Interactions between agents are usually designed from a global viewpoint. However, the implementation of a multi-agent interaction is distributed. It is well known that this difference between the specification and the implementation levels can introduce problems, allowing designers to specify protocols from a global viewpoint that cannot be implemented as a collection of individual agents. This leads naturally to the question of whether a given (global) protocol is enactable, namely, whether it can be implemented in a distributed way. We consider this question in the powerful setting of trace expressions, considering a range of message ordering interpretations (specifying what it means to say that an interaction step occurs before another), and a range of possible constraints on the semantics of message delivery, corresponding to different properties of the underlying communication middleware. We provide a definition of enactability, along with an implementation of the definition that is applied to a number of example protocols.
CITATION STYLE
Ferrando, A., Winikoff, M., Cranefield, S., Dignum, F., & Mascardi, V. (2020). On enactability of agent interaction protocols: Towards a unified approach. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 12058 LNAI, pp. 43–64). Springer. https://doi.org/10.1007/978-3-030-51417-4_3
Mendeley helps you to discover research relevant for your work.