Internet Steering Group Report 13/12/01

Table of Contents

Steering Group Membership

Introduction

Executive Summary

Background

What is Internet2?

Why a Research and Education Network?

The New Zealand Infrastructure's Capabilities

Issues

Capability Study

Steering Group Membership

Members of the Internet2 Steering Group are:

Neil James, University of Otago, Chair
Roger De Salis, Cisco Systems
John Hine, Victoria University of Wellington
Don Hollander
John Houlker, Industry New Zealand
Andy Linton
Simon Riley

Introduction

In August InternetNZ set up a Steering Group on Internet2 to promote some action in the Internet2 arena in New Zealand. The Steering Group decided to run a workshop with invitees from research and education organisations, networking industry suppliers, telecommunication companies, the government sector, and representatives of potential users of advanced Internet technologies. The workshop was held on the 16th of October with the following people involved:

Attending

Apologies

The people attending the workshop concluded that a research and education (R&E) network, able to peer with Internet2 activities, should be established in New Zealand. The immediate action point agreed at the workshop was to develop a paper explaining why it is important for New Zealand to join the Internet2 world, and why the establishment of an R&E network is a requisite part of this. The paper would address issues of value to research and education, and value to the knowledge society in general. It would also look briefly at the application side, including the media industries, and outline the costs and resourcing requirements for the establishment of a network.

This is that paper.

Executive Summary

New Zealand has expressed a desire to join those countries with "knowledge economies". Observation shows that these countries are investing heavily in the infrastructure needed to support the next generation of internet applications. There is no similar initiative in New Zealand. This paper argues that we must start down this path and that a high bandwidth Research & Education network is a logical first step.

As New Zealand's research and tertiary education communities move into a new era with a focus on cooperation, interdisciplinary initiatives and collaboration a first class Research & Education network is essential infrastructure. A properly provisioned R&E network will support research across all disciplines and provide for the future needs of the education sector.

New Zealand's current communication infrastructure has the capacity to support high bandwidth networks amongst the major centres and many provincial centres. Never the less there are issues such as standards, international standards and funding models that need detailed attention. This paper recommends the funding of a capability study to develop accurate requirements, costs, benefits and users of an initial high bandwidth R&E network.

Background

New Zealand wishes to join those countries with "knowledge economies". Toward this end it is currently reshaping its tertiary education sector and its various research funding mechanisms. A common theme is cooperation to establish the synergy that will create an innovative knowledge driven economy. Those countries which we consider to be our peers have uniformly included access to high bandwidth data services as infrastructure essential to support a knowledge economy. A parallel initiative is conspicuously absent in New Zealand.

With the Internet now firmly in mainstream use, various overseas countries are undertaking significant research and providing resources, including very high speed, moderate-cost networking, to various research and education institutions and other interested bodies. In the US the universities have, along with industry partners, developed the Abilene high performance network, and the National Science Foundation (NSF) has established the very high speed backbone network service (vBNS) in partnership with MCI Worldcom. Many other countries have developed or are developing advanced national research and education networks. Examples include Australia's national R&E network, AARnet, and the recently announced GrangeNet project which will take networking to a still higher level. In Singapore the Government has sponsored the development of SingNet to support R&E networking. There are similar developments throughout our region including Japan, Korea, and Thailand. The same pattern has emerged in Europe, and now in Central and South America.

The organisations creating these networks have established peer relationships allowing interconnection of the advanced Internet infrastructure around the world. In the Asia Pacific region the Asia Pacific Advanced Network (APAN) organisation has been established with the objective of fostering research into advanced applications that need or use significant bandwidth. Examples of these applications include distributed grid computing, low-cost interactive video, medical imaging research, and large scale data-set transmission or manipulation.

New Zealand is currently conspicuous by its absence from the group of nations involved in advanced networking and deployment of Internet2 technologies.

This paper argues that a national Research and Education networking body should be established to work with industry to ensure our universities and research institutions have affordable access to enhanced network services that are becoming essential for today's research. This new body would also work with research funding agencies to explore ways in which funding for research infrastructure might be established and would represent New Zealand in international forums on research and education networking. The establishment of this new body would enable New Zealand to take its place, in a peer relationship, in international initiatives such as the US Internet2 project.

The Internet2 Steering Group. The Internet2 Steering Group has been formed to initiate the introduction of Internet2 in New Zealand, enabling New Zealand to retain its place as an advanced country with reasonable levels of infrastructure.

The group does not see itself as a network manager or owner. The group's objectives include:

• Raising national awareness of the benefits of advanced network applications, and New Zealand's need to become a full partner in these developments.

• Marshalling, from a variety of interested parties including network companies, local authorities, local and central government, utilities, educational and research institutions, sufficiently adequate homogeneous resources to present an advanced capability to organisations requiring access to these applications. These bodies would then be able to connect to other institutions and like minded organisations both nationally and overseas via a sufficiently well provisioned connection where there is little or no concern about cost of connection, restriction of transmission medium, or ability to perform scientific or research or educational activities.

Note: Internet2 is the registered name for a United State universities´ project that aims to develop advanced Internet technologies and services. The term "Internet2" is used in this paper as a generic term to represent advanced Internet technologies and services.

What is Internet2?

Internet2 is driven by applications. While a good deal of the research and development work involves various aspects of networking it is all motivated by the desire to provide the next generation of applications.

A large number of applications are already in use using the Internet2 technologies and many development projects including many countries have been established. The Internet2 homepage at http://www.internet2.edu/ can be the start of a journey of exploration of these projects. More specifically some prepared information kits are now available at http://www.internet2.edu/html/infokit.html. The presentations found there feature applications from Health Sciences, Arts & Humanities, and Digital Video applications: http://www.internet2.edu/resources/Health_Sciences.ppt,
http://www.internet2.edu/resources/Arts_and_Humanities.ppt, and
http://www.internet2.edu/resources/Digital_Video.ppt.

For a comprehensive database of the applications being carried out on Internet2 and CA*net 3 (Canada's advanced network) see http://dast.nlanr.net/Clearinghouse/.

The applications described above have a number of common requirements such as high bandwidth, multiple channel synchronisation, naming and location services, security and authentication. These in turn place new demands on network protocols. Development of these applications have motivated research and development in many aspects of quality of service, directory services and public key infrastructure for security.

Why a Research and Education Network?

The New Zealand universities, in collaboration with the DSIR and MAF (and subsequently the CRIs) were instrumental in bringing the Internet to New Zealand in the 1980s and played a major role in encouraging Telecom NZ to introduce the first of the faster technologies for data networking, Frame Relay.

While in other countries organisations with a focus on R&E networks survived the Internet's development as mainstream commercial infrastructure this has not happened in New Zealand. The absence of such an organisation is reflected today in the lack of an information infrastructure focused on research and education. As New Zealand adjusts its research model introducing Centres of Research Excellence (which may in fact be distributed) and Collaborative Research Centres the lack of infrastructure becomes glaringly apparent.

To ensure that our research workers have opportunities to participate in national and international collaboration, it is vital that they have access to advanced networking technologies as represented by the emerging Internet2 world. Participation of the New Zealand universities and research organisations in the Internet2 initiative outlined in this paper is designed to ensure that New Zealand is able to participate in the high speed network initiatives already implemented or about to be implemented around the world.

A principal use of very high-speed networking is the provision of world-class facilities for education at all levels. Just one of the facilities that becomes available with Internet2 technology is fast and reliable video transmission. It is hoped that all readers can imagine a video-conference of such clarity and quality that it is possible to avoid significant hours and resources spent traveling to face-to-face classes or meetings. The current level of video-conferencing quality leaves an enormous amount to be desired. Unfortunately New Zealand's communication infrastructure is inadequate in many rural and provincial regions to support appropriate learning support technologies. There are several good local initiatives that aim to address this issue but without the support of a national initiative providing access to Internet2 technologies, a lot of the potential will not be able to be realised.

Internet2 is an opportunity to provide distant communications on a scale and quality that permits easy communication, without a terribly onerous cost penalty. This has an extremely positive potential with respect to the use of scarce resources (specialist teachers) and the current provision of physical travel infrastructure (example, roads, airlines etc).

The New Zealand Infrastructure's Capabilities

The infrastructure required to support Internet2 development consists of regional or national high speed networks. Bandwidth capabilities currently run from 1 to 10 gigabits per second. Different national networks, universities, research institutions and collaborating corporates connect at points named "GIGAPOPS" (gigabit points of presence). To initiate the use and development of Internet2 applications in New Zealand a high speed backbone network offering initial connections to similar organisations must be put in place. We argue that this can be accomplished today at modest cost.

National Data Transmission Capacity. Our national carriers normally transmit 2.5 Gigabit/sec, over a single fibre, or more likely over a single wavelength of a fibre. Current mainstream equipment permits the transmission of 64, 128 or 256 wavelengths over a single fibre. Fibre optic cables are typically laid containing 24 or 48 fibres.

Therefore, the provision of a single 24 or 48 core cable permits the current practical transmission capacity of 48 x 128 x 10 Gbps = 62,000 Gigabits per second. New Zealand's main centres are currently well covered by several such cables.

New technology, currently moving from research laboratories to production would potentially increase this to 40 gigabits/sec per wavelength, 256 wavelengths per fibre, for a cable capacity of nearly 500,000 gigabits per second.

Interfaces and Switching. The lowest speed connection currently offered by mainstream computer and network equipment vendors is 100 Mbps. The current sweet spot (i.e. that point offering optimal price/performance) is 1 Gigabit/sec Ethernet over copper. All these interfaces have fibre-optic options for modest cost increments. High speed interfaces of 2.5 Gbps and 10 Gbps are easily available.

Gigabit interfaces are now so inexpensive that they can be reasonably afforded by very small business’s or research laboratories, or even domestic homes, should the applications be of sufficient merit to justify the expenditure.

Issues

While the basic capacity and technology can be put in place for modest cost there remain other issues that must be addressed in the establishment of a high speed national R&D network. In this section we address several of these issues.

Technical Standards

In order to provision a single unified network of significantly greater capacity and capability than currently available, an initial base set of technical standards was suggested at the meeting. These are designed to ensure that organisations connecting to this network exhibit some basic technical comprehension and are capable of gracefully cohabiting on a shared resource. These standards are:

International Peering Point

The steering group has only considered Auckland as the logical place for international connections, due to current international fibre optical cable terminations.

In order to establish connection with similar overseas networks, at lest one international peering point will also need be established. AARNet, the Australian R&E network, has established peering points at the Pacific NorthWest GigaPoP. It is likely that New Zealand should take a similar path, though moves to develop an Australasian infrastructure must also be watched.

An Initial National R&E Network

New Zealand's advanced R&E network will initially be located to connect all the major academic and research institutions in New Zealand. This includes connections between Albany, Auckland, Manakau, Hamilton, Palmerston North, Wellington, Christchurch and Dunedin. The group agreed that capability must exist to connect the next level of institutions (polytechnics, colleges etc) and a longer term goal of the ultimately connecting every school and library in the country to this valuable educational research tool.

A key issue for the initial development is funding. New Zealand's current research funding mechanisms do not recognise the importance of infrastructure in supporting research. In the case of a R&E network the infrastructure will be critical to collaborative research, the development of new applications and capabilities across the research portfolios and as a platform for ICT research.

Capability Study

The meeting recommended that the next step be to obtain sufficient resources to provide a capability study. The tasks identified by the meeting for inclusion in the capability study include (but are not limited to):

• Existing Resource Availability. While we are confident that sufficient fibre optic cable is in place there is a need to identify one or more partners prepared to make capacity available and identify the locations around the country in which that capacity would be available.

• Financial Models. Different models for funding and managing the resources needed to implement and maintain a R&E network must be investigated. There are questions of ownership, agreed use, contracts with carriers, etc.

• Potential Users. Internet2 has been motivated by the next generation of applications. Users of these applications must be identified within New Zealand. Likely candidates are collaborative research groups, and the entertainment industry.

• Overseas Partners. New Zealand does not have the capacity to support the potential range of applications on its own. We need to identify like minded groups, government organisations, inter-government groups engaged in similar endeavours.

The final part of the capability study is to assemble all the knowledge gained into a conclusive report that will allow organisations and individuals to contribute successfully to a national resource.

©2002 The Internet Society of New Zealand Last updated 30 April 2002