Award Abstract # 1320977
NeTS: Small: Towards an Accurate, Geo-Aware, PoP-Level Perspective of the Internet's Inter-AS Connectivity

NSF Org: CNS
Division Of Computer and Network Systems
Recipient: UNIVERSITY OF OREGON
Initial Amendment Date: August 20, 2013
Latest Amendment Date: August 4, 2015
Award Number: 1320977
Award Instrument: Standard Grant
Program Manager: Darleen Fisher
CNS
 Division Of Computer and Network Systems
CSE
 Directorate for Computer and Information Science and Engineering
Start Date: October 1, 2013
End Date: September 30, 2017 (Estimated)
Total Intended Award Amount: $499,994.00
Total Awarded Amount to Date: $531,994.00
Funds Obligated to Date: FY 2013 = $499,994.00
FY 2014 = $16,000.00

FY 2015 = $16,000.00
History of Investigator:
  • Reza Rejaie (Principal Investigator)
    reza@cs.uoregon.edu
  • Bruce Maggs (Co-Principal Investigator)
  • Walter Willinger (Co-Principal Investigator)
Recipient Sponsored Research Office: University of Oregon Eugene
1776 E 13TH AVE
EUGENE
OR  US  97403-1905
(541)346-5131
Sponsor Congressional District: 04
Primary Place of Performance: University of Oregon Eugene
OR  US  97403-1202
Primary Place of Performance
Congressional District:
04
Unique Entity Identifier (UEI): Z3FGN9MF92U2
Parent UEI: Z3FGN9MF92U2
NSF Program(s): Special Projects - CNS,
Networking Technology and Syst
Primary Program Source: 01001314DB NSF RESEARCH & RELATED ACTIVIT
01001415DB NSF RESEARCH & RELATED ACTIVIT

01001516DB NSF RESEARCH & RELATED ACTIVIT
Program Reference Code(s): 7363, 7923, 9251
Program Element Code(s): 171400, 736300
Award Agency Code: 4900
Fund Agency Code: 4900
Assistance Listing Number(s): 47.070

ABSTRACT

Understanding where and how the more than 40,000 actively routed Autonomous Systems (ASes) in today's Internet interconnect is essential for meaningfully investigating a wide range of critical Internet-related problems such as the vulnerability of the Internet to physical damage. However, much of the published work on Internet topology has focused primarily on discovering the existence of such interconnections (e.g., logical connectivity such as AS-to-AS links or physical connectivity such as router-to-router links). Considerably less attention has been paid to the where and in how many different locations these interconnections have been established. For example, the often-studied AS-level view of the Internet is too coarse as mapping entire ASes to single geographic locations eliminates essential details (e.g. AS-level path diversity). At the same time, the popular router-level view of the Internet is not only too detailed, but also inherently difficult to capture.

Intellectual Merit: The main goal of this project is to design, develop and rigorously evaluate techniques to accurately map the geographic location of all the PoPs (Point-of-Presence) of a given target AS and determine the inter-AS connections that are established at each PoP (point of presence) of this AS. A significant fraction of the Internet's physical infrastructure (e.g. routers, switches) is hosted at a relatively small number of physical building complexes, called colocation (or colo) facilities or data centers that can be accurately geo-located. Thus, a core element of this project is the design of new targeted active measurement campaigns specifically developed to map a given colo facility by identifying not only all the PoPs of all the ASes present in that colo facility, but also the corresponding inter-AS connectivity that is visible to active probing at that location.

Broader Impact: The inferred PoP-level maps are leveraged to develop a new and improved simulation environment, called cBGP+, that is built on the existing simulator (cBGP) but supports real-world AS path diversity. This new simulator will enable many Internet stakeholders (e.g. ISPs, DHS) to meaningfully assess, evaluate, and predict the inter-AS reachability of the Internet in the presence of certain events or changes (e.g., political unrests or the results of natural or man-made disasters).

PUBLICATIONS PRODUCED AS A RESULT OF THIS RESEARCH

Note:  When clicking on a Digital Object Identifier (DOI) number, you will be taken to an external site maintained by the publisher. Some full text articles may not yet be available without a charge during the embargo (administrative interval).

Some links on this page may take you to non-federal websites. Their policies may differ from this site.

Bahador Yeganeh, Reza Rejaie, Walter Willinger "A View From the Edge: A Stub-AS Perspective of Traffic Localization and its Implications" IEEE/IFIP Network Traffic Measurement and Analysis Conference (TMA) , 2017
Reza Motamedi, Reza Rejaie, Walter Willinger "A Suvery of Techniques for Internet Topology Discovery" IEEE Communications Surveys and Tutorials , v.17 , 2015 , p.1044 1065
Reza Motamedi, Reza Rejaie, Walter Willinger "A Suvery of Techniques for Internet Topology Discovery" IEEE Communications Surveys and Tutorials , v.17 , 2015 , p.1044

PROJECT OUTCOMES REPORT

Disclaimer

This Project Outcomes Report for the General Public is displayed verbatim as submitted by the Principal Investigator (PI) for this award. Any opinions, findings, and conclusions or recommendations expressed in this Report are those of the PI and do not necessarily reflect the views of the National Science Foundation; NSF has not approved or endorsed its content.

 

Internet interconnections are the means by which networks exchange traffic between one another. These interconnections are typically established in facilities that have known geographic locations, and are owned and operated by so-called colocation and interconnection services providers (e.g., Equinix, CoreSite, and EdgeConneX). These previously under-studied colocation facilities and the critical role they play in solving the notoriously difficult problem of obtaining a comprehensive view of the structure and evolution of the interconnections in today's Internet were the focus of this project.

The main outcome of the project was the design, development and evaluation of mi2, a new approach for systematically mapping (both inferring and pinning) all private and public Internet Interconnections inside a given colocation facility. mi2 relies on the information derived from carefully designed traceroute-based measurement campaigns to infer the likely interconnections at a colcation facilityand leverages the Belief Propagation algorithm that has been shown to be an effective way for solving inference problems arising in diverse areas such as statistical physics, computer vision, and Artificial Intelligence. In particular, mi2 uses the Belief Propagation algorithm on a specially defined Markov Random Field graphical model to pin (geolocate) each end of inferred interconnections to the inside or outside of a target facility.   mi2 was evaluated by applying it to a diverse set of US-based colocation facilities. The results were also compared against those obtained by two recently developed related techniques.  The comparisons illustrate observed discrepancies that result from how the different techniques determine the ownership of (border) routers.

As an important by-product, the project also revealed drastic changes in today's Internet interconnection ecosystem (e.g., new infrastructures in the form of cloud exchanges that offer new peering options in the form of virtual private interconnections to new customers in the form of enterprise networks without an AS number). These observed changes have far-reaching implications for future efforts aimed at obtaining an accurate and comprehensive map of the Internet's interconnections.

 


Last Modified: 04/20/2018
Modified by: Reza Rejaie

Please report errors in award information by writing to: awardsearch@nsf.gov.

Print this page

Back to Top of page