Pharos network coordinates

Pharos is a hierarchical and decentralized network coordinate system. With the help of a simple two-level architecture, it achieves much better prediction accuracy then the representative Vivaldi coordinates, and it is incrementally deployable.

Overview
  • Pharos[4] is a fully decentralized NC system. All nodes in Pharos form two levels of overlays, namely a base overlay for long link prediction, and a local cluster overlay for short link prediction. The Vivaldi algorithm is applied to both the base overlay and the local cluster. As a result, each Pharos node has two sets of coordinates. The coordinates calculated in the base overlay, which are named global NC, are used for the global scale, and the coordinates calculated in the corresponding local cluster, which are named local NC, covers a smaller range of distance.
  • To form the local cluster, Pharos uses a method similar to binning and chooses some nodes called anchors to help node clustering. This method only requires a one-time measurement (with possible periodic refreshes) by the client to a small, fixed set of anchors. Any stable nodes which are able to response ICMP ping message can serve as anchor, such as the existing DNS servers.
  • The experimental results show that Pharos greatly outperforms Vivaldi in internet distance prediction without adding any significant overhead.

Insights behind Pharos
  • Simple and effective, obtain significant improvement in prediction accuracy by introducing a straightforward hierarchical distance prediction
  • Fully compatible with Vivaldi, the most widely deployed NC system. For every host where the Vivaldi client has been deployed, it just needs to run classic Vivaldi NC algorithm to join global overlay and local cluster, without deploying another NC client.
  • The anchors in Pharos is different from landmarks in Global network positioning (GNP),[5] which not only has to reply the ICMP ping but also need to reply the queries from all clients by sending their latest NCs. No requirement to deploy any extra software on the anchors.

See also

References
  1. S. Rhea; D. Geels; T. Roscoe; et al. (2004). "Handling Churn in a DHT" (PDF). Proceedings of the USENIX Annual Technical Conference (ATC'04).
  2. P. Pietzuch; J. Ledlie; J. Shneidman; et al. (2006). "Network-Aware Operator Placement for Stream-Processing Systems" (PDF). 22nd International Conference on Data Engineering (ICDE'06).
  3. J. Ledlie; P. Gardner & M. Seltzer (2007). "Network Coordinates in the Wild" (PDF). 4th USENIX Symposium on Networked Systems Design & Implementation.
  4. Y. Chen; Y. Xiong; X. Shi; et al. (April 2009). "Pharos: Accurate and Decentralised Network Coordinate System" (PDF). IET Communications. 3 (4): 539–548. doi:10.1049/iet-com.2008.0187. S2CID 11701454. Archived from the original (PDF) on 2013-12-03. Retrieved 2013-11-27.
  5. T. S. E. Ng & H. Zhang (2002). "Predicting Internet Network Distance with Coordinates-based Approaches". IEEE INFOCOM.
This article is issued from Wikipedia. The text is licensed under Creative Commons - Attribution - Sharealike. Additional terms may apply for the media files.