Current | 2004 | 2003 | 2002

7 October 2002

Collaborating at Speed

"Access to next generation internet is crucial to fulfilling the government's aims of encouraging world-class innovation and strengthening global connections" - Hon. J Anderton, Deputy Prime Minister, June 2002.

Executive Summary

An Innovative New Zealand

This capability study was commissioned by InternetNZ with support from Industry New Zealand, The University of Otago and Cisco Systems, to develop a plan for meeting the challenge presented by the Deputy Prime Minister - how do we, as a country, ensure that researchers, educators and business innovators have access to the next generation internet (NGI).

It has already been observed in international NGI forums that New Zealand is conspicuous by its absence. New Zealand remains one of the few countries in the developed world that has failed to invest in high-speed internet infrastructure.This is a rather strange position to be in. In recent years there has been so much talk about the knowledge economy and its importance for New Zealand's future economic growth, it is rather alarming to discover that much of the talk hasn't been transformed into action. The Government can claim with confidence that it has provided direction with the release of its policy framework for economic transformation - Growing an Innovative New Zealand - in February 2002. The responsibility now rests with individual sectors and the leaders in these sectors to translate policy into action.

One of the cornerstones of the Government's innovation strategy is to "increase our global connectedness"; another was to encourage greater collaboration within the research and education sector and with industry. This NGI study suggests a way that these goals can be achieved.

Applications Development

The Next Generation Internet enables us to communicate in ways that are not possible with today's Internet. With "bandwidth to waste", as one participant put it, we can start to explore new ways of collaborating, and new ways of sharing high-performance networking applications. This includes e-Science, Access Grids, Digital Video, Bioscience, E-Learning, Networked Virtual reality, Digital Libraries and Tele-Health, to name a few.

The development of new applications is at the heart of NGI networks. NGI networks certainly deliver higher bandwidth connections than are commonly in use today, but without new applications, they offer little advantage.New Zealand's specific requirement is for high-speed gigabit connections linking organisations engaged in collaborative research and education initiatives, as well as for the emerging new knowledge industries, especially those in the film and television post-production sector. These links should initially extend from Dunedin to Auckland, with early extensions to other regions. User access would start at 1Gbit/s with access speeds up to 40Gbit/s expected within 2-3 years.

Government Priorities

The Government's priority economic growth sectors - creative industries, biotechnology and ICT - would all receive a boost from the establishment of an NGI. Other Government research initiatives, including the Centres of Research Excellence (CoRE) and the Research Consortia, both of which rely heavily on collaborative activity, will quickly find the absence of high-speed internet capability a major impediment. Likewise, the PROBE initiative that aims to deliver high-speed bandwidth to all schools and most communities within two years would benefit from NGI backbone connectivity.

Open Access

A key finding from this study is the need for any sector with a requirement for high-performance internet connectivity to have access to the NGI. Many other countries have been able to justify establishing dedicated NGIs for the research and education sector, largely on account of their size. Usage of the network is governed by Acceptable Usage Policies (AUPs) to exclude any commercial (or commodity) traffic. This approach is not supported in New Zealand. The demand for bandwidth in the digital media and post-production sector already exceeds that required for research and education. Interconnecting with global NGIs will require some separation of commercial and non-commercial traffic at the international border, but this need not limit access within New Zealand.

The establishment of the CONZULSys Consortium, involving four Universities, and set up with the aim of acquiring a common shared library management system, is an excellent example of new "innovation economy" thinking. CONZULSys not only is a potential large user of a national NGI, but also provides a business model for collaborative activity. The NGI and CONZULSys initiatives should proceed in close consultation.Research FundingGovernment research funding through MoRST and its funding agencies - FRST, RSNZ and HRC - is being re-focused to more closely match the "innovation economy" priorities.

Access to a global NGI will help to ensure that New Zealand- based researchers are not disadvantaged. The support of the research funding agencies will be critical in exploiting the new opportunity.

Technology initiatives in the past have at times failed because all the funds are spent on the technology itself and there is no support for the applications. Every dollar invested in NGI infrastructure should be matched at least dollar for dollar with funding for research and applications development. Perhaps one of the most exciting aspects for New Zealander researchers will be the opportunity to attract research investment from other countries.

Broadband Coordination

The time is right to proceed with the implementation of an NGI. The opportunities presented by broadband networks have been known by technologists for over 15 years (and by science fiction writers for over 30 years), but most businesses and members of the general public have had little conception of what a broadband future might be like. This is changing rapidly. Helped along by initiatives such as Digital Trade Routes (high speed links to North America and Singapore) and PROBE (Provincial Broadband Extension), there is growing awareness of what is possible and in an increasing number of cases, frustration with today's limited and costly services.

New Zealand needs to be careful that our tiny market is not fragmented by narrow sector interests. There is a need for ongoing consultation and cooperation amongst these initiatives.

Infrastructure Development

The review of international developments is quite depressing, highlighting just how far New Zealand has slipped behind. On the positive side, we can learn from their experiences - Australia, Canada and the USA in particular have well developed NGI networks. We can also get some direction from the business models that these countries are using to manage and govern their networks.

In New Zealand, as in other countries, existing providers of telecommunication services typically have high capacity fibre infrastructure supporting their backbone networks. It is perhaps not surprising that all providers expressed a keen interest in participating in a national NGI. Even those with only local infrastructure (in Auckland and Wellington) felt they could assist with tail circuits to individual user sites.

The national telecommunication providers have recognised the provision of capacity for a national NGI is not so much an issue of capability but rather about compatible commercial agreements. National providers are very happy to support initiatives that will help grow new applications and new traffic, but are naturally very concerned about the potential threat to existing revenues from applications that provide NGI users with a bypass opportunity. For example, in Australia, AARNet is now a fully licenced carrier and is able to offer its user organisations significant advantages by treating voice and data as one, with appropriate quality measures.

The NGI committee will insist upon a transport model similar to that deployed by AARNET, which actively encourages a wide range of activities.AARNet has achieved its flexibility by seeking other suppliers of backbone infrastructure to avoid the business conflict of interest in reselling services from the traditional telecommunication providers. In Australia, Canada and the USA, this has been an option because of the healthy growth in competition for the supply of backbone wholesale bandwidth services.

New Zealand does not have this option today. Wholesale services are available from metropolitan area network providers (CityLink and Unitednetworks) and from international provider, Southern Cross Cable Network (SCCN), but they are not currently available between cities. This limits the options for a national NGI to a business partnership with an existing provider or the establishment of new infrastructure, which would probably need to be in partnership with an organisation with ownership of a national right of way (Transpower or Tranz Rail).

The Steering Group recommends that both options be explored in more detail when funding is assured.New Zealand is in a privileged position when it comes to international connectivity. With the Southern Cross Cable ring linking Australia, New Zealand, Fiji and Hawaii to the western sea border of the United States (Los Angeles and Seattle), New Zealand has access to as much capacity as we can afford (and then some!).

For start-up networks the cost of the 15-year Indefeasible Rights of Use (IRUs) can seem high, but not only does Southern Cross offer a range of capacity and pricing packages, AARNet has also opened the door to a collaborative arrangement which would mean that New Zealand could share their capacity. At face value this seems to be a very attractive offer, but of course the detail would need to be negotiated. Access to the global NGIs would then be provided via Pacific Wave in Seattle and the US Northwest GigaPop.GovernanceA New Zealand NGI requires clear governance and management structures.

Previous co-operative models used to establish the original Internet in new Zealand (Tuia Society) are no longer appropriate. Best effort services are no longer acceptable. Users of the Internet and its derivatives now expect 24 x 7 fully supported commercial services. The same would be expected of an NGI (except of course for experimental use of the network).

We are recommending 2 x 1Gbit/s access services so that even under experimental conditions, one link can remain in normal service.Resale models like ComNet have also proved difficult to sustain; unless the value-added service offered by the reseller is extremely compelling (price discounts are generally not enough, as these are easily matched by the infrastructure provider), the reseller can find it extremely difficult to compete with its own supplier.

The Governance models selected by UCAID (University Corporation for Advanced Internet Development) in the United States and FING (French Next Generation Internet Foundation) have special relevance to New Zealand. By assigning start-up responsibilities to a specific group like the University Vice-Chancellors, the establishment of an NGI could proceed relatively quickly. However, New Zealand must have a broader representation to ensure the needs the new knowledge industries are met as well as the research and education sector.

Funding

Now the important question. Who is going to pay? In support of its innovation framework, and its desire to promote better co-ordination between research, education and industry, the Government has established a number of funding pools with special incentives for collaboration.

Most of these relate to research activity, which is helpful once an NGI is established. However, the recently announced Partnerships for Excellence fund does provide reasonable sums of capital ($5 million upwards) for large projects. A key condition is that matching funding be obtained from private investors.

The Government has already given approval in principle to provide $25 million towards the development of a new Business School at Auckland University, subject to a satisfactory business case and of course the matching funding being raised. This precedent augers well for an NGI proposal.

But funding is not only required to build the network (whether this involves new infrastructure or infrastructure leased from existing providers). Ongoing funding is required to manage the network and support applications research. Our view that the most sustainable option is for users to share in these operating costs. Government and industry can of course support specific research initiatives with both cash and in-kind equipment and services.

A Way Forward

What is the way forward? It is now one-year since InternetNZ established a Working Group to promote the establishment of an NGI. It is time for action. We must commit to having the network operational within the next 12 months.

The first priority is to establish an NGI Consortium, with partners who are willing to make a direct financial contribution. This should be established no later than 1 November 2002. With contributions of $15,000 from each partner, as well as $100,000 from the Tertiary Education Commission's "Collaboration for Efficiency" Fund and a further $100,000 from Industry New Zealand's "Sector Initiatives Fund", there would be sufficient funding to meet the costs of the establishment management and governance team for the first twelve months.

A Project Director should be appointed by the end of 2002 and a Business Case presented to Government and Partners by March 2003. Subject to Government approval contracts could be let in May 2003, with the network operational in September 2003.

This is a challenging timetable, but the Steering Committee believes it is achievable.It is no longer a question of whether New Zealand needs an NGI, the challenge now is to see how quickly we can establish one.