Weighted gossip: Distributed averaging using non-doubly stochastic matrices
This paper presents a general class of gossip-based averaging algorithms, which are inspired from Uniform Gossip. While Uniform Gossip works synchronously on complete graphs, weighted gossip algorithms allow asynchronous rounds and converge on any connected, directed or undirected graph. Unlike most...
| Main Authors: | , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | en_US |
| Published: |
Institute of Electrical and Electronics Engineers (IEEE)
2012
|
| Online Access: | http://hdl.handle.net/1721.1/73148 https://orcid.org/0000-0003-2658-8239 |
| _version_ | 1826199844766613504 |
|---|---|
| author | Tsitsiklis, John N. Benezit, Florence Blondel, Vincent D. Thiran, Patrick Vetterli, Martin |
| author2 | Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science |
| author_facet | Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Tsitsiklis, John N. Benezit, Florence Blondel, Vincent D. Thiran, Patrick Vetterli, Martin |
| author_sort | Tsitsiklis, John N. |
| collection | MIT |
| description | This paper presents a general class of gossip-based averaging algorithms, which are inspired from Uniform Gossip. While Uniform Gossip works synchronously on complete graphs, weighted gossip algorithms allow asynchronous rounds and converge on any connected, directed or undirected graph. Unlike most previous gossip algorithms, Weighted Gossip admits stochastic update matrices which need not be doubly stochastic. Double-stochasticity being very restrictive in a distributed setting, this novel degree of freedom is essential and it opens the perspective of designing a large number of new gossip-based algorithms. To give an example, we present one of these algorithms, which we call One-Way Averaging. It is based on random geographic routing, just like Path Averaging, except that routes are one way instead of round trip. Hence in this example, getting rid of double stochasticity allows us to add robustness to Path Averaging. |
| first_indexed | 2024-09-23T11:26:40Z |
| format | Article |
| id | mit-1721.1/73148 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T11:26:40Z |
| publishDate | 2012 |
| publisher | Institute of Electrical and Electronics Engineers (IEEE) |
| record_format | dspace |
| spelling | mit-1721.1/731482022-09-27T19:32:45Z Weighted gossip: Distributed averaging using non-doubly stochastic matrices Tsitsiklis, John N. Benezit, Florence Blondel, Vincent D. Thiran, Patrick Vetterli, Martin Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Tsitsiklis, John N. Tsitsiklis, John N. This paper presents a general class of gossip-based averaging algorithms, which are inspired from Uniform Gossip. While Uniform Gossip works synchronously on complete graphs, weighted gossip algorithms allow asynchronous rounds and converge on any connected, directed or undirected graph. Unlike most previous gossip algorithms, Weighted Gossip admits stochastic update matrices which need not be doubly stochastic. Double-stochasticity being very restrictive in a distributed setting, this novel degree of freedom is essential and it opens the perspective of designing a large number of new gossip-based algorithms. To give an example, we present one of these algorithms, which we call One-Way Averaging. It is based on random geographic routing, just like Path Averaging, except that routes are one way instead of round trip. Hence in this example, getting rid of double stochasticity allows us to add robustness to Path Averaging. Swiss National Science Foundation (grant 5005-67322) National Science Foundation (U.S.) (grant ECCS-0701623) 2012-09-25T12:45:25Z 2012-09-25T12:45:25Z 2010-07 2010-06 Article http://purl.org/eprint/type/ConferencePaper 978-1-4244-7890-3 978-1-4244-7891-0 http://hdl.handle.net/1721.1/73148 Tsitsiklis, John N. et al. "Weighted Gossip: Distributed Averaging using non-doubly stochastic matrices." Proceedings of the 2010 IEEE International Symposium on Information Theory (ISIT): 1753-1757. © 2010 IEEE. https://orcid.org/0000-0003-2658-8239 en_US http://dx.doi.org/ 10.1109/ISIT.2010.5513273 Proceedings of the IEEE International Symposium on Information Theory (ISIT), 2010 Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use. application/pdf Institute of Electrical and Electronics Engineers (IEEE) IEEE |
| spellingShingle | Tsitsiklis, John N. Benezit, Florence Blondel, Vincent D. Thiran, Patrick Vetterli, Martin Weighted gossip: Distributed averaging using non-doubly stochastic matrices |
| title | Weighted gossip: Distributed averaging using non-doubly stochastic matrices |
| title_full | Weighted gossip: Distributed averaging using non-doubly stochastic matrices |
| title_fullStr | Weighted gossip: Distributed averaging using non-doubly stochastic matrices |
| title_full_unstemmed | Weighted gossip: Distributed averaging using non-doubly stochastic matrices |
| title_short | Weighted gossip: Distributed averaging using non-doubly stochastic matrices |
| title_sort | weighted gossip distributed averaging using non doubly stochastic matrices |
| url | http://hdl.handle.net/1721.1/73148 https://orcid.org/0000-0003-2658-8239 |
| work_keys_str_mv | AT tsitsiklisjohnn weightedgossipdistributedaveragingusingnondoublystochasticmatrices AT benezitflorence weightedgossipdistributedaveragingusingnondoublystochasticmatrices AT blondelvincentd weightedgossipdistributedaveragingusingnondoublystochasticmatrices AT thiranpatrick weightedgossipdistributedaveragingusingnondoublystochasticmatrices AT vetterlimartin weightedgossipdistributedaveragingusingnondoublystochasticmatrices |