Today's Internet has some 1.7 billion users, fosters an estimated $1.5 trillion in annual global economic benefits, and is widely agreed to offer a staggering array of societal benefits. The network sees enormous demand---on the order of 40 Tbps of inter-domain traffic and an annual growth rate of 44.5%. Remarkably, in spite of the Internet's importance and rapid growth, the core protocols that support its basic functions (i.e., addressing, naming, routing) have seen little fundamental change over time.

However, the Internet is now in the midst of its first fundamentally disruptive modification: a phase transition imposing extensive changes throughout the network's vast set of components. On the 3rd of February, 2011 the Internet Assigned Numbers Authority (IANA) allocated the five remaining blocks of IPv4 address space to the five Regional Internet Registries (RIRs). This event was a watershed moment for the Internet, in that it represented the exhaustion of the Internet's pool of unallocated IPv4 addresses. While eliminating poor utilization of existing allocated IPv4 address space and application of techniques for IPv4 address re-use (i.e., Network to Address Translation (NAT) and Dynamic Host Configuration Protocol (DHCP)) will likely provide short term relief, the eventual scarcity of this critical Internet resource will be a strong catalyst for change.

This project seeks to validate the hypothesis that the scarcity of IPv4 addresses and accompanying solutions will have profound effects on the Internet and many of its vital properties, including heterogeneity, openness, security, scalability, reliability, availability, concurrency and transparency. This project's efforts to understand these impacts are divided into three broad areas: (i) studies of the dynamics by which IPv4 addresses are allocated and how these address resources are subsequently used, (ii) understanding the near-term coping mechanisms and transition technologies (e.g., carrier grade NAT, tunneling), and (iii) observing the adoption of longer-term solutions (i.e., the new version of the Internet Protocol, IPv6). To pursue the above goals this project has formed a cross-organizational team of researchers with a long history of scientific Internet measurement (University of Michigan, International Computer Science Institute). The project is bolstered by a substantial supporting team that includes ISPs (AT&T, Merit Networks, Inc.), Internet number resource registries (APNIC, ARIN, AFRINIC, RIPE NCC, and LACNIC), and infrastructure providers (Verisign, Packet Clearing House, Arbor Networks) who will provide access to massive, often unique data.

Intellectual Merit: Such measurement---at-scale and across numerous core Internet protocols---presents not only a technical challenge in itself, but offers a once-in-a-generation opportunity to view massive, distributed systems under scarcity and transition. The behaviors observed will inform not only the development of global distributed systems, but also the sciences of network architecture and design, network protocols, and mobile networks. Specific areas of study in these domains potentially include: address transfers, address reclamation, address pollution, address allocation policies and allocation policy abuse, new address space topologies, addresses and identity, dynamic addressing, addressing and routing, addressing and mobility, transition strategies, adoption incentives, adoption tipping points, balkanization in transition, adoption measurement, negative adoption incentives.

Broader Impact: For many the Internet has become a ubiquitous aspect of everyday life. A major impact of this work will be directly ensuring the availability and reliability of the Internet through instrumentation, data sharing, and problem diagnosis with operators and networks throughout the global network as it undergoes this singular phase in its functioning. Further, in the spirit of scientific exploration commensurate with this unique phenomenon and measurement, this project will make available the extensive data collected as a means to spur scientific research beyond the specific research of the project team. Finally, based on the observations and studies, this project will develop principles guiding the development of new, more resilient network architectures and protocols.

Agency
National Science Foundation (NSF)
Institute
Division of Computer and Network Systems (CNS)
Application #
1111699
Program Officer
Darleen L. Fisher
Project Start
Project End
Budget Start
2011-08-01
Budget End
2015-07-31
Support Year
Fiscal Year
2011
Total Cost
$1,199,901
Indirect Cost
Name
University of Michigan Ann Arbor
Department
Type
DUNS #
City
Ann Arbor
State
MI
Country
United States
Zip Code
48109