Design and control of multi-agent systems correspond to the synthesis of winning strategies in games that model the interaction between the agents. In games with full observability, the strategies of players depend on the full history of the play. In games with partial observability, strategies depend only on observable components of the history. We survey two approaches to partial observability in two-player turn-based games with behavioral winning conditions. The first is the traditional longitudinal observability, where in all vertices, the players observe the assignment only to an observable subset of the atomic propositions. The second is the recently studied transverse observability, where players observe the assignment to all the atomic propositions, but only in vertices they own.
|Original language||American English|
|Title of host publication||Reachability Problems - 14th International Conference, RP 2020, Proceedings|
|Editors||Sylvain Schmitz, Igor Potapov|
|Publisher||Springer Science and Business Media Deutschland GmbH|
|Number of pages||15|
|State||Published - 2020|
|Event||14th International Conference on Reachability Problems, RP 2020 - Paris, France|
Duration: 19 Oct 2020 → 21 Oct 2020
|Name||Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|
|Conference||14th International Conference on Reachability Problems, RP 2020|
|Period||19/10/20 → 21/10/20|
Bibliographical noteFunding Information:
O.K—Supported in part by the Israel Science Foundation, grant No. 2357/19.
© 2020, Springer Nature Switzerland AG.