurring fiber markets for the moment.  And there are two points I would like to send out to the decision makers in Europe.  One, on the policy and regulatory side regarding operators.  Incumbent operators.  Needs to be hardly pushed to get on the track in Europe.  We have good examples.  We need to use them more.  On the lighter side I want to focus on the end consumer.  We need to realize better what the end consumer needs and wants and would like to pay for it and how they could pay for it.  That is the only way we will see an uptake in these networks and also get the business models rolling better.

>> MODERATOR: Thank you.  Qusai?

>> Q. SHATTI: Let me add to what my colleague said.  There has to be a model where the operators are accountable to all the countries that this cable serves, whether operational or even to the safety of that cable.  Because it is crucial of the safety of this cable is jeopardized, that means we are cut off.  That's it.  So, there got to be a model where all the parties involved, there is accountability model to operate.

>> MODERATOR: Thank you.

>> I think that in the current if I want to focus on in Iran, I think that if we let the private sector do something and start to build something, I think that would be beneficial for the country because they go faster and then the size of this business, which is the Internet business getting higher and the companies getting more and more motivated, I think the best would be to let the private sector in the country come and start business in this area.

>> MODERATOR: Thank you.  Michael, you have the last word.

>> M. KENDE: I just buildup on that a little bit.  I think it is important.  I would aim this at Governments mainly, that fiber is not an end in itself but a means to an end.  And we talk about creating an enabling environment and in the case of fiber, ting is very important how it is deployed to keep the cost down remove the road blocks, to make sure that it is competitive and put in place the way of measuring the results to make sure that there is not just ‑‑ I think someone talked about one road block that we heard about that one little area where there is no competition that raises the price and keeps the benefits from everyone.  I think it is important to watch it throughout and then we get into these content issues, which is a whole separate issue and I think the government would have less role there.  But I think it is really to create a good enabling environment that it is deployed the right way.

>> MODERATOR: Thank you very much.  So I would like to thank all of you in the audience for coming here.  I would like to thank the panelists and I think it is time for lunch.  Thank you.

(Applause)

 

*  *  *  *  *

This is the output of the realtime captioning taken during the IGF Istanbul, Turkey, meetings.  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. 

 

*  *  *  *  *