Simulations for Multi-Agent Systems with Imperfect Information

Patrick Gardy, Yuxin Deng

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

1 Scopus citations

Abstract

Equivalence-checking and simulations are well-known methods used to reduce the size of a system in order to verify it more efficiently. While Alur et al. proposed a notion of simulation sound and complete for ATL as early as 1998, there have been very few works on equivalence-checking performed on extensions of ATL* with probabilities, imperfect information, counters etc. In the case of multi-agent systems (MASs) with imperfect information, the lack of sound and complete algorithm mostly follows from the undecidability of ATL model-checking. However, while ATL is undecidable overall, there exist sub-classes of MASs for which ATL becomes decidable. In this paper, we propose a notion of simulation sound for ATL/ATL* on any MASs and complete on naive MASs. Using our simulations we design an equivalence-checking algorithm sound and complete for MASs with public actions.

Original languageEnglish
Title of host publicationFormal Methods and Software Engineering - 21st International Conference on Formal Engineering Methods, ICFEM 2019, Proceedings
EditorsYamine Ait-Ameur, Shengchao Qin
PublisherSpringer
Pages138-153
Number of pages16
ISBN (Print)9783030324087
DOIs
StatePublished - 2019
Event21st International Conference on Formal Engineering Methods, ICFEM 2019 - Shenzhen, China
Duration: 5 Nov 20199 Nov 2019

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11852 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference21st International Conference on Formal Engineering Methods, ICFEM 2019
Country/TerritoryChina
CityShenzhen
Period5/11/199/11/19

Fingerprint

Dive into the research topics of 'Simulations for Multi-Agent Systems with Imperfect Information'. Together they form a unique fingerprint.

Cite this