File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Optimal parameters for load balancing using the diffusion method in k-ary n-cube networks

TitleOptimal parameters for load balancing using the diffusion method in k-ary n-cube networks
Authors
KeywordsDiffusion Method
Distributed Scheduling
Load Balancing
Message-Passing Multiprocessors K-Ary N-Cube Networks
Parallel Processing
Issue Date1993
PublisherElsevier BV. The Journal's web site is located at http://www.elsevier.com/locate/ipl
Citation
Information Processing Letters, 1993, v. 47 n. 4, p. 181-187 How to Cite?
AbstractThe diffusion method is a simple but practical method for dynamic load balancing in distributed memory message-passing multiprocessors. It is iterative in nature and is fully distributed. Its efficiency depends on a parameter, the diffusion parameter, which dictates how excess workloads are to be diffused away. We derive optimal values for this parameter for the cases of the k-ary n-cube and its variant, the n-dimensional torus. With these optimal parameters, the diffusion method is guaranteed to terminate at a balanced state in optimal time.
Persistent Identifierhttp://hdl.handle.net/10722/152241
ISSN
2023 Impact Factor: 0.7
2023 SCImago Journal Rankings: 0.404
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorXu, CZen_US
dc.contributor.authorLau, FCMen_US
dc.date.accessioned2012-06-26T06:36:42Z-
dc.date.available2012-06-26T06:36:42Z-
dc.date.issued1993en_US
dc.identifier.citationInformation Processing Letters, 1993, v. 47 n. 4, p. 181-187en_US
dc.identifier.issn0020-0190en_US
dc.identifier.urihttp://hdl.handle.net/10722/152241-
dc.description.abstractThe diffusion method is a simple but practical method for dynamic load balancing in distributed memory message-passing multiprocessors. It is iterative in nature and is fully distributed. Its efficiency depends on a parameter, the diffusion parameter, which dictates how excess workloads are to be diffused away. We derive optimal values for this parameter for the cases of the k-ary n-cube and its variant, the n-dimensional torus. With these optimal parameters, the diffusion method is guaranteed to terminate at a balanced state in optimal time.en_US
dc.languageengen_US
dc.publisherElsevier BV. The Journal's web site is located at http://www.elsevier.com/locate/iplen_US
dc.relation.ispartofInformation Processing Lettersen_US
dc.subjectDiffusion Methoden_US
dc.subjectDistributed Schedulingen_US
dc.subjectLoad Balancingen_US
dc.subjectMessage-Passing Multiprocessors K-Ary N-Cube Networksen_US
dc.subjectParallel Processingen_US
dc.titleOptimal parameters for load balancing using the diffusion method in k-ary n-cube networksen_US
dc.typeArticleen_US
dc.identifier.emailLau, FCM:fcmlau@cs.hku.hken_US
dc.identifier.authorityLau, FCM=rp00221en_US
dc.description.naturelink_to_subscribed_fulltexten_US
dc.identifier.doi10.1016/0020-0190(93)90030-Den_US
dc.identifier.scopuseid_2-s2.0-0027668059en_US
dc.identifier.volume47en_US
dc.identifier.issue4en_US
dc.identifier.spage181en_US
dc.identifier.epage187en_US
dc.identifier.isiWOS:A1993MA23500003-
dc.publisher.placeNetherlandsen_US
dc.identifier.scopusauthoridXu, CZ=8701888000en_US
dc.identifier.scopusauthoridLau, FCM=7102749723en_US
dc.identifier.issnl0020-0190-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats