Director of the Telecommunication Standardization Bureau (TSB), Malcolm Johnson, has presented a report to the ITU Council for its October 2009 meeting. The report analyzes the possibility of the ITU taking on a role as an IPv6 address registry for Member States and concludes by suggesting that the ITU take action toward this goal.
This proposal is based on resolution 64 of the WTSA-08 (Johannesburg 2008) which directed the TSB Director to “(…) study the question of IPv6 address allocation and registration for interested members and, especially, developing countries and to report to the 2009 session of the ITU Council”.
The Number Resource Organization (NRO) appreciates concerns held by so many stakeholders regarding the uptake of IPv6, particularly now, as the IPv4 unallocated address distribution regime draws to its logical conclusion. The NRO appreciates, and shares, the concerns of the ITU-T in ensuring that the global communications system is a common global asset that can be used equitably and fairly by all. It is vital, however, that the global community recognize that the Internet is a global resource that already carries a massive level of investment in the provision of goods and services, and that, as we move forward, we must be mindful of all aspects of continuity and stability.
The NRO believes that the TSB Director’s report goes far beyond the terms of reference laid down in Resolution 64. Rather than simply undertaking the requested studies, the report assumes that the objective of those studies is to turn the ITU into an IPv6 registry, a position not agreed on by the Member States at the WTSA-08.
The report of the TSB Director to the ITU Council is partially based on a study conducted by the NAv6 program of Malaysia. This study was widely criticized by various stakeholders, including governments, for its lack of scientific rigor and for being tailored around pre-determined conclusions. It is clear from the study’s findings that the objective was not to provide an impartial analysis of the challenges currently facing the Internet, but rather to advocate a parallel structure to the RIRs. This is made explicit in the Executive Summary:
“This document has been created to propose and review the creation of an additional parallel structure to the RIRs for the allocation and distribution of IPv6 addresses to the global community as requested by some member nations to the ITU.”
The NRO believes that establishing an IPv6 registry under the ITU, responsible for allocating Internet addresses to Member States, has the potential to create significant technical problems for the global Internet and is contrary to the established principles for Internet Governance which emerged from the Geneva (2003) and Tunisia (2005) summits, constituting a backward step from the important advances that have been achieved regarding the formation of multi-stakeholder models of Internet Governance. At the same time, the establishment of such a registry will not achieve the stated objectives of the plan’s proponents.
These arguments are explained more fully in the following paragraphs.
1. Objectives
Some objectives in establishing an ITU-controlled IPv6 administration mechanism include:
1.1 Assuring the availability of a certain number of IPv6 addresses for the use of each Member State
There is no practical way to limit the use of IP addresses to a certain country. The topology of the Internet is different from the topology of classic telephony.
IP networks do not have topologies based on national boundaries. In very large countries, it is possible that the networks of some operators are contained within the borders of the country, but in much of the world it is normal to have networks that cover several countries. In such cases there is no way to limit the use of specific IP addresses to within a certain country.
Further, if an addressing scheme based on countries were to be deployed, network operators would certainly obtain IP addresses in those countries where it is easiest or cheapest. Addresses supposedly allocated to one country may then be used in transnational networks, defeating the purpose of nation-based allocation.
1.2 Ensuring the equitable distribution of IPv6 addresses
The TSB report proposes that the ITU develop rules for the allocation of IP addresses to Member States. These rules would presumably replace the existing allocation criteria used by the RIRs, which determine eligibility based on demonstrated need for addresses, in favour of pre-allocating of IPv6 addresses directly to ITU Member States based on parameters such as population or size of their economy (GDP).
In such discussions, the most populated countries would surely argue that population be the basis for IP address allocation, while the larger economies would certainly argue for GDP as the main criterion. The NRO notes that neither criterion is usually associated with developing countries.
1.3 Avoiding repetition of the errors made in the early stages of IPv4 distribution of IPv4
It is important to highlight that the situation we face with IPv6 today is absolutely different to the situation with IPv4 in the early days of the Internet.
In the early days of the Internet, allocations of IPv4 were made to organizations in those countries that were first connected to Internet. These allocations, which far exceeded the needs of those receiving the addresses, were made mainly in the United States.
The creation of the Regional Internet Registry (RIR) system fundamentally changed this practice, moving to a system in which allocations were made based on need, demonstrated and documented by the applicant, in accordance with defined policies that are developed in an open, participative and transparent way. Internet Governance discussions in recent years have recognized this as a successful, responsible and inclusive system.
The uneven geographic distribution of IPv4 address space, produced before the creation of RIRs, is unlikely to have any significance once IPv4 addresses have run out, as the entire world will need to adopt IPv6 at roughly the same time. In addition, it is likely that the RIRs which distribute to the least developed regions (LACNIC and AFRINIC) will have remaining IPv4 address stocks for longer than the other RIRs, due to recovery and redistribution of unused addresses and equal distribution of the last part of the central stock. This will mean that LACNIC and AFRINIC will be able to continue making allocations of IPv4 in their regions, cushioning the potential impact of IPv6 transition on their communities.
In September 2006, the global Internet community approved a policy allocating each RIR a “/12” block of IPv6 addresses to be distributed to their respective communities. “/12” represents a huge number of Internet addresses – based on standard assignment practices, “/12” represents addresses for 4,500 trillion End Users (or networked devices). Each RIR has been allocated the same number of IPv6 addresses, ensuring that network operators in all countries already have access to the addresses required for their networks.
2. Route Aggregation
One of the principles guiding the RIRs is the route aggregation. This means assigning IP addresses in a manner that they can be announced to the Internet split into the smallest number of separate blocks. It is these announcements of separate blocks of reachable IP addresses that comprise the Internet routing table, which is used by all the routers throughout the Internet.
The more announcements there are, the larger the routing table. An expanding routing table necessitates more frequent updates of core Internet routing equipment, larger memory and more processing capacity. Route aggregation aims to minimize the Internet routing table and thereby minimize the load on the Internet’s core infrastructure.
The establishment of a parallel system to the RIRs would have a significant negative effect on route aggregation, resulting in the need for larger and more frequent investments in routing infrastructure. The negative impact of this would most likely be felt most keenly by developing countries.
3. Resource Conservation
Resource conservation is another core principle of the RIR system. Although the number of IPv6 addresses is almost unimaginably large, it is still important to acknowledge that this is a finite resource. It must be administered responsibly, using conservation criteria, in order that the Internet may continue to evolve, even in dimensions not foreseen today.
The creation of a parallel IPv6 administration system, competing with the existing system in terms of price or policy, would likely increase address consumption.
4. Regression of the Internet Governance Systems
In recent years, and especially since 2003, when the first phase of the WSIS was developed in Geneva, the systems of Internet Governance have evolved significantly, consolidating a global, multi-stakeholder system that would have been difficult to imagine only 10 years ago. As time has gone on, the participation of all stakeholders in these governance mechanisms has improved significantly.
The RIRs participation in traditional governmental proceedings, such as APECTEL, CITEL, eLAC, OECD; the permanent interaction of the governments with the RIRs, the Government Roundtables organized by RIPE NCC; the permanent reports issued by the NRO to the Governmental Advisory Committee of ICANN (GAC); the creation of government working groups in several RIRs, the open and frequent communication with Law Enforcement agencies; the growing participation of government representatives in the RIRs meetings, and the permanent open dialogue, are just some examples of the relevant progress that has been made in the consolidation of this new vision of the relationships between various stakeholders.
The creation of an IPv6 registry under the ITU, with rules defined by Member States and focused on assigning IPv6 addresses to Member States, would constitute a significant regression in this progress and against the principles laid out in the Tunis Summit for Internet Governance. Creating a non multi-stakeholder governance mechanism will also diminish the motivation for governments to participate in existing governance structures.
Conclusion
The current IP address administration system is based on open, participative, transparent and multi-stakeholder policy making processes that reflect the needs of network operators. It constitutes a successful governance model that appropriately channels the concerns and expectations of all stakeholders, including governments.
If ITU Members States have unsatisfied expectations or interests, the existing system offers guarantees and opportunities to address those concerns. The NRO has offered to work together with the TSB Director in addressing any possible concerns from the Member States, and we reaffirm that offer here.
The project of creating an IPv6 registry, parallel to the existing RIR system, located in the ITU with the objective of allocating IPv6 addresses to Member States, does not offer any improvement over the current system. It does, however, present huge doubts, risks and negative aspects, especially for developing countries. Such a system will not make the Internet more affordable for citizens of countries in the developing world; it will not extend the reach of users from local to global. Rather, it will create additional costs and impediments, and act as a barrier to effective competition and innovation.
Last modified on 14/09/2018