Canada and the Next Internet Revolution

Canarie Builds Next Generation Optical internets Government and R&E Leadership Push IP Cost Advantage

Special Report Shows Prospects for Customer Owned Fiber Networks in Rare Innovative National Environment

pp. 1-2

This special issue of the COOK Report documents and analyzes a profound revolution that is underway in telecommunications in Canada. There, under the leadership of Canarie the Advanced Internet Development Organization, the Canadians are building a nationally connected community owned infrastructure of dark fiber. They are lighting the infrastructure with IP over gigabit Ethernet over glass.

As one of their presentations says, they offer "A proposed strategy to make Canada the most networked country in the world and the first to have low cost Gigabit Internet infrastructure available to virtually all schools, hospitals, libraries and businesses by 2005."

As this issue shows however this is not future hype. Large sections of the public infrastructure are operational now. We argue meanwhile the United States, having invented the Internet, is in the process of giving away all leadership in its implementation. Internet 2 is being run as a subsidy program for the connection of universities - one that is devoid of the innovation of Canarie. While 75% of the school districts in Quebec either have completed or are installing their own dark fiber networks as a part of Canada's national public grid the FCC-imposed e-rate in the US means spending 2.25 billion dollars a year to subsidize the obsolescent copper plant of the local phone company. The program forces schools to buy service year after year and effectively prevents communities from building their own infrastructure.

Meanwhile, market forces under facilities-based telecom deregulation in Canada are pushing for the ad-hoc establishment of a private/public-sector partly customer owned and operated IP over gigabit Ethernet fiber network linking schools and municipal governments to province wide network which in turn link to the Canarie national trans-Canada optical backbone for research and education traffic. The result is the first large scale national infrastructure that operates (except where it must interconnect) completely independently of the global public switched telephone network.

Canarie's Role in Charting Canada's Telecom Future

Increasing Bandwidth & Lowering Costs for R & E Community St Arnaud Offers Overview of OBGP, Fiber Infrastructure Growth, Condominium Fiber Business Model and Scaling Issues

pp. 3-8

We interview Bill St Arnaud the Director of Network Projects at Canarie. Bill explains the development of the first Canarie Networks and goes on to explain that he serves the Canadian research and education community by developing ways to make broadband Internet access several orders of magnitude less expensive for his clients. One of these ways has been in the development of news methods of building customer owned - or as Canarie likes to call them - "customer empowered" dark fiber networks.

In one of the papers on the Canarie site we find "lower prices for fibre is leading to a shift away from carrier-owned infrastructure and towards more customer- or municipally-owned fibre, as well as to innovative sharing arrangements such as fibre "condominiums". "Condominium" fibre is "un-lit", or "dark" fibre that is installed by a private contractor on behalf of a consortium of customers, with customers owning the individual strands of fibre. Each customer/owner "lights" their fibres using their own technology, thereby deploying a private network to wherever the fibre reaches, perhaps including carrier COs and ISPs. The business arrangement is comparable to a condominium apartment building, where common expenses such as management and maintenance fees are the joint responsibility of all the owners of the individual fibres."

The Canadian government has made a decision to use Canarie to build a public sector national infrastructure that can use the thousand fold cost advantage of the new IP over gigabit Ethernet over glass to enable a national broadbrand infrastructure for education, government and research telecommunications - in short everything that could be considered non commercial.

In order to leverage most effectively the ability of their public customers to use their networks to communicate with each other, Canarie has been developing (as we have reported in an interview with St Arnaud in our November issue) optical extensions to the Border Gateway Protocol (BGP) or in this case OBGP. This will permit each of the new k-12, municipal and university public sector networks to reach out and peer with likeminded counterparts by using OBGP to take a lambda or wavelength of light and establish a virtual optical cross connect at a mutually agreed upon exchange point. For the first time these networks can manage and direct their bandwidth without having to buy any carrier services or having anything to do with carrier clouds.

St. Arnaud's strategy is to test his concept and submit before the Spring IETF an Internet draft for formal IETF standardization. If this procedure went smoothly the work could be finished and OBGP could be found in the capabilities of all Cisco and Juniper routers by year's end. Unfortunately its is impossible to predict whether the standards process will go smoothly. We learned from Bill that Canarie has now raised the stakes by establishing a second OBGP group at Carlton university in Ottawa with the task of writing the code necessary to get the job done and testing that code in an inexpensive optical switch available now from JDS Uniphase or soon from a new start up named Edgeflow. This switch sits in front of the customers router and will give each of them OBGP capability regardless of what comes out of the standards process. The result will be that without dependence on a carrier many small entities will be able to generate and manage their own gigabit Ethernet bandwidth in a way that a year ago only a Fortune 1000 company could in anyway likely to be able to do. "Customer empowered" - indeed. St Arnaud however is clearly making a major wager on OBGP. In this issue we summarize his October 30 2000 proposed architecture design for the establishment and funding of CA*net4. Our summary shows that a major purpose of the new network will be to use and develop OBGP within the customer owned national infrastructure that Canada is building.

Finally we look at St. Arnaud's new draft "Scaling the Internet" first posted on December 12, 2000 and most recently revised on January 3, 2001. Here he looks at bandwidth demand for large scale internet transit backbones like UUNET's that are reported to be doubling yearly. He raises a hypothesis that also points out another new and disturbing trend. It begins to look like the numbers of open network connections between computers are growing even more rapidly. As peer to peer connections move to personal computers huge numbers of machines can suddenly maintain perhaps dozens of connections per machine. Even if one supposed that the growth of new users of the net would slow or even cease, there are numerous reasons to suppose that the numbers of connections will continue to grow. This process is demanding the acquisition of even more backbone bandwidth and its demands may be doubling every four months. Mike Odell of UUNET has spoken of needing petabit backbone links within two years.

Hopefully CAIDA or a similar entity will be able to test this hypothesis within the next few months. It it is verifiable, it will demand that major changes in internet topology be made. Ironically the widespead use of OBGP to get as much bandwidth at the edges of the network as possible and off the largest carrier operated transit backbones may be the only way to continue to scale the growth of the Internet.

A Revolution in the Cost of New Fiber Networks

IMS Executive Explains His Role in the Creation of Fiber Brokering and Condominium Dark Fiber Networks

Plummeting Cost Makes it Possible to Bring Fiber to Most End Users

pp. 9 -14

We interview Robert Proulx, vice president telecommunications of IMS Experts-Conseils. Robert explains how during the past four years he developed the art of brokering and condo fiber builds at first in Quebec and now in seven different Canadian provinces. At first he worked for Hydro Quebec, the provinces electric utility building and marketing a commercial fiber network on their right of way. Bell Canada became nervous, stepped in and bought all the network's extra capacity. It moved too late because by the time it did, Proulx knew who had fiber, who did not and where the fiber was. He realized that he had a business model for leveraging the knowledge he had developed. His customers still wanted fiber. They'd just have to do it on their own.

He explained what happened in the following way. "I'm the engineer hired by the organization to build a network. The first thing I do is to find a partner who wants to be an investor in the project. The partner becomes an investor in the project and I say we have a project that will be owned by five or six companies and we prepare the document and we go out with an RFP to build. And the same contractors that are being used by Bell Canada will build my network. It's an RFP to build. Not to provide services, not to provide dark fiber. It's to build network that will be owned by a consortium. But we also are responsible for building the consortium itself.'

Now Proulx would have a much more difficult time were it not for the regulatory environment in Canada. In the second half of the 1990's the cable TV companies fought for and obtained laws giving them open access to the poles and conduits owned by the incumbent local exchange carriers and provincial electric utilities. Anyone with a non dominant carrier license which is easy to get can demand and obtain access to poles and conduits from which to string new fiber. The payment to the pole owner is set at uniform inexpensive rates. It was this easy and affordable access to right of way (established uniformly by national policy) that facilitated the planing and then implementation of Canada's on going boom in the construction of customer owned dark fiber networks.

One of the most significant of Proulx and IMS' achievments has been the network they brokered for RISQ. RISQ is the Reseau d'Informations Scientifiques du Quebec, an organization owned by all the universities in the Province of Quebec. It exists as part of the Canarie program furthering province wide fiber nets. A carrier-built network without sharing would have ost 100 million dollars for 3,500 kilometers of new fiber and right of way.

As the map of the St Lawrence River Valley of Quebec shows on page 12 of this issue, Proulx knew with exquisite detail where the fiber was. He pursued a successful strategy of giving the owners buy-in to the project by exchanges of strands of fiber. This created a situation where the only parts that he had to build was to fill in the gaps between already existing fiber.

Clash of Broadband Private vs Public Sector Business Models Will Impact Canadian Economy

Francois Menard Describes Complexity of Canadian Regulatory, Carrier, and Content Struggles

Suggests That Viable Outcome Likely Only With Community Networks Infrastructure

pp. 15 - 35

The interview with Francois Menard at nearly 22,000 words is the longest we have ever published. There is a good reason for the length. Menard knows the intricacies of the content combined carrier business model from the inside and talks about them with a detail not found elsewhere. But he does far more than this. He understands as few others the efficiencies and power of the 'pure' internet model. The conversation shows how the impact of a potentially pernicious business model combining ownership of content and network infrastructure is likely to lead to a market share ware between Videotron and its new content based owner Quebecor, along with other cable carrier partners, on the one hand and Bell Canada, and other incumbent local exchange carriers,on the other hand.

For a period of time in the purest of dot com plays it looks as though Quebecor may use its pension fund fueled acquisition of Videotron to give away internet access in order to gain market share. Menard sees the commercial players battling it out to maintain vertically controlled cost inefficient corporate empires. When the battle dies down most Canadian ISPs will be destroyed and the few corporate visitors may try to recoup their battered bottom lines by raising rates drastically

Our conversation finally takes the struggles for commercial market share and shows how they will impact developing national public infrastructure in Canada and ultimately globally. What is presented is a synthesis of commercial, policy, economic and technical impacts of the ongoing technology deployment. The deployment is shaped on the one hand by efforts of the vertically integrated content mega-corps that have acquired or been acquired by huge carriers.

These people see the Internet as a way to broaden their control over their 19th century empires. On the other hand the 21st century optical Internet technology is decentralizing control and ownership pushing it to the edges and into the hands of community or customer owned - "empowered as Canarie likes to say" - networks. It is unusual for anyone individual to be expert in both these areas. It is even more unusual for such a person to be able to describe the likely impact of the ongoing collision between these forces that are both very much opposed and inextricably linked. Menard does just this. The vast scope and complexity of what he describes is the major reason for the extraordinary length of the interview. What is articulated here has, we believe, never been said before openly and in one place.

He described for us how he acquired his knowledge. "In working with several Internet Service Providers faced with the prospect of seeing the business models which they had pioneered destroyed by the established incumbent carriers turned into ferocious competitors, he became especially cognizant of the potential repercussions that the recent wave of mergers in the telecommunications industry and the even more recent mergers between Internet access, cable, telephony, television and printed press conglomerates could have on the Internet Service Provider industry. Having spent the last few months building an Ethernet metropolitan optical network for a competing cable television carrier, he realized that he could take his vision much further with community networks, and decided to join IMS to pursue his vision."

A few sound bytes: "While I want to know what's the true cost of an ad on TV all I have is an offer from Quebecor. The offer states that the only kind of advertising I sell you is this bundled scheme of newspapers, Internet, and TV."

"What we have in fact been explaining is the difference between the new business models of incumbent telecommunications carriers going wild with media mergers and the traditional ISP's which cannot afford to buy into this. What ISP's are trying to provide is a monthly subscription based service, not cross-funded by advertising or e-commerce taxation or captive portal kinds of services."

Menard is interested in the potential of municipalities to form network buying cooperatives on behalf of their residents. He suggest that it could "preserve the quality of the telecommunications service that we have today in a world where we clearly suspect that that if we leave it to the private sector, it will degrade rapidly as the telecommunications services become focused on content exclusivity and advertising-funded. The utilities and municipalities can change this as they are entities clearly known to have no economy of scale from the selling of advertising. IMS also can do its share of the effort to alter this trajectory, but it cannot be out there on all markets at the same time. Then again, it may not be all energy utilities and all municipal administrations which may be receptive to these new opportunities."

"We could imagine certain areas where the public sector would no longer be purchasing any services from the incumbent carriers aside from local PSTN interconnection. The business that would be left for the carriers in these areas would be formed of a mixture of shopping malls, business condos, and residential. It's a lot of business taken away from the incumbent carriers. From that perspective, the landscape of telecom will require a tremendous amount of consolidation once all these guys are no longer buying services from the incumbent carriers. I would suggest that we've proven that most phone companies have no interest in providing services based on the best networks that they can design because it would cannibalize too much their existing businesses. As the public sector in Canada is showing to this date, you have to build your IP network on condominium fiber if you want it to reap all of its potential benefits."

"The task at hand is to structure the resale of the telecommunications facilities of the electric utilities so that they provide telecommunications infrastructure with a level of openness suitable for implementing competition between telecommunications service providers which base their revenues on selling advertising and those who base their revenues on selling services."

"It used to be because of old technology that you needed public subsidy to provide services at affordable prices for the most people. Only fifteen years ago, mechanical switches for placing a telephone call across the country were still widely in operation. The responsibilities for managing this network were colossal in comparison to today's technologies. Looking at the current state of the art technologies we must ask if residents are now well served by far distant entities which define what services are with the aid of a public subsidy license from national regulators. Or should we eliminate the public subsidy license in favor of what can now be built by the residents of those very communities in a free market?"

"Scaling the Internet" - Excerpts from a Draft

Bill St Arnaud in New Paper Examines Traffic Issues His Hypothesis Suggests That Scaling Problems May Force Traffic Off Backbones

pp. 36 - 37

Excerpts from

Optical Community Networks - A Canarie Presentation - September, 2000

pp. 38-44

Excerpts from . The amount of cost information on these slides for building various community and municipal networks is extremely useful.

CA*net4 Design Document - A Canarie Presentation - October 30, 2000

pp. 45 - 49

Excerpts from

The design philosophy is fascinating. "Research and Education networks must be at forefront of new network architecture and technologies. But should not be duplicating leading edge developments occurring in private sector. Should be undertaking network technology development that is well ahead of any commercial interest. But any network architecture can only be validated by connecting real users with real applications and must solve real world problems. Test networks per se are not sufficient. There is a growing trend for many schools, universities and businesses to control and manage their own dark fiber. Can we extend this concept so that they can also own and manage their own wavelengths? Will "empowering" customers to control and manage their own networks result in new applications and services similar to how the PC empowered users to develop new computing applications?"