Shaping a New Science Infrastructure

Global Collaboration Mandates Inventive
Solutions to Resource Constraint Problems

Last month (April 2009) we began to explore Harvey Newman’s discovery of the necessary role for data networks in the world of research in high energy physics. We traced its development in the global network that enables the use of the Large Hadron Collider and examined the Collider itself and the research  that it enables.

The concluding portion of our interview with Harvey in this month’s issue explores the nature of the global optical virtual private network he created and the technology innovations that his group have made as part of their effort to make an unprecedented amount of data available to the global community of high energy physicists and do so in such a way that they gained the most cost-effective use of the technology on which they depend.

Harvey explains how his MonALISA application that is used in monitoring the LHC global VPN enables operational costs to be kept at a reasonable level and describes why, with the rapid escalation of the amount of data needed for delivery into the hands of a distributed global network of researchers, capital costs of hardware and optical wavelengths are a more significant issue than operational expenditures.   The interview covers via funding strategies chosen for the development of the multi tiered LHC Network.  A key insight is the funding of local universities to set up regional centers as part of a network in a way that enables them to become global players. To make his monetary resources grow as fast as possible he also explains how all the development of FDT (the fast data transfer application) has enabled him to, by means of optical circuit switching, run his wavelengths at about 85% of capacity or even higher rather than the more normal roughly 30% common to the best effort Internet.

He explains in some detail how over the last seven or eight years he has participated in the development of the global new e-science community where as the representative of high-energy physics he has gone into countries in Latin America, Eastern Europe, India and places far outside the general purview of European or American science to get local leaders to work with their governments to establish fiber-optic research networks that expand the reach of e-science in ways where local governments find the necessary investment in the enabling optical network infrastructure to be a cost-effective means of increasing local intellectual capital and competitiveness.

We close with a detailed look at  EVO (Enabling Virtual Organizations), the peer-to-peer client/server based tele-presence collaboration tool that he has developed in parallel with the network.  Its capabilities and uses are extremely impressive.  Its costs are also low. But it is not open source and is dependent on licensing through Caltech.

Finally, the tools that his group have developed have the potential to be used to solve some of our infrastructural problems. Unfortunately, predisposition is that federal funding should not be used to give away something that a private corporation could sell, is the primary shaper of policy.  As long as this is the case, the benefits of these publicly funded efforts are less likely to pay social dividends in some of the areas of education, public administration and local government where they could otherwise become extremely useful.

Symposium:  Issues of Policy and

Summary – As the new administration looks at policy and the economic meltdown continues, regulation is up in the air.  We start with the February 25 Supreme Court Decision that essentially removed anti-trust remedies to telco pricing practices and move onto a discussion that says muni-networks that survive in the face of hostile duopoly may have no choice but to offer their own services to citizens.  From there the discussion moves into support structures and control there of where we trace a spirited argument between Jim Baler and Erik Cecil over if or how fundamental regulatory premises should be swept aside.  From that point on we conclude with a discussion of the impact of the Chicago School in regulatory economics and a behavioral economic point of view as being something that merits a new approach in creating an alternative approach.

Feb 25 Decision voids Antitrust as Remedy for carrier abuse. p. 43

Cecil: As Brian notes, if the rates are legal - e.g. approved or, in this day and age, even countenanced by a regulatory authority,  .  .  .  you are going to have a damn hard time of convincing a judge that there has been an antitrust violation. 

Savage: But even if what AT&T was alleged to have done would not be an antitrust violation, that does not mean (or at least does not have to mean) that its conduct is not a form of "unjust" or "unreasonable" or "discriminatory" conduct, or some similar formulation based on regulatory, rather than antitrust, principles.

Cooper: The courts will be the last to change.  The judiciary is filled with lawyers who grew up and entered legal practice during the ascendance of Chicago School market fundamentalism.  . . . I have pushed hard for both the DOJ and FTC under Obama to start a claw back campaign, reviewing all the bad decisions that have been so harmful, so that when the judge cites some lousy precedent, they can pull out the study and show why that precedent was wrong.

Why Fiber is Always
Superior to Copper p. 49

Rood: I think the entire idea that it is the high speed broadband, that brings benefits is misguided, it is the very different market structure that raises competitive levels with open access to fiber, that brings benefits. . . A nation, or any sovereign entity, should do an OPEX calculation comparable to the one discussed in the study above at least once, and then ask themselves the simple question: are they trading off the CAPEX required for rewiring against higher OPEX with DOCSIS and DSL technology paths, and is it this trade off that brings higher societal costs when compared to their capital outlay?

Cecil: p. 53 any anger directed at any incumbent business per se on this list is misdirected to the extent it is really frustration with market and regulatory design. Outside of documented bad acts there are no "good" or "bad" actors here.  All are rational economic actors given their business models and methods. 

I think the greater point, too often lost in the specifics, is that the overall regulatory design is simply out of date.  From this lawyer's perspective, for example, the regulatory system in the U.S. is deeply flawed not necessarily because I think it does favor incumbents, but primarily because it is so imprecise and unpredictable to the point of having very little internal consistency.  This is actually worse than always favoring incumbents, but I think we can do a whole lot better than that.

Municipalities Do Need to Offer Services  p. 55

St Arnaud: We have been in long series of discussions with an out of territory ILEC. The local team were very keen on our business model as they knew that they only way they could compete with the local ILEC is by someone else underwriting the cost of the infrastructure. But when the local team took the business case back to head office in their home territory the project was killed. The senior management was terrified that other ILECs would invade their turf with a similar strategy.  The ILECs are much more interested in protecting their local monopoly rather than competing in someone else's backyard.

So we are stuck looking to provide cable and Internet services ourselves.  This actually requires a bigger investment than deploying the actual fiber (at least for a small scale project), because you have to purchase volume Internet transit to be competitive. But the biggest challenges is arranging for delivery of cable TV services. This is a Byzantine world if I ever saw one and much more difficult and costly than arranging for Internet service.  I know that many municipal networks like Layfette and others have run into the same problem.

Budde: It looks to me that the future has to be in a utilities based open fibre network. In order to built up extra higher margin revenues a value added infrastructure company/ or division should be attached to the utility operating data centres, content hosting, cloud computing, billing, network management, and offer that to individuals (UCG) and companies. In other words stay totally away from the content itself, ISP operation, etc.

Once you have this structure in place content owners and content aggregators will on their own terms start building their business models on that infrastructure and they might consider using the value added infrastructure services, but for the rest want to run this business totally separate and independent from the network operator. Of course they will only become interested if such a network has sufficient mass, that's why very little activity is taking place on the current FttH networks as hardly any of them have reached any mass that warrant these media and services companies to start using them. They rely on the operators building a network based on the principle: 'built and they will come'. They don't want to make any commitment upfront but are happy to come onboard once you have got the network in place.

So, in most situations, only utilities based organisation can afford to built such networks and this again makes munies an ideal partner in building FttH networks, for all the 'common good' reasons that we have been discussing extensively

New Ways to think About Infrastructure  p. 58

Kevin Barron
: Fiber is public infrastructure; government builds public infrastructure.  . . . If we stop the petty squabbles over layers 0 and 1, we could actually see some of the 10-to-1 return on investment in BB infrastructure. My fear though, is that even with the collapse of the market, the religious dogma of the "free market uber alles" will continue to obscure the fact that fiber is basic infrastructure (which *requires* public ROW's in most populated areas). We continue to be captives of that dogma, even while our economy melts down, our retirement plans go up in smoke and unemployment hits record highs as a result.

The bogeyman of "big government" has been used for 20 years in this country as a smokescreen to effectively capture it and primarily turn it into a hired gun for industry.

Cecil: The solution I specifically proposed was to eliminate the fee-generating dance - because when I'm running and managing all sorts of these kinds of litigation I get tired of the fees.  And it's no secret that these disputes tend to be self-perpetuating.  It is, therefore, the very basis of a framework that pits all sorts of common interests against each other. . . . [On p. 61 Cecil proposes a way to “rationally fund” dark fiber utilities.

What I do not think is useful, however, is any perpetuation of failed regulatory models that result in incessant grinding of opposed interests that do not necessarily have to be opposed.

Policy Strategy

Budde: The way I would like to handle this is to maintain the more radical call for change, of course in the hope that we will be able to get at least some of this implemented at a minimum we will get some people thinking. (As a matter of fact we do already have some very senior people in the Administration at least interested in these more radical approaches.) This will create cracks in the system and than when that happens we need to be opportunistic on how to wiggle ourselves into these cracks to establish changes and that will be a matter of give and take, more along the lines Jim is talking about.

In one of our previous discussions we discussed the pros and cons of structural changes vs. grass root changes (e.g muni broadband). If, through guerrilla warfare, we could get FttH muni penetration reaching 10% or so I bet we will start seeing cracks appearing in the incumbent fortresses and that would accelerate the change enormously. Consequently our bold activities when well executed don't have to see the system start to change overnight, we can be the lever that actually allows the system to implode on itself. You can see in some of the European and Asian countries how quickly things can change once the incumbents decide that they have milked the old system for long enough and start moving toward a more open network environment. It can happen rather quickly (2-3 years)

This is important in framing issues.  The Overton Window moves.  You start by pushing an idea that is outside of it, and it sounds nutty, but things that had previously been seem as radical suddenly look more acceptable by comparison.   . . . .

So let's talk about radical and even unthinkable ideas.  If we talk about the unthinkable long enough, our ideas that are merely considered sensible will have a chance at becoming policy.  And it can keep moving.

What Behavioral
Economics Tells Us p. 75

Savage: What behavioral economics tells us in the short run is several things:

(1) Intelligent regulation is not easy.  You actually have to know the devil is going on in the markets and industries you want to regulate.

(2)  Actual people, in their role as consumers, make systemic and predictable errors in making certain types of decisions.  (Someone listed some of the key ones, e.g.: status quo bias, option overload, fear of loss of current benefits more than outweighs equivalent potential of gain).  It follows that people who would sell to them have systemic and predictable incentives to exploit those errors.  These are not problems that are fixed or avoided by nominal “competition.”  Therefore, from a consumer protection standpoint, one should focus on the details of consumer transactions more than we have.

(3)  If, as noted above, firms can maximize their profits by exploiting flaws in the way people are wired up to make decisions, then assuming the
firms themselves are rational profit-maximizers (but see below) they will misallocate their own resources towards those types of activities (e.g., new marketing plans, long-term contracts, etc.) and less towards what neoclassical theory would say they would do (product improvements or price decreases to attract customers).  This suggests that some greater degree of regulatory inquiry/oversight into product development processes might be appropriate.

(4)The flip side of people making bad decisions in the context of their individual choices as consumers is people making bad decisions in their role as
managers of businesses.

Peering p. 77

I see force in the argument that in certain sectors of the economy; and telecoms is one of them, the US" legal system" lacks the essential characteristics to be called a 'legal' system at all: it is unequivocally a political system. After all, the FCC is a creature of Congress, passing rules is an executive branch function, weak or ineffective judicial oversight in a system that allows the political appointment of enforcement agencies and judges is a questionable legal system at best: if the legal system is not separate and independent from the political system it fails one fundamental test for the rule of law.


Shaping a New Science Infrastructure
Global Collaboration Mandates Inventive
Solutions to Resource Constraint Problems

Editor’s Introduction                                                                             p. 1
Building the Global VPN                                                                        p. 1
Capital Constraints Meet Network Management and
Growth Needs                                                                                         p. 4
Lambda Switching Advancements – VCAT and LCAS – FDT
Gets More out of Each Lambda                                                               p. 7
These Networks Need Dynamically Provisioned Circuit Overlays         p. 13
Digital Divide Issues                                                                             p. 15
Monitoring Regional Connectivity                                                        p. 20
Digital Divide Workshops – Putting Brazil on the GLIF                      p. 21
Other Workshops                                                                                    p. 24

Enabling Virtual Organizations
A Low Cost but Highly Efficient
Global Telepresence System
p. 27
The Economics of EVO                                                                            p. 28
More about the Architecture and Use                                                     p. 30
OpenGL’s Use in EVO                                                                              p. 34
A Multiplicity of Capabilities                                                                   p. 36
Summation                                                                                              p. 40

Symposium Discussion  February 18, 2009 - March 17 2009

Issues of Policy and Infrastructure

The Feb. 25 Supreme Court Decision  on
Anti-trust in Telco Pricing;
p. 43

By the Time Economic Thought Impacts the Law it’s Already
Ten to 20 Years Behind Current Thinking                                            p. 45

Why Fiber is Superior to Copper -- The
Importance of Technology Choice
p. 49

Municipalities Do Need to Offer Services
Given the Obstacles that Vertically
Integrated Duopolies Impose p. 55

Reorder the Process                                                                                  p. 57
Ways of Thinking about Infrastructure from Technicaland Legal Points of View p. 58
Two Very Different Legal Approaches                                                   p. 60

Rationally Fund Dark Fiber Utilities p. 61
“Interstructure”                                                                                        p. 63
Defining an Agenda to Match Our Challenges p. 64
The Overton Window                                                                                p. 67

Behavioral Economics Versus the
Chicago School
p. 70
The Real Culprit - Clinton Captured by Wall Street                          p. 72
Approaches to the “How To’s” of Regulation                                       p. 74
What Behavioral Economics Tells Us                                                     p. 74

Problems of Free Peering                                                                   p. 77

Executive Summary                          p. 81