Internet Measurements

Internet Measurements

Artur Ziviani (National Laboratory for Scientific Computing (LNCC), Brazil)
Copyright: © 2008 |Pages: 7
DOI: 10.4018/978-1-59140-993-9.ch034
OnDemand PDF Download:


In the mid 90’s, the Internet has started its metamorphosis from a tool restricted to the scientific community into a crucial component of the modern information society. The evolution in the last 15 years from 2,000 to 180,000 active BGP (Border Gateway Protocol) entries as of March 2006 (Smith, 2006) witnesses this metamorphosis and is an indication on how complex Internet has become. Possibly, the most significant consequence of the huge success of the Internet is that the common goal and collaborative spirit that used to guide its players no longer holds because of the large diversity we find in today’s Internet. Users, commercial access providers, governments, telecommunication operators, and content providers usually have opposite interests in using the network, commonly leading to a situation where they co-exist in a tussle (Clark, Wroclawski, Sollins, & Braden, 2002). For instance, large network operators need to be interconnected to obtain and offer universal connectivity, even if they are often fierce competitors. As a result, the heterogeneity and fully distributed administration of the Internet, allied to its extensive geographic coverage and to its dynamism in applications and network traffic, impose great challenges to the characterization of the structure and behavior of the Internet as a whole (Floyd & Paxson, 2001; Spring, Wetherall, & Anderson, 2003a).

Key Terms in this Chapter

P2P Traffic: P2P traffic refers to network traffic generated by distributed applications (i.e. Gnutella, KaZaa, or Skype to name a few) that organize the communication among its participants using the peerto- peer (P2P) communication model in contrast with the classic client-server model.

Probe Packet: A packet used in an active measurement experiment to collect knowledge on a given network parameter of interest.

Self-Similarity: A process is said to be self-similar if its behavior is roughly the same across different spacial or time scales.

Active Measurement: An active measurement sends probe packets and the result of the journey of this packets through the network is monitored to estimate network characteristics.

Autonomous System: An autonomous system (AS) corresponds to a routing domain under the control of a single administrative entity (such as a university or a business enterprise, for instance).

Passive Measurement: Passive measurement refers to the process of monitoring network traffic without injecting new traffic or affecting the existing one.

Long-Range Dependence: Long-range dependence (LRD) characterizes process that keep statistically significant correlations across large time scales.

Traffic Matrix: A traffic matrix (TM) represents the volume of traffic between all possible pairs of sources and destinations in a given IP domain.

BGP: The border gateway protocol (BGP) is a protocol for exchanging routing information between autonomous systems.

Complete Chapter List

Search this Book: