DocumentCode :
1810065
Title :
Measuring the mixing time of a network
Author :
Foukas, Xenofon ; Carzaniga, Antonio ; Wolf, Alexander L.
Author_Institution :
Sch. of Inf., Univ. of Edinburgh, Edinburgh, UK
fYear :
2015
fDate :
April 26 2015-May 1 2015
Firstpage :
2749
Lastpage :
2757
Abstract :
Mixing time is a global property of a network that indicates how fast a random walk gains independence from its starting point. Mixing time is an essential parameter for many distributed algorithms, but especially those based on gossip. We design, implement, and evaluate a distributed protocol to measure mixing time. The protocol extends an existing algorithm that models the diffusion of information seen from each node in the network as the impulse response of a particular dynamic system. In its original formulation, the algorithm was susceptible to topology changes (or “churn”) and was evaluated only in simulation. Here we present a concrete implementation of an enhanced version of the algorithm that exploits multiple parallel runs to obtain a robust measurement, and evaluate it using a network testbed (Emulab) in combination with a peer-to-peer system (FreePastry) to assess both its performance and its ability to deal with network churn.
Keywords :
distributed algorithms; peer-to-peer computing; transient response; Emulab; FreePastry; distributed algorithm; distributed protocol; gossip algorithm; impulse response; network churn; network mixing time; network testbed; peer-to-peer system; Algorithm design and analysis; Eigenvalues and eigenfunctions; Estimation; Heuristic algorithms; Peer-to-peer computing; Protocols; Time measurement;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer Communications (INFOCOM), 2015 IEEE Conference on
Conference_Location :
Kowloon
Type :
conf
DOI :
10.1109/INFOCOM.2015.7218667
Filename :
7218667
Link To Document :
بازگشت