A Practical Navigator for the Internet Economy

ICT and E-Science as an Innovation Platform in The Netherlands

A National Research and Innovation Network
 What Can the US Learn from Dutch Experience?


Introduction  pp. 1-3

This issue examines the continued evolution of both fiber infrastructure and optical networking in research, education, and tech transfer networks and network projects in the Netherlands.  It also looks at the Dutch planning process for directing economic investment in ICT in ways designed to achieve, through innovation, maximum economic impact.  

The introduction and conclusion extensively discusses the programs of ICTRegie – the Dutch ICT Research and Innovation Authority.  They reflect a process set up in 2005 whereby Government research funding is put through a process that is designed to create proposals that are in keeping with nationally defined goals allied to the nation’s social needs as well as its economic strengths.

The mission statement is what President Obama should emulate in his Office of the CTO. “First to introduce unity and consistency to the strategic direction of the ICT research and innovation by means of the development of a national strategy that enjoys broad support and, second, to ensure ongoing strengthening and appropriate dynamism of a Dutch ICT knowledge infrastructure geared towards high social and economic yield.”

A New National Infrastructure, pp. 4-14


Kees Neggers, SURFnet Managing Director, explains how innovative technologists used the intersection of smart Dutch telecom liberalization policies, and then the dot com boom to build competing fiber rings around Amsterdam creating an attractive environment that allowed the city to become a global hub for layer three traffic at AMS-IX and layer one and two optical light waves at NetherLight.

Kees explains the sequence of events leading to SURFnet’s solving the last mile connection by means of fiber for all its members in 2004 and becoming the first national all optical research network at the end of 2005.  This enabled what they call hybrid networking – doing as many tasks a possible at layer one and two and as few tasks as possible with much more expensive electrical and routed layer three networks.

Working with an all optical network made the assignment of light paths possible, first to institutions, then to departments and ultimately to individuals.

The five year GigaPort project running from January 1 2004 through December 31 2008 made optical private networks possible for SURFnet members.  It also made possible the establishment of a resource management platform built on a foundation of grid computing, a layer of web services allowing users to connect to applications and to instruments as well as super computers.  On top of the resource management platform is one of generic e-science services (the light blue box in the diagram on page 11) in other words software tools that support research in individual fields.

Within GigaPort one has cutting edge technology supporting not only basic research but also tech-transfer from universities to companies like Philips, Unilever and IBM.  It also has valuable experience with the complexities of making it work through ways to build and administer the corporate networks, to the software tools needed to take advantage of the lightpaths to needed training to enable users to benefit from the technology at their disposal.

Most significant is the decision of ICTRegie to recommend continued support for SURFnet and GigaPort as a basic platform for technology development in the Netherlands.  As Harvey Newman says at the very end of this issue, while the private sector is critical without roads open to all we will be placing a huge handicap on our possibility for further progress.

As Kees Neggers said: In our view innovation needs the network and hardware resources plus the software and the people that have the knowledge of how all of these layers interact with each other and in addition to this knowledge you still need a major outreach effort to involve the users for which it has been created.

[And Cees de Laat added]:  From the light blue box you need people working with every discipline to make it happen, because the people in the disciplines themselves cannot do that.  You can not expect the biology professor to understand how to go out and identify and hire the people necessary to teach him and his colleagues how to use the high performance tools.
 

Hybrid Networks and e-Science Future
Development, pp. 15 -40


Cees de Laat takes us through the events that began with SURFnet’s 2001 decision to work with Starlight in Chicago and CENIC in San Diego as well as a few other locations to explore the use of dedicated light waves of one to ten gigabits.  The experiences of the iGrid2002 meeting led Cees and his colleagues to begin work with circuit switched layer 2 lightpaths wherever possible.

Working with extremely demanding users who needed their own lightpaths led Cees and his group to think about tailoring the network to the needs of the user so that, unlike layer 3 routed networks that are stupid procrustean beds into which everything must fit, the networks could become much more flexible and adaptive for the population that the network is designed to serve.

As Cees explains: “what I am describing is a kind of peculiar programming language or programming environment where your networks are just subroutines, your data are more subroutines, your solutions to solve the problem are also subroutines. You then say optimize and solve and it will then work out the most optimal way to achieve what you want to have done.

What all this means – and this is the most fundamental thing to get one’s mind around – is that your network becomes just part of your programming environment.

Normally you have your data and your computing and a “stupid network” that is an unmodifiable “given” that you have to play along with.  But here your network is just part of your toolset.”

Later he elaborates: “In our group back home we are creating objects for virtualizing and programming wavelength switches and photonic devices which directly talk to fiber; Micro-Electro-Mechanical Systems (MEMS) devices that can connect fibers so that we have flexibility at the fiber layer. And at the Ethernet layer, we can do similar things manipulating Virtual Local Area Networks (VLANs). We address the Ethernet layer and the packet routing when we need that.

If you can manipulate all these layers and have also vertical and horizontal knowledge in every layer, you can do the magic and you get a perfectly integrated multilayer hybrid network that is optimized for your application.”
 
Cees goes on to elaborate the kinds of e-Science projects that these lightpath networks support.  He then demonstrates in some detail his VL-e (e-Science Virtual Laboratory) applications.

I said to him: what you are explaining is how to build a networking environment that combines this use of equipment, technology, and energy in such a way to give the user applications that meet his demands at the minimal necessary cost?

De Laat: “Yes. But to do this one must expose the network to the users so that they understand what is happening and can see the benefits of not always working at the highest and most expensive layer -- namely the routed layer. One needs to make sure that your users understand the benefits of operating in different layers of their optical network. The users need to grasp that if they endow their applications with the ability to intelligently traverse the layers they can open up all manner of increased possibilities to better performance at less cost.”

For the most part the rest of interview describes how routing or path finding is done in his hybrid optical network.


Interactive Networks, p. 41-49


I have written up with illustrative photos a demonstration by Rudolf Strijkers of what a user interface to application embedded networks looks like and the capabilities it permits.

What we discuss is a prototype of a user interface with a multi-touch-sensitive screen that allows a user to tap on the tools he wishes to select and with his finger to draw the paths he wishes in order to activate nodes on a programmable network.

As Rudolf pointed out: This is the first prototype of what we are calling “Interactive Networks”. In interactive networks humans become an integral part of the control system to manage the next-generation of programmable networks and Grids. The main design principle is this; by virtualizing the configurable and programmable properties of network elements as software objects, any aspect of a network infrastructure can be manipulated from computer programs. What we show here is an implementation of an interactive control system concept for user programmable networks, which applies the architectural concepts we have developed in our research.

At the end he concludes: With the information given by the network, and currently we support continuous measurement of delay, jitter, bandwidth and throughput, one or more operators at once can write programs that automate decision-making. This opens the way for automated network adaptation in a user-friendly environment.

For example, now we can say, certain paths should avoid busy parts of the network. By using only standard functions in Mathematica, it is already possible to write a simple program that uses the real-time throughput information to continuously reroute one or more paths that avoid busy parts of the network.


Conclusion  pp. 50-58


The technology that is discussed here is inevitable and inexorably extensible and usable at reasonable cost where open access fiber is available.  The countries that implement this openly and broadly as an infrastructural system of roads and highways will derive enormous competitive educational and economic benefit.  Those that don’t will never see either the  educational or the economic benefits.

We touch on the OptiPuter and OptiPortals and their significance for science education. We also extensively discuss the Netherlands ICTRegie five year plan and show how it can be used to further distributed edge based collaboration and consensus in allocating economic resources.  We offer a seven point national fiber infrastructure plan for the new administration.

Be it a new ARPA or an American ICTRegie, our new administration MUST grasp the present opportunity to sweep out the wreckage of cronyism and speculative capital. In the national and public interest like Roosevelt did in the 1930s, it must build out a “rural electrification system” - this time an interstate highway system of optical fiber.

I conclude with a mention of BT becoming a services based cloud computing platform that I made to Harvey Newman.

Harvey Newman:  I agree with this. And it makes me think - Often, to make one's case for innovating at an incumbent, one is constrained to say "What is the business model of the Internet ?"

In  this  case  what  they  are asking is: How do companies ("how do I") make a profit out of this ?

So I say - Only the highway-building model  ("building the nation from the bottom up") will get us out of the past.

And if we do not do this - there is strategic as well as economic risk. Since any nation can do this.”



Contents


ICT and E-Science as an Innovation Platform in The Netherlands

A National Research and Innovation Network
What Can the US Learn from Dutch Experience?


Before Resource Use Careful Deliberation and Coordinated
Planning Needed                                        p. 1
E-Science                                                p. 2

Adopting a New National Infrastructure

How Fiber Became the Basic Platform for Education, Commerce, Technology Transfer, and Research in the Netherlands

Sparking the Fiber Revolution                                    p. 4
The Significance of AMS-IX                                p. 5
The Germination of the Hybrid Network Approach in 2001             p. 6
Adoption of Lightpaths - Not only for Science but also for
ICT Departments                                                p. 7
Dutch to American Analogies and the E-Science Virtual Laboratory    p. 9
The Optical Network and its Platform for Collaboration Become
Basic Infrastructure                                        p. 11

The Basis for a Future Internet?

Optical Hybrid Networks and e-Science as Platforms
for Innovation and Tech Transfer                     p. 15


How the World Changes When You Give the User his Own
Network    p. 17
Treating the GigaPort as Tech Transfer Infrastructure                p. 20
Lambda Grid Developments and the Need for Hybrid
Networking        p. 21   
WiVR: a Window into Virtual Reality                                p. 22
The CineGrid Project                                                p. 23
What These Large Scale Projects Teach Us About User
Populations    p. 24
How Low Can You Go?                                    p. 26
The Global Fiber Infrastructure                                    p. 26
The Significance of SURFnet6 as a Photonic Network                p. 29
A Photonic Switch                                        p. 30
Resource Description Framework Language                     p. 32
Network Description Language as Routing Protocol                p. 35
Terabit Networking                                        p. 38

Interactive Networks

A User Interface for Application Owned
Lightpath Networks   p.41

Running Mathematica                                    p. 47


Where Is All This Headed?   
                      

For Further Investigation - the OptiPuter                             p. 51
It’s the Process and the Planning                            p. 52
The American Approach Needs To Move in the Dutch
Direction        p. 55
Some Thoughts on What is Needed                            p. 56

Appendix:  E-Science Recommendations               p. 59

Executive Summary                                 p. 61