文章基本信息

标题：Computing Probabilistic Bisimilarity Distances for Probabilistic Automata
本地全文：下载
作者：Giorgio Bacci ; Giovanni Bacci ; Kim G. Larsen 等
期刊名称：LIPIcs : Leibniz International Proceedings in Informatics
电子版ISSN：1868-8969
出版年度：2019
卷号：140
页码：1-17
DOI：10.4230/LIPIcs.CONCUR.2019.9
出版社：Schloss Dagstuhl -- Leibniz-Zentrum fuer Informatik
摘要：The probabilistic bisimilarity distance of Deng et al. has been proposed as a robust quantitative generalization of Segala and Lynch's probabilistic bisimilarity for probabilistic automata. In this paper, we present a novel characterization of the bisimilarity distance as the solution of a simple stochastic game. The characterization gives us an algorithm to compute the distances by applying Condon's simple policy iteration on these games. The correctness of Condon's approach, however, relies on the assumption that the games are stopping. Our games may be non-stopping in general, yet we are able to prove termination for this extended class of games. Already other algorithms have been proposed in the literature to compute these distances, with complexity in UP cap coUP and PPAD. Despite the theoretical relevance, these algorithms are inefficient in practice. To the best of our knowledge, our algorithm is the first practical solution. In the proofs of all the above-mentioned results, an alternative presentation of the Hausdorff distance due to Mémoli plays a central rôle.
关键词：Probabilistic automata; Behavioural metrics; Simple stochastic games; Simple policy iteration algorithm