Policy Paper on Connectivity
1. Executive Summary
3 Technical Background
3.1 Peering and Transit – How thousands of Networks become the Global Internet
4. Special Issues in Connectivity
4.1 Access for Scientists
4.2 Access for Rural Areas
4.3 Access for Citizens via a Civil Society Stakeholder Body
5.The Ecuadorian Political, Economic, and Infrastructural Framework
5.1 Existing Infrastructure and Policy Goals for Unbundling, Structural Separation and Sale of IRUs
5.2 Celec EP (Corporación Eléctrica del Ecuador – Celec EP)
5,5 CEDIA – The Ecuadoran University Network necessary for global connectivity to Collaborative Science
5.6 Formulation of a Vision for “Higher Education”
6. Alternative Models
6.1 Case Study 1: Brazil, Netherlands
6.2 Case Study 2: guifi.net
7. Policies to Assist the National Broadband Plan and Strategies for Expanding Internet Use
7. 1 Policy Goals of the Broadband Plan and the Three Basic Strategies
8. Ecuadoran Policy Recommendations
8.1 A single overriding basic principle
8.2 Policy for Bringing guifinet to Ecuador
10. Why I Withdraw this Paper [Extract Only]
Full Paper with All Notes and Active Links DOC (24 Pages): Cook on Connectivity
Full Text NOT Footnotes NOT Links Below the Fold
The premise of this paper on “Connectivity” is that the single most important component needed for the construction of a social knowledge economy in Ecuador is a national broadband infrastructure. Such infrastructure is needed to underlie and enable the nation’s basic knowledge acquisition.
This paper is not talking about ordinary commercial Internet, telephony, cellular or TV services. Instead it explains how Ecuador can most cost effectively connect to the global fabric of networks based on the TCP/IP protocol and interconnected with each other by means of informal agreements to share Internet standards. Inherent in the country’s national plans is the need to build the capability of connecting on a global basis to the development of scientific knowledge in what René Ramirez calls to the wealthy North. In order to develop a self-sustaining economy based on locally developed knowledge, Ecuador needs to nurture its own generation of scientists and plug them into the on going globally networked development of new knowledge.
To do this it must develop an infrastructure that cost-effectively connects with the global system of Research and Education Networks. It must also develop an infrastructure within the country based primarily on fiber-optic transmission that will connect the brightest minds in the nation at high-bandwidth and minimal cost. By forming the networked groups discussed in the other FLOK policy papers they may build a social knowledge economy that to the maximum possible extent is independent of the intellectual property rights of the so-called developed world.
This paper has requirements involving complexity and therefore length that are somewhat different from the others. It must explain the technology of Internet connectivity on a global level. It must also introduce it readers to the workings of the global system of research and education networks. It must illustrate the role that these networks play in the development of large-scale scientific projects in areas such as bio-diversity, pharmaceuticals and climate change that underpin the nation’s ability to become a source of knowledge rather than of raw materials.
To make policy that will allow Ecuador’s resources to be used with maximum cost-effectiveness, it must make clear the technological difference between voice and data communications. And it must begin to show that, with fiber optics the Networked transmission of all information via voice or data is now doable on a cost basis that is orders of magnitude cheaper than century-old telephony.
The most critical issue is the existence of fiber-optic networks extending from the two largest cities into the hundred to 150 next largest cities and towns. Building such an infrastructure from scratch is very expensive. But Ecuador does not have to do this because it has not one but three different corporations with fiber-optic networks that connect all population centers except for the remotest villages. Two of the three are owned by the government and therefore should be open and amenable to the productive changes in policy needed to make interconnection to their networks much easier and cheaper. However these networks appear not to be operating in a manner that is facilitating their use in the least expensive possible dissemination of knowledge.
In order to help readers understand what is already there and available to be put into the service of policy making in the interest of the nation and its people, this paper devotes some detail to explaining the size, scope and operation of these three companies. The first being CNT the national phone company. The second being the fiber-optic grid that connects the national system of hydropower companies – CELEC Transelectric and the third being privately owned Telconet that from the point of view of Ecuador’s need to build the most modern TCP/IP data network capability as well as connect internationally is best situated.
Compounding the increasing complexity of these interlocking issues is the need for Ecuador to invest in its own research and education network. Such a network called CEDIA does exist and currently functions with no government support. If the government’s plans for Yachay and Minister Ramirez’s other three world-class universities are to succeed, it is absolutely necessary that the cost-effective fiber-optic attachment for these universities and the current members of CEDIA be achieved.
In looking at broadband connectivity the last mile to the homes and workplaces of the citizen and schools that educate the nation’s children is inevitably the most expensive gap that must be filled. And in this case there is very good news because of the example that has been done with guifinet in Catalonia. There is absolutely no reason why Ecuador should not be able to build its own guifinet. And in this case good news is that it can be done very inexpensively by the ordinary citizen producing an infrastructure that is held and operated as a commons something that is completely in line with the social and economic organization proposed by the by the five streams of the Flok project.
To achieve this regulatory changes are mandatory. Businesses with capital to invest must be given the certainty of understanding what their costs will be by being enabled to buy IRUs on dark fiber. Terms and conditions for interconnection of network infrastructure to the POPs of the three existing fiber-optic-based companies must be established. Related policies can and should be designed to keep the nation’s investment in its networks open to new interconnection and growth and investment on a predictable and rational basis.
Finally there should be established a civil body similar to the CGi of Brazil that advocates on behalf of the Internet by means of keeping all stakeholders informed of what is being done in these areas by the other nations of Latin America as well as those elsewhere in the world.
The political program of the Ecuadorian government under the Raphael Correa administration is based on two national plans. The first covers the period from 2009 to 2013, and the second the period from 2013 to 2017. These plans posit the building of a Society of Good Living based on a Social Knowledge Economy.
The goal of a social knowledge economy is explained by Minister Rene Ramirez in his January 2014 essay “Towards Intellectual Independence”. He advocates that the Ecuadoran government must invest in the building of a world class telecommunication network based on fiber optic transmission where possible and wireless where not. The goal must be affordable Internet access for every Ecuadorian citizen. Such is the absolute prerequisite to the acquisition of the knowledge the nation needs to emancipate itself.
The principal objective of Ecuador’s national broadband plan is to “Improve the quality of life of Ecuadorians through use, introduction and appropriation of new information and communication technologies.” Announced in November 2011, Ecuador’s national digital strategy, has the goal of having Internet access in place for half of the country’s households by 2015, up from a current level of 29%. The 2013- 2017 National Plan has the goal of increasing Internet access in schools to 90.0%.
Ecuador has substantial fiber assets belonging to three different companies – two of which, CNT and Celec Transelectric – the government owns. These companies are discussed in some detail below. Government policy could encourage these assets to be made available under conditions that would not financially harm them. That internet is a basic infrastructure, like roads, water and sewer and electricity, must be understood by the citizens and the government of Ecuador.
Peering and Transit – how thousands of networks become the global internet
Possibly the most important overall objective is to show how to lower the cost of Internet access. To do this, we must first understand how Internet bandwidth between nations and within a nation is delivered and priced. Secondly we must understand the importance of government leadership that forbids telecommunication companies from setting prices in such a way that they take undue advantage of their possession of the infrastructure needed to connect their citizens to the network.
Networks interconnect with each other at Internet exchange points. Ecuador has had its own exchange point for more than a decade. An exchange point is housed in a secure building into which fiber optic cables and antennas for wireless transmission are brought.
The use of a few meters of fiber from the equipment rack of one provider to that of a second provider enables traffic from the network of the first provider to flow into and through the network of the second provider and vice versa.
At these exchange points networks privately negotiate conditions of interconnection. If they are of approximately equal size that is, if they send each other in equivalent amounts of traffic, they “peer” or interconnect without charge to each other. If they are of different sizes the smaller networks buy connectivity in the form of a certain quantity of bandwidth from the larger. This purchase of bandwidth is referred to as the purchase of “transit” that is the ability sending data across someone else’s network and have that network deliver it anywhere in the world. Transit costs at major exchange points can be as cheap as a dollar per megabit per second. This is a very important subject with a complex history. It is the playing field on which ISPs have for the past 25 years pushed each other around as they seek to gain economic advantage against a competitor .
These exchange points become exceptionally valuable because they can keep traffic local and get it from one part of the nation to another part without having to send it over a very expensive international link where it has to be sent back again using up very expensive international transit bandwidth in the process.
New networks that want to send traffic to any distance have to operate by the purchase of transit from existing Internet service providers. The cost of that transit can vary enormously and to be anywhere from a dollar per megabit per second to $1000 per megabit per second or even more. The variation in cost depends on what the market will bear. Such cost in turn depends on and access to a nearby Internet exchange point and the amount of competition there .
Special Issues in Connectivity
Access for Scientists
Whether in the North or South, collaborative cooperative scientific research is being done mainly by research and education networks. These networks interconnect with each other by means of optical fiber and communicate with each other by means of separate colors of light or ”light’ paths”. In areas like particle physics, climate change research; life sciences, genomics and molecular design what is called enhanced science is being done increasingly openly. Cedia Ecuador’s research and education network should be able to join what is referred to as the Global Lambda Integrated Facility via interconnection at Ecuador’s nap to the increased capacity of the Pacific Caribbean undersea cable due to be turned on in June of this year.
Membership in the GLIF is open to anyone who chooses to send people to any of its two meetings per year. Participation demands the possession of an optical network and the ability to connect that network to a growing series of Global Lightpath Exchanges. Over the past five or six years what has become operational is the ability to place software on a user’s workstation that will enable the user to dip into the network and directly open a lightpath of a Gigabit or 10 Gb or even more as a direct circuit to the workstation or to the computing grid of another user located anywhere in the world and to do this, of course, in real time. Just in the past year light path linked supercomputers have been applied to climate modeling as a huge “virtual machine.”
Access for Rural Areas
Bringing global Internet connectivity in rural areas will be difficult and expensive. It is not horribly expensive to bring a strand of fiber suspended from polls to a rural village. But getting connectivity the last kilometer to every individual home is another matter. Wireless will not work well. It depends on line of sight connectivity and foliage of any kind, let alone that of Amazon jungles will block the signal. The most cost-effective approach would be to enable an Ecuadorian version of guifi,net to build-out from a fiber connection at a CNT office in every small town. Such a network would be built in small steps from one house to the next by line of sight over a period of many months. Satellite is not an advise option because the signals would be blocked by foliage and the latency involved for the signal to travel from Earth to orbit and back would make the network connection for applications dependent on low latency like VoIP difficult.
Notwithstanding these difficulties Mintel funded a significant satellite project in 2012. As part of an “important government program to eliminate ‘digital illiteracy,’ Hughes Technology announced the selection of Telconet to provide an HX System and broadband satellite terminals for the Ministry of Telecommunication and Information Society (Mintel) rural connectivity project in Ecuador.” White papers from Hughes describe technical ”solutions” to these problems including ground connections to enable switchover in case of failure. As a matter of policy, future investment in terrestrial infrastructure by means of inter-connection with existing fiber and a guifnet build out from those interconnections would better serve citizens. A drawback from Hughes is the uncertainty as to whether NSA has added backdoors to the software. Hughes makes drones for the Pentagon. Brazil for anything satellite would likely be a better choice on behalf of privacy.
Access for Citizens via a Civil Society Stakeholder Body
It is remarkable that there is no multi stakeholder body that advocates for the development of the Internet in Ecuador. The existing AEPROVI (“Ecuadorian Association of value-added providers and Internet”) which runs NAP.EC did not evolve around the interests of internet traffic and the ways in which the nation’ will Is large carriers could work together no make the provisioning of small scale Internet service economically feasible. Instead we only have the phone companies acting as telcos and the cable tv companies acting as cablecos, each in their own corporate interest. No one is at the table on behalf of Internet as a source of knowledge.
This can be changed:
For example, Brazil has the Comitê Gestor da internet (CGi). It is a multi stakeholder body. The Brazilian Internet Steering Committee (CGI.br) was created by Interministerial Ordinance 147 of May 31st, 1995, which was amended by Presidential Decree 4,829 of September 3rd, 2003, with the purpose of coordinating and integrating all Internet service initiatives in Brazil, as well as promoting technical quality, innovation and the dissemination of the services available. It is composed of 21 members. Nine from the goverment, four from the corporate sector, four from civil society, three from science and technology and one “internet expert.” Advice from an Ecuadorian equivalent would be more productive in the long run than having some government ministry request bids for the purchase of Hughes Satellite Systems technology for rural service. In the future such an acquisition should start with a new Ecuadorian Internet steering committees input and recommendation as to the most cost effective way to use technology to solve the problem.
The Ecuadorian Political and Economic Framework
Existing Infrastructure and Policy Goals for Unbundling, Structural Separation and sale of IRUs
Ecuador has three companies — two owned by the government and one privately owned– of national scope that are engaged in the provision of telecommunications services. This section describes the companies and their network assets. As the result of the existence of these companies, there are three large 10,000 kilometer plus fiber networks connecting all of the nation’s towns and cities. As all these networks are connected to the Ecuador NAP in Quito and Guayaquil they are well connected to each other and to international bandwidth. What is left is a last kilometer problem of connection to a home or business. Policy that implements unbundling and preferably structural separation into wholesale and retail services; as well as policy that mandates the sale of IRUs would make a major difference.
Unbundling means that the incumbent network provider is forced to sell wholesale access to its infrastructure at tariffed prices that are published and available to any company that wants to purchase and resell the services. Structural separation takes unbundling a step further and would divide the last mile provider in to a company that is responsible for the maintenance of that infrastructure and acts as a wholesaler to anyone who wants to provide services over it. This is usually regarded as a superior solution because under Unbundling the last mile provider as an incentive to favor its direct customers as opposed to those of its competitor. The use of IRUs will be discussed in detail later but the most important fact is that without the sale of IRUs, there is no way for a what the deliverer of Internet service that does not have access to fiber to compete with a company that does.
Celec EP (Corporación Eléctrica del Ecuador – Celec EP)
Ecuador’s state power generation holding company Celec, which was created in February 2010, bringing together generators Hidropaute, Electroguayas, Termoesmeraldas, Termopichincha and Hidroagoyán along with transmission company Transelectric
This page describes services in effect for 25 years that are necessary essentially to monitor electric transmission. It explains that it has recently seen the need to deploy fiber “as part of its telecommunications network.” It adds that “Taking advantage of the new legal framework and the introduction of competition, CELEC EP – entered the market TRANSELECTRIC Carrier Services of Ecuador.” A map of their fiber network is linked to below .
INTERNET Services: According to its website page: CELEC EP – TRANSELECTRIC offers its customers services “BROADBAND” which derives navigation High speed Internet, allowing access to technologies such as voice, video and data that require the transfer of large amounts of information. Among the many advantages are the ability to access a wide range of resources and products to meet the demand for reasonably access customer information.” According to its tarriffs page E-1 Internet prices are only $113 a month . This is an excellent price. What is not clear is to whom and under what conditions this service is available.
The second company is a private company called Telconet which has been in existence since the 1980s and serves primarily enterprise customers. An enterprise is the term used to identify a large commercial customer. In the case of Telconet we are referring to major Ecuadoran corporations and the offices of all multinational corporations doing business in the country. Since the company is private, information about its activities; its customers; and its prices is scarce.
Telconet is a predominantly Cisco shop which means it is dependent on the proprietary technologies that Minister Ramirez would like to break free from. Here is what it says about itself in this respect. “The Next Generation Network (NGN) of TELCONET is a communications platform that enables you to access services from 1 Mbps to 10 Gbps multiple lambdas. In every city where TELCONET operates, it has Cisco certified platforms.” It goes on to relate in great detail its catalogue of Cisco services.
This presents a policy opportunity. Because there is no internet advocacy group in Ecuador like the Brazilian Committee, any source of independent expertise is very scarce. Like Microsoft, Cisco is very adept at selling its solutions to those with purchasing authority. Just like the old phrase “no one ever got fired for buying IBM,” a recommendation to purchase a Cisco solution has been hard to argue against. However in other environments – the mesh wireless of guifinet – there are more open opportunities for new investment.
Open hardware alternatives do not exist for large scale traffic that must travel between internet autonomous systems. At the large system enterprise level the silicon in these machines in many cases is custom-made. Cisco and Juniper are the most restricted environments. Huawei is more open. Since Telconet is privately owned it would be unlikely to embrace a Chinese solution and once you buy in to a walled garden architecture the possibility of replacing it becomes cost prohibitive.
Telconet exists to serve business clients. It touts what it calls unified communications saying that “Telconet poses to integrate all technologies in a intuitively and way easy to use regardless of the end user device.” It adds that it offers “Support for multiple vendors such as Cisco*, Avaya*, Microsoft*, Polycom*, Denwa* and Asterisk.” Finally it has announced nationwide wifi services by the end of next year.
Telconet has a subsidiary called Andinean Cable. Through this is it is said to have a 26% stake in the upgrade of PCCS the Pacific Caribbean Cable system to be complete in June of this year by Alcatels Submarine Networks The PCCS consists of Cable & Wireless Communications, Setar, Telconet, Telefonica Global Solutions and United Telecommunication Services (UTS) The cable lands in Manta Ecuador. In August 2013 Mintel announced Florida-Ecuador consortium PCCS cable, in the 3rd quarter of 2014 will improve ‘by 160 times’ the internet capacity ‘throughout EcuIador. ’
According to Telconet’s web site “The PCCS will be the main output of Ecuador for Internet content and with it, TELCONET will increase reliability of communications and help reduce the risk of breakdown in communications.” Within the possibilities of service offered by TELCONET’s PCCS are the following: [Sales] under IRU model. And Rental capabilities into various format 10G,,1G, STM16, and STM64.
Note that this “IRU model” appears to be a new development. It appears that due, very likely, to the increased bandwidth capacity of the PCCS upgrade Telconet Management is now willing to sell international IRUs. One would hope that they expand to selling ones on their domestic routes as well.
CNT is Ecuador’s national carrier. It is first and foremost basically a Telephone Company which like most fund companies is much better at giving voice and mobile service than Internet as a tour of its website compared to that of telco that quickly shows. The following two paragraphs are offer a concise summary of its history. It has a nork of more than 10,000 km of fiber something that in theory give said the capacity to seetwll IRUs. however I have been unable to find any reference in the literature indicating a willingness to sell an IRU. Several sources have complained off record that it refuses to do so. The very sophisticated data network enterprise related information on Telconet’s website is missing at CNT whose website positions itself as a telephone company rather than as a data services company like Telconet. When discussing policy issues it should be understood that there is a huge cultural difference between a telephone company like CNT and a company like Telconet that could properly call itself an internet company. Because of the packet switched technology of the internet as opposed to time division multiplexed technology is usually an order of magnitude cheaper in the cost of delivery of information phone company management the world oversees the technology improvement as an opportunity to improve its profit rather than pass savings on to its customers.
CNT was formed on October 30, 2008, as the National Telecommunications Corporation, CNT SA by the merger of Andinatel SA and Pacifictel SA. Then on January 14, 2010, the CNT SA , became a public company, under the name of, CNT EP NATIONAL TELECOM CORPORATION . Finally, on July 30, 2010, the formation was completed with the acquisition of mobile phone company ALEGRO . Since the government owns CNT and Celec Transelectric, one would assume that it could exert influence in their behavior and especially in the prices they charge .
CNT advertises “fast boy” internet for 18 dollars a month – very expensive for a county with an annual per capita income of less than $6000. The page calls the connection broadband and says downloads are unlimited but does not say anything about actual speeds delivered. The best overall source on Telecommunications in Ecuador is an almanac published by Budde.com. This document also contains copious information on CNT.
“The company has a nationwide presence with a fiber optic backbone of 10,000 km as depicted above. The fiber cable is installed underground throughout their ducts in ring configurations to provide network reliability. The backbone is provisioned with DWDM technology, with a total maximum capacity of 192 (10Gbps) lambdas. My efforts to find this information on CNT’s current site were unsuccessful.
Given the challenge of the internet, it is not surprising that CNT ‘s reaction is one of protecting their “turf”. This means that there will be a natural reluctance to unbundle infrastructure and support. Not surprisingly the tendency has been one that tends to maintain the status quo with CNT in complete charge of its network infrastructure.
Investment in services. Complete transparency is needed. Not only should bids be public; but the responses to bids should also be public. When CEDIA went out for backbone service, the nation’s laws required a public bid. However they did not require that the price and terms of the bids be made public. That was a fatal flaw. Level (3), Telefonica and CLEC Trans-electric could have bid and did not. The policy lesson is that public bidding serves no useful purpose if the results of the bids are allowed to be delivered under the terms of non disclosure. Policy needs to mandate transparency through the entire process if money is to be wisely spent.
CEDIA – The Ecuadoran University Network necessary for global connectivity to Collaborative Science
In 2012, with CNT as the only other bidder Telco.net, won a ten year contract to provide service. From Telconet’s web site we read: “The Advanced Academic Network of Ecuador (CEDIA Network ) operates over TELCONET NGN, linking major universities, polytechnic Institutes science and technology organizations in the country with fiber optic platforms whose capabilities are exclusive to this network with 1 Gbps capacity.”
As of 2011 CEDIA has had no direct government funding . In early 2014 the situation is unchanged.] Also as of early 2014 it connects 30 of the nation’s 57 universities. Cedia is also responsible for purchasing access to the commercial internet for all its members, a task that currently requires 8 Gbs per second connectivity. It does not want to function as an ISP because of the administrative burdens that would entail.
“The CEDIA budget for 2014 is about $4,500,000.00. 70% of this budget is used to pay for infrastructure to local operators and RedCLARA. The remaining is used to finance research projects, researchers and technicians training, and enhance our infrastructure (supercomputer, cloud, etc.).”
The same source who requested anonymity also said:
“The main problem in Ecuador has been the difficulty to identify suppliers of dark fiber willing to support the project. CNT, the national telecom company has simply stated that they are not interested, which has forced us to rely on private national telecom companies, mainly Telconet.” he later pointed out that the contact with CNT regarding the Ecuadorian segment was made by RedCLARA directly. “We are aware of this outcome because RedCLARA provided this information [that is it told us that CNT was not interested] after the feasibility study was conducted.”
The source of the above map is found here.
Formulation of a Vision for “Higher Education”
In looking at the Internet as a source of connectivity to the knowledge of the wealthy developed nations of the northern hemisphere, Ecuador appears to face a conundrum. It has only three PhD granting universities — something recognized by the current administration’s commendable policy for subsidizing the education of its current youth up to and including the doctoral level at foreign universities. This is an issue that is also acknowledged by the Correa administration’s plans for Yachay, the University of the Amazon, the Education University and the Arts University mentioned below. This is a cultural situation impacting CEDIA that as an outsider I find it exceptionally difficult to grasp. I have spoken with José Andrade the Ecuadorian citizen who left his country at the age of 18 and enrolled in Stanford University. He went on to earn his PhD in Silicon Valley and is now full professor of Mechanical Engineering at Caltech. It makes absolute sense that he would be advising the Correa administration on educational policy.
From a lengthy telephone interview with him it is very clear that the Administration is highly concerned about the quality of its universities and the dearth of PhD’s they produce. This concern makes perfect sense when thought about in the context of Renée Ramirez’ interest in developing a policy where Ecuador acts in such a way as to build strong bridges connecting Ecuador to the development of scientific knowledge within the countries that he refers to as those of the wealthy North.
However, what I have come up against in the research I’ve done over the past few months has been a lack of knowledge on the part of key decision-makers in Ecuador of the role played over the last 30 or so years by the Research and Education Networks in the rest of the world. Ecuador has a small research and education network (CEDIA), but CEDIA appears not to be well understood by those in the government with decision-making authority. A key question, to which I do not know the answer, is how to inform those in the national regulatory authority as well as those the national phone company CNT that CEDIA is not a wholesale buyers Cooperative for Internet connectivity for Ecuadorian universities. Instead CEDIA is a critical group of people with links to other research and education networks in other South American countries. Virtually all of these other Latin American R&E networks, as far as I can find out, are much better supported within their own nations. CEDIA gets no government funding and indeed a recent regulation makes it very expensive from now on for a University in Ecuador to even consider joining CEDIA as a member.
As will be explained with the case study on the GLIF, there is a global science internet to which CEDIA is Ecuador’s only link. The government should be investing a relatively small amount of money that would enable the new universities that the administration wishes to build, as well as CEDIA’s existing 30 members, to participate in those developments. Only 90 days ago at the University of Leiden, under the sponsorship of the Netherlands Center for Enhanced Science there was the founding meeting the Data FAIRPort, an international group that will establish the protocols by which the vast scientific data sets of the semantic web will, it is to be hoped, be barred from being patented and privatized being kept instead in a global commons to the benefit of all of humanity. One of the organizers is working on behalf of Latin American nations. However, Ecuador appears not to be at this table. If Renee Ramirez’s social knowledge economy is to succeed Ecuador should be represented. It is recommended that CEDIA should be included in the summit in Quito next month so that it will have an opportunity to state its case.
If Yachay and Inkiam the Unversity of the Amazon, and Uniartes: the University of the Arts and UNAE: National Education University, are Minister Ramirez’ major projects for world class education, then these institutions should have world class broadband attachments. The infrastructure from which connections can be built is there but affordable prices are not. Solve the problem for CEDIA without subsuming CEDIA into a “public organization” that gives the government complete control over it, and you will solve it for the four new universities You will do so because you will have created infrastructure to which they all can be cost effectively attached. Create a fiber ring or establish necessary IRUs and the marginal cost of attaching new institutions drastically declines.
Case study 1: Brazil, Netherlands
For connectivity to scientific research in the rich North, Ecuador has the example of the globe’s research and education networks. They are not all of equal quality. The United States had one good one called National LambdaRail but it was put out of business by one called Internet2. A problem both in the United States and Europe has been that some R&E networks placed more emphasis on government support to offset operational expenses for ordinary internet connection than they did to develop new operational capabilities for the network. In Canada while Bill St Arnaud was director of Canarie it was a notable exception., while in the Netherlands SURFnet starting all the way back in 1986 became the best in the world.
In this very small and old European nation about the same physical size and population as Ecuador, pragmatic farseeing public servants put together over a period of more than 20 years a network designed to build a national knowledge infrastructure. The Netherlands found itself with the good fortune not of oil but with natural gas from fields that touched the Netherlands shore in the extreme northeastern part of the country. Up until about 2002 the nation was attempting to use the natural gas income to build physical infrastructure: airports canals and highways. But then a young woman who was a farseeing member of the Dutch legislature suggested that the country pour many millions of euros per year into SURFNet, its world leading optical network— the goal being the building of a knowledge infrastructure. The result changed the way knowledge wss formed and disseminated.
By means of grid computing the GLIF enables scientists in many fields of expertise to collaborate in the real-time manipulation of data sets using grids established for the purpose of such varying goals as research into cancer and climate change. BioSciences, biodiversity and pharmaceuticals are major ranges of exploration of what two or three years ago was known as eScience and now is referred to as enhanced science. The Netherlands has a new Center for Enhanced Science now ifinishing its third year. Its purpose is to provide software engineers knowledgeable in the areas of science being researched who can work directly with the scientists in the creation of algorithms to run directly on supercomputer architectures.
Currently in South America only Brazil is a full participant in the GLIF. Barriers to Ecuadorian participation at both levels (GLIF and e-science) are not insurmountable. eScience is necessary because scientists using high performance computing are now facing tasks demanding parallel access in real time to the worlds largest super computers, to specialized software development that matches their needs to rapidly changing architectural and hardware capabilities and to the opportunities made possible by the huge data sets resulting from the exploration of the semantic web.
Without an infrastructure that provides these, the administration’s plans for Yachay as a world-class center for innovation are likely to remain unrealized because leading scientists will not join a university that lacks the infrastructure necessary to support world class eScience. The best over all view of what e-science is about is contained in the following presentation by Tony Hey VP Corporate research Microsoft: eScience, Semantic Computing and the Cloud Towards a Smart Cyber infrastructure for eScience.
Dr. Hey also recounts the extremely important progress being made towards open access in the publication of science research. In a series of detailed blog posts on France, Brazil and Latin America, China, Australia, and Germany, and in a separate and equally valuable series called A Journey to Open Access. Included are chapters on the situation n the United Kingdom and the United States. The shear volume of literature detailed in these reports emphasizes the importance of Ecuadoran internet connectivity to the globe’s scientific research.
Case Study 2: guifi.net
guifi.net, begun in the farmlands about 80 km north of Barcelona a dozen years ago, has become the largest wireless network in the world. It is user built and most interesting from the point of view of the FLOK project, it is built on the premise that its infrastructure is a commons. It is legally structured so that the radios involved and their antennas belong to the communities in which they are placed by means of something called the Wireless Commons License .
Originally it is built on a mesh network principal where, to connect to the network, all you have to do is connect to the nodes of your nearest neighbor provided you can place an antenna so that your radio can see his radio by line of sight. guifiNet however no longer likes being called a mesh network because it is using citizen laid optical fiber wherever possible. Fiber that is very cheaply imported from China and where small towns are very eager to make right-of-way available to connect to their citizens by fiber to the next town five or 10 km away
In 2004 guifi.net filed paperwork with RIPE the European and Middle Eastern Internet routing Registry. This enabled it to be recognized as a telecommunications carrier and gained it legal rights of coexistence alongside Telefonica the same global capital company that owns one of Ecuador’s commercial mobile systems and has an ownership stake in one of the undersea cables connecting Ecuador. It also gained guifi.net the ability to use the border Gateway protocol through the assignment of an Autonomous Systems Number and the ability two handout Internet protocol IPv4 and v6 blocks needed to route traffic through the network mesh.
What guifi.net has built is a basic wireless TCP/IP infrastructure that is supplied and expanded by each radio added to the overall network and which then can be used to carry higher layer application traffic such as email and web services. It has enabled the growth of between 75 and 100 small companies called guifinet professionals that, for a monthly fee of a few dollars, will provide Internet related services to users who are too non-technical to provide their own. In addition to this situation the libraries of many small towns in Catalonia for a few hundred dollars per year buy commercial Internet connections to which citizens of the town can connect at no additional charge by means of establishing a simple network proxy.
Guifi.net connects through international gateways in Barcelona. It is scrupulously neutral in its approach. It essentially runs infrastructure only and as long as its users are not breaking any laws, does not care what they do with their use of the network. Although the network commons license very carefully elaborates that a user may not use his or her access to the network in such a way as to harm the ability of other users to share in the common infrastructure. Another freenet project, similar to guifinet, is being built in Kansas City Missouri in an urban and very poor area serving people who can in no way begin to afford the $70 a month that Google wants for its gigabit fiber in the same city. *
The hardware on which guifinet runs is mass produced and not terribly expensive. It is getting cheaper and stronger all the time. Unlike the routing software that support the largest wireline networks – Cisco, Juniper, Huawei – the guifinet operating system is opensource and undergoes continual testing and improvement.
Guifinet has worked at unprecedented scale in Spain — Catalonia and Castelon. It is citizen built commons owned and operated, It is totally compatible with the aims of FLOK. As long as CNT and celec transelectric permit it to interconnect with their POPs in Ecuadoran towns, it will work very cost effectively. Connections are line of sight but use roof top antennas and power over ethernet cable to go from the antenna to the computer inside the home, school or business.
Policies to Assist the National Broadband Plan and Strategies for Expanding Internet Use
Policy Goals of the Broadband Plan and the Three Basic Strategies
The 2013 to 2017 National Plan merely specifies a goal of 90% of Ecuadorian schools connected to the Internet by 2017, p. 62. The 2011 National Broadband Plan posits a similar objective to be achieved by means of the three strategies .
STRATEGY ONE — The following strategies come from the above cited document.
The first overall strategy of the national broadband plan namely to ensure competition by “establish[ing] mandatory sharing the physical infrastructure of telecommunications networks.”
This is excellent. Efforts can and should start with mandatory interconnection for all network providers. But it should also be extended to the unbundling of network infrastructure at a wholesale level for use by separate Internet service providers. And most significant of all, it could include what is known as structural separation. Namely, the formation of a company in control of last kilometer whose sole purpose is to sell wholesale access to that infrastructure to companies which wish to compete in the provision of services for the so-called last kilometer. A related way of achieving this is for legislators to require publicly owned companies in this case CNT and Celec Transelectric to sell indefeasible rights of utilization on strands of dark fiber within their networks to those who wish to purchase them.
STRATEGY TWO — The second strategy of the broadband plan to “Promote the granting of authorization convergent certificates and encourage price reductions in Broadband service” will depend on reductions achieved in the cost of delivering service. In this case the regulator is Supertel .
“Convergent certificate” should mean you may provide internet service, where this is defined as any data service, using any legal software that runs on a TCP/IP stack. Also that should mean how you do it or what you do with it is up to you.
However, it seems this also means that under the current way of doing things, the would-be service provider must get permission for fixed internet to homes and business. But that someone who wants to expand the economy by offering a new business service of internet to smart phones must go through another authorization process for providing internet via mobile service. Or someone else will face the need for another authorization for service to cyber cafes, yet another for telecenters. And so on.
Apparently Supertel requires separate authorization for every kind of imaginable internet service. Not surprisingly this creates regulatory uncertainty for anyone who wants to offer service.
Policy recommendation – abolish service authorization and replace it with general authorization to provide any form of internet service
According to Budde.com “in late 2013 the regulator shut down Univisa in Cuenca and Azogues for delivering services in areas where it was not licensed.”
Therefore, even if you could buy an IRU on dark fiber, unless you used the IRU for private purposes, the current process of “Service authorization” creates a situation where the government regulator can look at what you are doing with the network and if you are doing something it does not like, it could tell you to stop or worse shut you down.
POLICY Recommendation: The idea for convergent certificates of authorization is excellent. Instead of certification for each kind of service offered, let’s just have one certification for internet service.
STRATEGY THREE — Encourage the new deployment of broadband infrastructure in underserved areas.
Do this by enabling the selling of IRUs Doing this enables the organization that wishes to provide service to gain control over the physical media necessary for such provision at a known cost for a period of usually between 15 and 20 years. By this action the motivation to very carefully plan how to light the fiber on which one’s service depends is put directly into the hands of the people involved in providing the new service. If one wants according to this third of the three strategies listed “to encourage the new deployment of broadband infrastructure in underserved areas,” there is no better course of action to take.
We must understand that without doing this, the current owners of fiber have no incentive to cut prices or extend access. The telco mindset says to them never give access to fiber which, once you have it, is an almost infinite resource to those outside your company.
The prices charged by those operating as telephone companies are always about an order of magnitude and sometimes more than the cost of delivery. Until this mindset is overturned, the market for Internet services in the areas of: “public education, healthcare and e-government” will remain needlessly small. Insistence on cost effective opening of the existing fiber infrastructure via IRUs should enable new market entrants to provide much better service in the mobile area. One would hope that the regulator would be given the ability to demand hearings on pricing and interconnection policy. Decision makers should ask “What relationship does price have to actual costs?”
The are admirable aims. However, reasonably priced and accessible Internet service in Ecuador are unlikely to be available anytime soon because of the following problems.
Licensing for an Internet service provider is an insecure cumbersome process where virtually every field of activity requires a separate license. Roughly 200 ISP licenses exist in the country with the six largest having 93% of the market. Note further that such licenses do not encompass Internet cafés of which there are approximately 2000.
Regulations requiring CNT to unbundle access to its network exist but are apparently have not been issued. Without very public well-enforced rules for interconnection and tariffed prices it is impossible to see how meaningful competition to CNT will develop.
But we must pause and look at the signal this sends to anyone in Ecuador with the capital to invest in becoming an Internet service provider. It means that such a person has no way of making a reasonable long-term plan for the provision of service over a given geographical area because determining the cost of bandwidth the very “gasoline” or fuel needed for such a business over a long-term period of time is impossible. The cost will be determined by the owner of the fiber which in most cases is CNT a competitor of any new company wishing to offer service. Such a policy seems not to be in the national interest of the Ecuadorian people.
Ecuadoran Policy Recommendations
A single overriding basic principle — Enable the cost effective and therefore sustainable building of the network infrastructure on which the actual ability to create and then maintain Ecuador’s social knowledge society will depend. Do this by driving down the cost of bandwidth both in Ecuador and from Ecuador to the rest of the world.
Also emphasize the need to steer away from proprietary network architectures and hardware design. The goal should be for entrepreneurial Ecuadorians to have access to dark fiber and light that fiber with hardware designed to deliver relatively open architectures. (Truly open architectures are not available.) This will ensure operational flexibility and lower costs. The points to be made here are somewhat complex. CNT operates wireline as an effective monopoly without competition. Therefore it has no incentive to build out an improved network. Having control of its own fiber, it has what is called a facilities-based monopoly. Given the infinite carrying cpacity of fiber, overbuilding, CNT is not a rational option. If, for the good of the nation, CNT were to be forced to sell IRUs, it would be giving the purchaser an opportunity to compete against it. Such competition would not be cheap — certainly not compared to guifinet wireless. If a competitor to CNT were to invest in a greenfield build while there is no such thing as truly open hardware architecture, the use of Huawei would be a more open and cost effective choice than Cisco or Juniper. Furthermore as Peter Lothberg, a well known Swedish network architect showed at RIPE last September, an optical network architecture can be designed using alternatives that are more open and more simple than those of Cisco and Juniper. Lothberg’s client was Deutsche Telkom.
It is difficult to recommend policy decisions tailored to fit very complex and rapidly changing technologies in only a few words. Here are a few general guidelines.
For the good of all of Ecuador ask CNT to reduce prices and by doing so increase the size of the market.
By means of the creation of a civil society based organization for the advancement and coordination of the internet in Ecuador establish a body that can give accurate, up to date and neutral information on the development of policy. Advice for the government should be free of economic motivation. The model for this organization would be the the Comitê Gestor da internet (CGi) . The good of the nation, not the profit of the providing organization should determine policy.
Ask the regulator to the extent possible to recommend criteria for purchase of new operating equipment that will take advantage of any existing open source software and open network design principles that encourage interoperability. Discourage making decisions at lower layers of the stack that may restrict business entry at higher application layers. Bidders for contracts for equipment and services should be informed that bids that best demonstrate adherence to these criteria will be favored. These should be interim policies awaiting the establishment of more specific open critera decided on by the new Ecuadoran CGi,
Do not place barriers that will prevent Ecuadorian universities from purchasing low-cost bandwidth in ways that are done in other Latin American countries and indeed in the rest of the world. Transit bandwidth costs for traffic that stays within Ecuador are about a dollar per megabit per second at NAP.ec. Domestic traffic should be priced to reflect such costs. International traffic will be more expensive, but it can be pulled in and stored – cached is the technical term – locally,
Permit the sale of indefeasible rights of utilization (IRUs) by understanding that this will allow other Ecuadorian businesses to invest in offering Internet bandwidth and drive down prices charged by CNT.
Making a wise investment in Ecuador’s long-term economic future by opening up a wholesale market by means of insisting that CNT unbundle access to its copper as well as to its fiber. And that Celec Transelectric do the same with its services.
Promote conditions that facilitate interconnection of every one with ISP authorization. Publish the conditions required for someone wishing to offer Internet service to do so by interconnecting at a point of presence of CNT, Celec Trans-electric and Telco.net. Do this as well for any Internet service provider who wishes to build out its own physical network starting from the location of NAP.ec in Quito and Guayaquil.
Transparency is paramount. Although bids for public services must be done in public, when they are unsealed, the amounts offered for the sale of such services are subject to nondisclosure. The process of making bids in public is good. It offers transparency. However, making the results of the bidding process subject to nondisclosure negates the original effort at transparency since it creates a situation where the requested provider of service can offer outrageous prices and or restrictions in the absence of public knowledge. Therefore the use of nondisclosure for making the results of bids public should be banned.
Policy for Bringing guifinet to Ecuador
In this case it will be a matter of identifying those with technical knowledge and interest in the building of wireless networks based on a Commons infrastructure. Working with those in Catalonia Italy Greece, Germany and Argentina and the United States holding a series of seminars in a handful of Ecuador’s largest cities it should be possible to spread the understanding needed, as well as the personal connections, to began serious build-out’s before the end of 2015.
Once it catches on, the process becomes self-sustaining. What is needed is it legal support of the idea of a Commons infrastructure of network transmission and the ability to ensure reasonable terms of interconnection with commercial networks. In other words the commercial companies must be required to permit colocation and interconnection of wireless routers and switches at their remote points of presence. (PoPs).
From a technology point of view it should enable the provisioning of 5 or 10 or 20 Mb per second of bandwidth to end-users where the exact amount needed would not need to depend on a top-down uniformity of some architect’s prior design. End users could size the radios they need and add more bandwidth by adding more radios.
This would be a very cost effective way of investing in a country where teledensity although it has dramatically improved under the Correa administration is still lower than the Latin American average of 18%.
Understand what guifinet is and how it could be built in Ecuador. What has been built is a TCP/IP based fabric that can carry any TCP/IP using applications that its users wish to employ. It is this fabric that is held as a commons. It gains its legal standing and governance through a five person foundation. The foundation is very carefully operated in such a way that it cannot be captured by the action of any user or user group .
Brazilian Internet Steering Committee
www.cgi.br/ nnd http://www.cgi.br/pagina/about-the-cgi-br/148
Budde Paul (2013) Ecuador – Telecoms, IP Networks, Digital Media and Forecasts
CEDIA – Ecuador’s Research and Education Network
CEDIA Network Map is visible at http://www.telconet.ec/en/services/internet2
A 44 page booklet of CEDIA services is available here.
A Global Network Map
Service Map of Fiber Network
The March 2013 design paper for celec transelectric new optical transport network is found here
CNT Ecuador’s National Carrier
Source of CNT map page 21 of Estrella paper on Cedia.
Ecuador 2009 National Plan
Ecuador 2011 National Broadband Plan “The Digital Strategy 2.0 Ecuador”
National Broadband Plan http://www.supertel.gob.ec/ and
Ecuador 2013 National Plan http://www.buenvivir.gob.ec/
Ecuador’s NAP (Traffic Exchange Point)
Internet Society (November 2013) Connectivity In Latin American and the Caribbean, the Role of Exchange Points.
Estrella, Deigo (2011) “Assessment Of National Research Educational Network In Ecuador And Beyond, “Senecyt, Quito
Estrella paper on Cedia page 21 is the Source of CNT map
http://repositorio.educacionsuperior.gob.ec/bitstream/28000/775/1/T-SENESCYT-0360.pdf Diego Estrella provides Information on the formation and history of CNT, pp. 20-21
Freenet Mesh Software Development – http://battlemesh.org/
GLIF maps http://www.glif.is/publications/maps/GLIF_5-11_World_2k.jpg
GuifiNet and Global Freenets
Guifi Net http://guifi.net/en.
Wireless commons license http://guifi.net/es/ProcomunXOLN%23ProcomunXOLN
Guifinet – what it is, how it spread, maps and capabilities
Do-It-Ourselves Telecommunications: Guifi.net Commons Coming to US from Catalonia – In Spain & US We Are Now Building Network Infrastructure Held as Commons
November December 2013 COOK Report on Internet Protocol
http://www.cookreport.com/pdfs/nov-Dec_2013_CRpp.pdf The Global Free Network Movement[The Right to Telecommunications Self-Determination with Examples from Spain, USA, Argentina and France
January February 2014 COOK Report on Internet Protocol
Citizen Involvement in Freenets Living Labs in Catalonia and Hackerspaces in Oakland California — Guifinet to Join i2Cat Foundation for Collaborative Use of Both Infrastructures
Hughes and Satellite
Hughes Company http://www.hughes.com/technologies/satellite-systems
Hughes Broadband Connectivity
High-Quality Rural Wireless Services with Optimized IP
Lothberg, Peter (2013) TeraStream – A Simplified IP Network Service Delivery Model https://ripe67.ripe.net/presentations/131-ripe2-2.pdf
Ramirez, Renee (2014) Towards Intellectual Independence
Research networks and knowledge
SURFNet and eScience
COOK Report (Jan-Feb 2010) Building a National Knowledge Infrastructure. Download URL is http://www.cookreport.com/pdfs/knowledge.pdf
COOK Report (February August 2011) Fast Thinking: a Research and Education Network Renaissance. http://www.cookreport.com/wp/?p=362
For more information on eScience in 2014 see the videos and white papers at http://www.esciencecenter.nl/publications/
Hey, Tony A Journey to Open Access
and a general presentation by Professor Hey,
Data Transmission http://www.telconet.ec/en/services/datatransmission
Pacific Caribbean Cable Service http://www.telconet.ec/en/services/pccs
Pacific Caribbean Cable Service http://www.alcatel-lucent.com/press/2012/002755
Pacific Caribbean Cable Service
Turner, Brough: (March 2014) Internet Peering Impact on Network Architecture and Economics http://www.slideshare.net/Brough/internet-peering-annotated
10. Why I withdraw this paper*
The basic reason is that Daniel Vasquez, who apparently is effectively in charge of the entire budget, has unilaterally changed the conditions of my travel and told me so only after I enquired about what he had promised. When I complained to him, he sent me back a nasty threat that one more communication from me of any kind and he would give the order to end the collaboration. I have worked hundreds of hours over months on this. As I wrote earlier I will give up copyright only when paid as promised. Daniel has abused my trust. I regret offering to do this work and I insist that it not be used under any circumstances.
Reality turns out to be different. The reality was that Daniel Vasquez and Xabier Barandiaran two “hacktivists” from Spain had befriended Carlos Prieto who had worked for René Ramirez and now was Rector at IAEN and had budgetary authority. Michel visited Quito last September and, with much fanfare, the plan was announced. Michel set out to gather his research team. Inexplicably during that same period of time Carlos fired relatives of Renee Ramirez and all “hell” broke loose. “Hell” that after the appointment of a new rector was resolved in mid March. By the way Daniel, did you inform Michel of the firing and the possible consequences when it took place or was it after the fact?
* A lengthy detailed denouncement of the two Spanish citizens that have hijacked and destroyed the Ecuadorean initiative has been deleted. This extract captures the essence of the denouncement.