A Practical Navigator for the Internet Economy

Kick Starting the Commons and Preserving the Wealth of Networks

Why the Cooperative CAIDA Commons is Needed to Preserve Our Capability for Innovation

How to purchase this issue. $350 or $1400 group.

March 30, 2007 Ewing, NJ -- The May issue contains an essay on the economic significance of the Common's Project, and interview with Tom West CEO of NLR, a joint discussion of the commons with originators kc Claffy and Sascha Meinrath, and symposium discussion.

Executive Summary

Preserving the Commons, pp. 1-5

The re-verticalization of the telecom industry in the United States has created a situation where the ability of the open end to end internet to serve as a vehicle for continued wealth generation is endangered. The two giant phone companies see the net as a mechanism for the delivery of carefully controlled triple play. End of discussion.

Their inter carrier arms used to belong to the Co-operative Association for Internet Data Analysis (Caida). No longer. They have withdrawn behind closed non disclosure protected doors. They now own a substantial proportion of the internet infrastructure in the US and they have quickly enclosed it in an opaque vertical silo. Outsiders can find out nothing about performance or actual costs.

It is as though the airlines had been allowed to privative all data about their performance including all findings of the National Transportation Safety Board. The original Internet commons has been enclosed. The equity building operations of the giant telcos is being used in a short term short sighted manner to enclose the capability of the internet that has permitted innovation and wealth creation.

Our April issue introduced the plan to establish an open users cooperative based peering and transit backbone via the co-operation of NLR, Caida and an alliance of community networks. This issue explores, in more depth, the idea of a need for a neutral and open Commons.

Our Introduction looks at user centered activities at the edge of the network as the source of a new and ever broadening range of innovation and collaboration. Borrowing from Yochai Benkler's May 2006 monograph, the Wealth of Networks, it demonstrates how the Internet used as a public Commons makes possible the new user owned and controlled ability to engage in economic productivity and shape the development of politics and culture. In comparing Tom Vest's work (Measuring the Wealth of Networks) to that of Yochai, we complain that no one is really covering the totality of the infrastructure from the fiber ducts to the political effects. We try to recount - politely - that Yochai's optimistic new world of wealth created by users at the edge rests on a Foundation of uncertain strength - one that in the US the phone companies and cable cos have tried but so far failed to monopolize - a failure whose continued duration is uncertain.

We conclude that the CAIDA formulated Commons offers a critical opportunity to recreate a user owned and controlled infrastructure that can expose actual economic costs in running state of the art IP optical networks and in so doing enable local communities to move more confidently forward ion building their own infrastructure. We point out that the economic forces that have benefited from the wealth creating capability of the internet better invest in the Commons to enable its continuation. Specifically some folk need to see the stakes and make the investment to see that the Commons gets the necessary financial support which so far it does not have.

Tom West and NLR, pp. 6 -16

In an Interview Tom West tells us in some detail how National Lambda Rail was created and has grown into a national foot print with state connected optical networks that for the first time give America's Reasearch and education community the ability to operate directly and light its own networks when and where needed. An activity that may be three times as cost effective as the managed capability of Internet2.

The critical lesson to take away is that with the falling prices of the optics necessary to light user-owned fiber that the economic power for all constituents of just doing it yourself is overwhelming. Internet2 's approach may have been acceptable in the past. In the future it is no longer economically viable.

NLR does research and offers its members interconnection to global research networks and well as member managed features like Transit Rail. While Internet2, which as far as basic network functionality has never had that much to offer its members since they do not directly operate it, has been around for more than a decade NLR was established at the critical moment of low prices for fiber IRUs and with the necessary leadership to show what is possible when you own rather than rent.

Unfortunately in mid 2005 Doug van Houwelling approached NLR for a merger. Negotiations continued for a year before the collapse and Internet2 withdrew its two ownership shares in NLR. Although most thought the matter settled, Googling intensively on the merger discussions and aftermath shows that the US Education and research community was upset at having to pay for two research infrastructures and felt that it had missed an opportunity for great cost savings. Suddenly and out of the blue on March 10th came an announcement that the merger was on again. This time with a schedule that dispenses with both CEOs by completion in 2009. I admit strong bias on behalf of NLR and if the merger goes through hope it will not be at the expense of the NLR business model.

Meinrath and Claffy Provide More Detail on the Commons, pp. 17-22

In a short interview Sascha Meinrath (with extensive comments by kc Claffy) answers questions about the process behind the formation of the Commons and explains the synergy behind its founding forces.

The commons has multiple critical purposes depending on through whose eyes it is seen. KC makes clear that from her point of view it is a science experiment to measure users traffic. As KC puts it: More specifically, commons participants are agreeing to be on a (self)*regulated* backbone, where the regulations are the results of confirmed empirical analysis of what the backbone needs to perform well.? I have seen it as on opportunity to become independent of the controlling role of the phone companies are exerting throughout the national infrastructure. But as I elaborated during my panel at Pulver's Spring VON a third point of view is to think of the commons in more neutral terms - namely as a means of preserving an environment friendly to the edged based user dominated innovation that made the internet an engine of economic growth in the first place.

Sascha points out that ?Privacy issues have certainly risen to the forefront in the last decade. The have been fostered by corporate practices of not protecting consumer data and of commoditizing data to make a profit. People therefore have a well-justified set of reasons to be concerned about this. Different standards are one of the main reasons to go with a research community for the study of this data. CAIDA has been doing traffic research for a decade and has an important reputation to up hold in terms of its standing in the scientific community and it terms of the applicable institutional review boards that are built into protecting consumer, participant, and user data.

When I asked about deep packet inspection, kc had an excellent answer: There is no magic pixie dust here. if people want to be able to answer questions about the prevalence of spam and p2p traffic on this network, more than the tcp/ip header is often required. DPI is a normal part of network engineering anyway, if you prevent privacy-respecting DPI from being used for research, it's going to be the only disallowed use of DPI.

Symposium Discussion: pp. 23-49

An excellent report from ILSR. p..23

Localizing the internet: Five Ways Public Ownership Solves the US Broadband Problem, the Institute for Local Self-Reliance (ILSR) argues that public ownership of municipal wireline and wireless broadband access infrastructure can generate significant return for localities in terms of payback and revenue generation to pay for other municipal services. ILSR works with citizen groups, governments and private businesses in developing policies designed to extract the maximum value from local resources.

Susan Crawford, p. 25

A look at Susan Crawford's excellent paper The Internet and the Project of Communications Law. The paper criticizes the nearly exclusive focus of communications policy on the private economic success of infrastructure and ?application? providers, and suggests that communications policy be focused on facilitating communications themselves.

Susan links the on-line formation of ideas, processes and business to economic growth. She then demonstrates that internet which is access to each other is the incubator of and locus of progress in the area of sustainable growth. She demonstrates that an economically viable nation state depends on this for its sustainable future. She concludes by showing that structural separation of transport and content is the ONLY way to achieve these ends.

PON Architecture, p. 27

When and under what conditions do PONs make sense?

Bill St.Arnaud: PON of any flavour does not make sense. The biggest drawback with PON, whether it is WDM-PON, G-PON, BPON, EPON etc, is the fact that the Optical Network Terminal (ONT) at the customer premises must be under the control and operation of the carrier. This is required to insure that there is compatibility and synchronization in the upstream signal - usually based on a time slice protocol or an assigned wavelength. Snip

Herman Wagter: I agree with Bill. PON means that over 70 % of the value chain is locked-in (physical layer, transport layer), which drastically reduces the option value (economic sense). This makes no sense for infrastructures which are supposed to have an economic life of 30 years or more. You hurt your economy's potential to evolve and grow.

Coluccio: I don't think anyone here, including myself, is "for" PON. In an earlier post, Bob made a comment suggesting that wavelengths over PON would be preferable, and I merely demonstrated that it was already available.

Streaming Video: p. 31

Tom Evslin: I do understand, thanks to Tim Nulty, that an operator like Burlington Telecom can't get the content from the content license holders without agreeing to stream it but think this could (and should change) over time so that content owners and creators - NFL or indy producers, for example - do agree to make content available ala carte in which case we'll want to get it all through the "Internet" portion of the pipe which will need to be broader.

Encouraging the Growth of Netcos, p. 34

Editor: Denmark climbs to the top in the rankings of the World Economic Forum's Global Information Technology Report for 2006-2007. This discussion sheds some light as to why.Take a look at this article, from the January issue of Linux Journal:

It's about a small outfit in Copenhagen that does nothing but deploy single-play Internet connectivity. (I've written about them before, in other postings. Forgive the redundancy.)

What is the Cost of Building and Lighting Your Own Fiber? p. 37

Editor: One of the best discussions yet on this important subject.

Mark Cooper: What the Bell's said they would do made sense in a franchise environment, which is not popular among those on this list who revile the regulatorium. Here is the way I see the math.

Verizon's cost to deploy FIOS in its urban and suburban territories (it is selling off or not deploying in rural areas) is about $1300 per home passed (which is two and half times the figure for Denmark). If we assume a cost recovery factor of 25 percent, Verizon needs to recover about $27/month in capital costs from FIOS passed homes. Based on cable's current operating statistics it costs about $30 per month in operating costs to deliver video plus high speed Internet. That means FIOS needs a monthly charge of $57 per home passed. If you assume that they get half the homes they pass, that means they need a charge of $114 per subscriber.

The current price level they are getting is about $100, which includes telephone service, whose costs are $10 to $20 per month. They need about $130 per month. In other words, Verizon has a shortfall of $30 per subscriber per month. The shortfall arises precisely because there are redundant facilities in a natural monopoly situation. If the costs were $500 home passed (as in Denmark, but how much would it cost to lower the contention ratios to achieve the level of service that Verizon is offering?) or there were a regulated franchise the problem would be solved.

In the former case competition would prevent abuse and affordable high speed internet would be delivered. In the latter case, regulators would do a better job than the unregulated crummy duopoly we have in the U.S. The cost of capital goes way down (the cost recovery ratio is less than 20 percent) and there is no competition for subscribers. Monopolists might not be able to convince all consumers to subscribe (since, being lazy and slovenly monopolists, they are shitty at innovation and marketing). That is why they should be required to run open networks, so that innovation without permission can drive demand.

Later Cooper in response to Bob Frankston: The problem is political, not economic and always has been. I understand that engineers just wish the politicians will get out of the way, but politics is an inevitable and essential part of social life, as Benkler makes clear in his new book.

Cooper: We have a range of costs, from $500 to $2500 from the for profit sector depending on density, residence type and the extent of bloat in the company. What are the alternatives? We have a barn raising but no cost estimate, which could be free with volunteer labor and donated assets. Do we have estimates from municipal networks, wired and wireless?

Hertz: Dark fiber is just that; unlit, without electronics but often in a cable with other lit fibers, and needing only electronics to activate it. Buying lambdas is one step up. Prices are entirely dependent on location and seller. Rough range: $200-$2000 per fiber per mile for 20 year deal (IRU - Indefeasible right of use) plus recurring fees for O&M, ROW, licensing, power, unscheduled maintenance (e.g., backhoe blight), etc. Some deals we have done throw in the O&M for a one-time upfront bump to the purchase price but that is not common. Typical deal minimum is two fibers.

For the complete issue you must subscribe.


Kick Starting the Commons and Preserving the Wealth of Networks

Why the Cooperative CAIDA Commons is Needed to Preserve Our Capability for Innovation

The Wealth of Networks p. 2
Social Networks Versus the Old Regime p. 3
Yochai Benkler on Property and the Commons p. 3
The Commons as the Preserve for Innovation and
Wealth Creation p. 4

Why National Lambda Rail is the Most Cost Effective Approach for the US Research and Education Community
CEO Tom West Explains the Advantages of Do-It-Yourself

Why Not Own and Operate Your Own Network? p. 6
The Cost of a Wavelength p. 9
Economics, Organization and the Regional Optical Networks p. 11
Transit Rail p. 12
Rural Health Care p. 13

Why the NLR - Internet 2 Merger Talks Are on Again - Stakes Are Extremely High

A Coup d’Etat? p. 15
Economics a Forcing Factor? p. 16

The Co-organizers - kc Claffy and Sasha Meinrath- Flesh Out Details of Commons Project

Two Different Groups Interacting Produce a Synergistic Agenda p. 17
Traffic Security p. 18
The Purpose and Status of the Project p. 19
Establishing an Operational Foundation p. 21

Symposium Discussion Feb 17 - March 17 (pages 23 - 49)

Excellent ISLR Study of the Benefits of
Community Broadband p. 23

Mapping and Preserving the Wealth Producing Capability of the Internet, p. 25

PON Architectures, p. 27
Streaming Video p. 31

Encouraging the Growth of Netcos p. 34
What is the Cost of Building and Lighting Your Own Fiber? p. 37
Cooperatives p. 42
Cost Analysis p. 44
Pricing of Dark Fiber p. 46

Executive Summary p. 50