Skip to main navigation Skip to search Skip to main content

Faster possibility detection by combining two approaches

  • Cornell University

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

22 Scopus citations

Abstract

A new algorithm is presented for detecting whether a particular computation of an asynchronous distributed system satisfies Poss Ф (read "possibly Ф”), meaning the system could have passed through a global state satisfying Ф. Like the algorithm of Cooper and Maxzullo, Ф may be any global state predicate; and like the algorithm of Gaxg and Waldecker, Poss Ф is detected quite efficiently if Ф has a certain structure. The new algorithm exploits the structure of some predicates not handled by Gaxg and Waldecker's algorithm to detect Poss cI, more efficiently than is possible with any algorithm that, like Cooper and Maxzullo's, evaluates Ф on every global state through which the system could have passed. A second algorithm is also presented for off-line detection of Poss Ф. It uses Strassen's scheme for fast matrix multiplication. The intrinsic complexity of off-line and on-line detection of Poss Ф is discussed.

Original languageEnglish
Title of host publicationDistributed Algorithms - 9th International Workshop, WDAG 1995, Proceedings
EditorsJean-Michel Helary, Michel Raynal
PublisherSpringer Verlag
Pages318-332
Number of pages15
ISBN (Print)3540602747, 9783540602743
DOIs
StatePublished - 1995
Event9th International Workshop on Distributed Algoriths, WDAG 1995 - Le Mont-Saint-Michel, France
Duration: Sep 13 1995Sep 15 1995

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume972
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference9th International Workshop on Distributed Algoriths, WDAG 1995
Country/TerritoryFrance
CityLe Mont-Saint-Michel
Period09/13/9509/15/95

Fingerprint

Dive into the research topics of 'Faster possibility detection by combining two approaches'. Together they form a unique fingerprint.

Cite this