Time Efficient Self-stabilizing Stable Marriage - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Reports (Research Report) Year : 2016

Time Efficient Self-stabilizing Stable Marriage

Joffroy Beauquier
  • Function : Author
  • PersonId : 913519
Janna Burman
Marie Laveau
Connectez-vous pour contacter l'auteur


“Stable marriage” refers to a particular matching with constraints, in- troduced in the economic context of two-sided markets. The problem has a wide variety of other applications in different domains, such as Cloud computing, Internet content delivery or college admissions. Most of the solutions known up to now are given for the synchronous model, in which executions proceed in rounds, and assume initialization. In this paper, we consider a distributed and asynchronous context, without initializa- tion (i.e., in a self-stabilizing manner - tolerating any transient faults) and with some confidentiality requirements. The single already known self-stabilizing solution [24], based on Ackerman et al.’s algorithm [1], is in O(n 4 ) moves (activation of a single node). We considerably im- prove on this previous result by presenting a solution with an complexity of O(n 2 ) moves (and O(n 2 ) asynchronous rounds), relying on Gale and Shapley’s algorithm [14]. This algorithm runs also in O(n 2 ) moves, but in a centralized synchronous context. Moreover it is not self-sabilizing and a corruption cannot be repaired locally, as noticed by Knuth [21].
Fichier principal
Vignette du fichier
GSA distributed asynchron.pdf (390.34 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-01266028 , version 1 (01-02-2016)
hal-01266028 , version 2 (13-09-2018)


  • HAL Id : hal-01266028 , version 2


Joffroy Beauquier, Thibault Bernard, Janna Burman, Shay Kutten, Marie Laveau. Time Efficient Self-stabilizing Stable Marriage. [Rapport de recherche] LRI - CNRS, University Paris-Sud. 2016. ⟨hal-01266028v2⟩
170 View
347 Download


Gmail Facebook Twitter LinkedIn More