The Internet is running out of unique identifiers. Consumption of these identifiers continues at a rate that requires near-term attention to issues in not only the technical but also economic and socio-political domains.
We wish to support informed public policy through our objective analysis of questions concerning address space exhaustion, concentration of address ownership, and IPv6 adoption. We hope to foster discussion that recognizes the following empirical conditions:
- the current Internet has become critical infrastructure for governments, organizations, and individuals throughout the world,
- the Internet must become more scalable and sustainable, which requires an infusion of capital and skilled labor, and
- the organizations that own and maintain core Internet backbone infrastructure currently have neither incentive nor spare capital to invest in fundamental architectural innovation of the Internet.
Research topics are:
We analyze the current controversy regarding predictions of IPv4 address space exhaustion and provide an independent opinion regarding possible future trajectories based on the best publicly available data. We also analyze the related question of concentration of address ownership.
In October 2005, CAIDA's analysis of IPv4 consumption rates (sponsored by ARIN and presented at the October 2005 ARIN meeting) used an exponential model that predicted that, at 2005 IPv4 address allocation rates, IANA would allocate all unused IPv4 space by 2008, with exhaustion of the additional multicast and special-use space following in late 2008 and early 2009.
Several analyses of IPv4 address consumption rates and lifetime projections exist. Most notable are Geoff Huston's "IPv4 Address Space Report"  and Tony Hain's "A Pragmatic Report on IPv4 Address Space Consumption" . These analyses predict that, if current allocation rates prevail, IANA will exhaust all available space sometime between 2009 and 2016.
With respect to the ownership of address space, the clear trend is toward an increasing concentration of ownership of the address space in a relatively small number of organizations. Between 2% and 11.2% of organizations hold between 56.4% and 63.4% of the address space. We also find that the vast majority of organizations hold fewer than 8 allocations each.
We analyze the IPv6 consumption rate by examining the top-level allocations made by IANA to RIRs.
Although IPv6 allocations have occurred sporadically over the past 6 years, an overwhelming majority of the allocations, in terms of address space coverage, have been made in the last two years. This recent acceleration in growth suggests increasing interest in IPv6. RIPE and APNIC hold the majority of the allocations and account for 91.2% of the total allocated IPv6 space. ARIN has a comparatively small amount at 8.2% of allocated IPv6 space, while LACNIC has only negligible IPv6 allocations.
Though currently constrained by available funding and resources in this area, CAIDA would like to maintain and eventually automate graphs of IPv4 concentration. These graphs and corresponding analysis of allocated, routed, and occupied address space help to begin to answer the question, what drives the demand for IPv4 address space?
Detailed analysis of IPv4 address space concentration.
IPv4 Census Map
The IPv4 Census Map displays IPv4 prefixes that were inferred to contain at least one active host from both passive and active measurements. Thus, the data and map give us a lower bound on IPv4 address utilization.
IPv4 WHOIS Map
A visualization of the characteristics of IPv4 "whois" data from the five Regional Internet Registries (RIRs). Each pixel in the full-size image represents a single /24 network containing up to 256 hosts. The color of each pixel corresponds to where the address is routed used, unrouted assigned, unrouted unsed, unrouted unassigned, or reserved.
The projected rates of IPv4 address consumption make compelling arguments for expediting scenario planning and discussions of economics, ownership and trust issues. CAIDA emphasizes that technical analysis alone cannot effectively drive these discussions. In the following presentations, CAIDA Director and PI, kc claffy, describes how the particular problem of address consumption is one of many operational Internet problems that has resisted a satisfactory technical solution for over a decade:
- "apocalypse then": ipv4 address space depletion (ARIN, Oct '05)
- [Ii]nternet [applications] drivers of growth: 2005-2015 (OIDA, Jun '05)
- top problems of the Internet and what can be done to help (AUSCERT, '05)
- Additional presentations on policy