Internet Measurement Data Catalog

Featured Collections and Publications

CAIDA skitter AS Links Topology - 2854 files, 2000‑01‑02 to 2008‑02‑08

Autonomous system (AS) topology derived from skitter traces (an AS is roughly an ISP). Possible uses include studying statistical and topological properties of AS graphs, constructing realistic Internet topologies for modeling and simulation, and studying AS relationships. Data collection has been continuing for over 7 years as of 2007.

CAIDA Witty Worm Data, public access - 7 files, 2004‑03‑20 to 2004‑03‑25

Information useful for studying the spread of the Witty worm, as observed by the UCSD Network Telescope over a 5-day period in Mar 2004. This dataset consists of public-access files that do not individually identify infected computers. Data available include time, duration, country, and connection speed distributions of infected hosts. This public-access dataset does not include packet traces of traffic generated by infected hosts. Possible uses include modeling worm propagation. Statistics: 55,909 infected IP addresses.

CAIDA OC48 Traces 2003-04-24 - 26 files, 2003‑04‑24 to 2003‑04‑24

Anonymized packet header traces collected in both directions of an OC48 link on Apr 24, 2003 (1 hour). This link is a west coast peering link for a large ISP. Possible uses include research on the characteristics of traffic, including application breakdown, security events, geographic and topological distribution, and flow volume and duration. Statistics (both directions): 13GB of traces, 203 million packets, and 96GB of observed IP traffic.

CAIDA Backscatter-2004-2005 - 63 files, 2004‑05‑26 to 2005‑12‑01

Information useful for longitudinal study of denial-of-service (DoS) attacks. This dataset consists of 5.5 billion IPv4 packets sent by DoS attack victims in response to spoofed attack traffic. This backscatter from victims was collected by the UCSD Network Telescope, one week of data per quarter, between May 2004 and November 2005. Possible uses include modeling DoS attacks, understanding victim populations, and using real packet traces to validate algorithms for detecting or classifying malicious traffic. This last use is particularly valuable because it is extremely challenging to artificially generate the kind of real-world noise present on the Internet.

Symbols indicate collections or publications with annotations of type  notice or  alert