TOP
Search the Dagstuhl Website
Looking for information on the websites of the individual seminars? - Then please:
Not found what you are looking for? - Some of our services have separate websites, each with its own search option. Please check the following list:
Schloss Dagstuhl - LZI - Logo
Schloss Dagstuhl Services
Seminars
Within this website:
External resources:
  • DOOR (for registering your stay at Dagstuhl)
  • DOSA (for proposing future Dagstuhl Seminars or Dagstuhl Perspectives Workshops)
Publishing
Within this website:
External resources:
dblp
Within this website:
External resources:
  • the dblp Computer Science Bibliography


Dagstuhl Seminar 13322

The Critical Internet Infrastructure

( Aug 04 – Aug 09, 2013 )

(Click in the middle of the image to enlarge)

Permalink
Please use the following short url to reference this page: https://www.dagstuhl.de/13322

Organizers

Coordinator

Contact



Schedule

Motivation and Goals

The Internet was designed to offer open data transfer services on a planetary scale. However, its success has turned it into a mission-critical infrastructure of vital importance for most countries, businesses, and industries. The aim of this seminar is to bring together the research and network operator communities to discuss and analyze the Internet as a critical infrastructure. We will address the vulnerability of the Internet from a number of different angles (e.g., physical infrastructure, control plane, data plane, services, etc.), with an emphasis on the core transport infrastructure as well as the content delivery side. The seminar will contribute to a better understanding of the Internet as a system of interdependent elements and pursue extensions of current research perspectives to consider novel (and maybe unusual) approaches to studying the local or region-specific substrates as parts of the Internet's global ecosystem.

Rethinking Perspectives on the Internet Backbone
Analyzing the mutual impact between ASes, the vulnerability and efficiency of the backbone requires the identification of ASes and their role in mutual transit. In particular, stakeholders do not want to ground their internal data exchange on weak third parties. In Internet terms, AS interconnections between key players of a country should be part of a transparently visible Internet ecosystem. However, the Internet is a globally distributed network without boundaries, which makes the identification of locally relevant subparts hard. This seminar aims at being a platform to leverage new and uncommon research perspectives that go beyond the Internet backbone as a globally distributed system.

Methodologies to Analyze the Internet Structure
To analyze the Internet as a critical infrastructure, a clear picture is required about the kind and granularity of data needed to obtain relevant results and draw valid conclusions, even if the available dataset is restricted. Sampling and inference are common methods to assess the impact of the limited view on the real Internet. Current approaches to model the Internet backbone need to be revisited to reflect the Internet as critical infrastructure. The mapping of logical Internet nodes (ASes) to concrete entities (companies, points of presence etc.) as well as its annotation with meta data (e.g., administrative contact points) have been identified as important to cover the Internet structure from a non-technical perspective.

Paradigms Overlaying IP Connectivity
Delivering content to the end users is one of the main objectives of the Internet. In the early Internet, end users accessed content directly from a primary source. With the advent of CDNs this has changed. A single CDN operates as replication and distribution network for many content publishers, which brings data closer and more efficiently to end users. In fact, a very large portion of the current Internet content is maintained by only a limited number of CDNs, creating a limited competition in this area. Until now, this oligarchy has not been thoroughly studied, especially in the context of the Internet as a critical infrastructure.


Goals of the Seminar:

The research questions to be pursued and answered include:

  • How can we define and extract a locally relevant view of the globally distributed Internet?
  • Which metrics are appropriate to measure the importance of Internet stakeholders and their mutual relationships?
  • Which countermeasures and improvements are feasible to protect the Internet as critical infrastructure without narrowing its flexibility and openness?
  • To what extent can we analyze the Internet structure in short time frames?
  • What is the role of specific ASes for reliably interconnecting the Internet infrastructure of a country?
  • How can we reveal weak transits and unintentionally strong dependencies between ASes and specific regions of the world?
  • How can we predict Internet scale consequences of large scale problems (what-if-questions)?

Summary

The Internet was designed to offer open data transfer services on a planetary scale. However, its success has turned it into a mission-critical infrastructure of vital importance for most countries, businesses, and industries. The aim of this seminar is to bring together the research and network operator communities to discuss and analyze the Internet as a critical infrastructure. We will address the vulnerability of the Internet from a number of different angles (e.g., physical infrastructure, control plane, data plane, services, etc.), with an emphasis on the core transport infrastructure as well as the content delivery side. The seminar will contribute to a better understanding of the Internet as a system of interdependent elements and pursue extensions of current research perspectives to consider novel (and maybe unusual) approaches to studying the local or region-specific substrates as parts of the Internet's global ecosystem.

Rethinking Perspectives on the Internet Backbone

Analyzing the mutual impact between ASes, the vulnerability and efficiency of the backbone requires the identification of ASes and their role in mutual transit. In particular, stakeholders do not want to ground their internal data exchange on weak third parties. In Internet terms, AS interconnections between key players of a country should be part of a transparently visible Internet ecosystem. However, the Internet is a globally distributed network without boundaries, which makes the identification of locally relevant subparts hard. This seminar aims at being a platform to leverage new and uncommon research perspectives that go beyond the Internet backbone as a globally distributed system.

Methodologies to Analyze the Internet Structure

To analyze the Internet as a critical infrastructure, a clear picture is required about the kind and granularity of data needed to obtain relevant results and draw valid conclusions, even if the available dataset is restricted. Sampling and inference are common methods to assess the impact of the limited view on the real Internet. Current approaches to model the Internet backbone need to be revisited to reflect the Internet as critical infrastructure. The mapping of logical Internet nodes (ASes) to concrete entities (companies, points of presence etc.) as well as its annotation with meta data (e.g., administrative contact points) have been identified as important to cover the Internet structure from a non-technical perspective.

Paradigms Overlaying IP Connectivity

Delivering content to the end users is one of the main objectives of the Internet. In the early Internet, end users accessed content directly from a primary source. With the advent of CDNs this has changed. A single CDN operates as replication and distribution network for many content publishers, which brings data closer and more efficiently to end users. In fact, a very large portion of the current Internet content is maintained by only a limited number of CDNs, creating a limited competition in this area. Until now, this oligarchy has not been thoroughly studied, especially in the context of the Internet as a critical infrastructure.

Original Goals of the Seminar

The research questions to be pursued and answered include:

  • How can we define and extract a locally relevant view of the globally distributed Internet?
  • Which metrics are appropriate to measure the importance of Internet stakeholders and their mutual relationships?
  • Which countermeasures and improvements are feasible to protect the Internet as critical infrastructure without narrowing its flexibility and openness?
  • To what extent can we analyze the Internet structure in short time frames?
  • What is the role of specific ASes for reliably interconnecting the Internet infrastructure of a country?
  • How can we reveal weak transits and unintentionally strong dependencies between ASes and specific regions of the world?
  • How can we predict Internet scale consequences of large scale problems (what-if-questions)?

The complexity of the Internet makes it equally complex to give complete answers to these questions. This seminar helped us to start touching the questions. During our discussions it was clear that it is not only important to continue the work on these challenges but that it is also worth to follow up with a more specific focus on measurement aspects.

Acknowledgments

The editors of this report would like to thank all participants for very fruitful and open-minded discussions! In particular, we thank the operators for sharing practical insights.

We gratefully acknowledge the Dagstuhl staff for helping on all administrative coordination, for their patience, and most importantly for providing an extremely inspiring environment.

Copyright Georg Carle and Jochen Schiller and Steve Uhlig and Walter Willinger and Matthias Wählisch

Participants
  • Bernhard Ager (ETH Zürich, CH) [dblp]
  • Lothar Braun (TU München, DE) [dblp]
  • Randy Bush (Internet Initiative Japan Inc. - Tokyo, JP) [dblp]
  • Georg Carle (TU München, DE) [dblp]
  • Nikolaos Chatzis (TU Berlin, DE) [dblp]
  • Kenjiro Cho (Internet Initiative Japan Inc. - Tokyo, JP) [dblp]
  • David Choffnes (Northeastern University - Boston, US) [dblp]
  • Alberto Dainotti (San Diego Supercomputer Center, US) [dblp]
  • Roland Dobbins (Arbor Networks - Singapore, SG)
  • Anja Feldmann (TU Berlin, DE) [dblp]
  • Timothy G. Griffin (University of Cambridge, GB) [dblp]
  • Thomas Häberlen (ENISA - Athens, GR) [dblp]
  • Ethan Katz-Bassett (University of Southern California - Los Angeles, US) [dblp]
  • Stefan Katzenbeisser (TU Darmstadt, DE) [dblp]
  • Rossella Mattioli (ENISA - Athens, GR)
  • Matthew Roughan (University of Adelaide, AU) [dblp]
  • Jochen Schiller (FU Berlin, DE) [dblp]
  • Johann Schlamp (TU München, DE) [dblp]
  • Thomas C. Schmidt (HAW - Hamburg, DE) [dblp]
  • Yuval Shavitt (Tel Aviv University, IL) [dblp]
  • Georgios Smaragdakis (T-Labs - Berlin DE & TU Berlin, DE) [dblp]
  • Rade Stanojevic (Telefónica Research - Barcelona, ES) [dblp]
  • Steve Uhlig (Queen Mary University of London, GB) [dblp]
  • Matthias Wählisch (FU Berlin, DE) [dblp]
  • Walter Willinger (AT&T Labs Research - Florham Park, US) [dblp]
  • Bill Woodcock (PCH - San Francisco, US) [dblp]

Related Seminars
  • Dagstuhl Seminar 17511: The Critical Internet Infrastructure Revisited (2017-12-17 - 2017-12-20) (Details)

Classification
  • modelling / simulation
  • networks
  • security / cryptology

Keywords
  • Internet
  • Backbone
  • Internet Services
  • Critical Infrastructure