License: Creative Commons Attribution 3.0 Unported license (CC BY 3.0)
When quoting this document, please refer to the following
DOI: 10.4230/LIPIcs.DISC.2017.59
URN: urn:nbn:de:0030-drops-79649
URL: http://dagstuhl.sunsite.rwth-aachen.de/volltexte/2017/7964/
Go to the corresponding LIPIcs Volume Portal

Massoulié, Laurent ; Varloot, Rémi

Brief Announcement: Rapid Mixing of Local Dynamics on Graphs

pdf-format:
LIPIcs-DISC-2017-59.pdf (0.3 MB)

Abstract

In peer-to-peer networks, it is desirable that the logical topology of connections between the constituting nodes make a well-connected graph, i.e., a graph with low diameter and high expansion. At the same time, this graph should evolve only through local modifications. These requirements prompt the following question: are there local graph dynamics that i) create a well-connected graph in equilibrium, and ii) converge rapidly to this equilibrium?

In this paper we provide an affirmative answer by exhibiting a local graph dynamic that mixes provably fast. Specifically, for a graph on N nodes, mixing has occurred after each node has performed O(polylog(N)) operations. This is in contrast with previous results, which required at least Omega(N polylog(N)) operations per node before the graph had properly mixed.

BibTeX - Entry

@InProceedings{massouli_et_al:LIPIcs:2017:7964,
  author =	{Laurent Massouli{\'e} and R{\'e}mi Varloot},
  title =	{{Brief Announcement: Rapid Mixing of Local Dynamics on Graphs}},
  booktitle =	{31st International Symposium on Distributed Computing (DISC 2017)},
  pages =	{59:1--59:3},
  series =	{Leibniz International Proceedings in Informatics (LIPIcs)},
  ISBN =	{978-3-95977-053-8},
  ISSN =	{1868-8969},
  year =	{2017},
  volume =	{91},
  editor =	{Andr{\'e}a W. Richa},
  publisher =	{Schloss Dagstuhl--Leibniz-Zentrum fuer Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{http://drops.dagstuhl.de/opus/volltexte/2017/7964},
  URN =		{urn:nbn:de:0030-drops-79649},
  doi =		{10.4230/LIPIcs.DISC.2017.59},
  annote =	{Keywords: Markov chains, Mixing time, Dynamic graphs, Local dynamics}
}

Keywords: Markov chains, Mixing time, Dynamic graphs, Local dynamics
Collection: 31st International Symposium on Distributed Computing (DISC 2017)
Issue Date: 2017
Date of publication: 12.10.2017

DROPS-Home | Fulltext Search | Imprint | Privacy Published by LZI