A Verification by Abstraction Framework for organizational Multi-Agent Systems
Nicolas Gaud, Vincent Hilaire, Stéphane Galland, Abderrafìâa Koukam, Massimo Cossentino · 2008
Software agents and multi-agent systems (MAS from now on) are recognized as both abstractions and effective technologies for modelling and building complex distributed applications. However, they are still difficult to engineer. The reason is that when massive number of autonomous components interact it is very difficult to predict that the emergent organizational structure fits the system goals or that the desired functionalities will be fulfilled. Verification approaches try to evaluate whether or not a product, service, or system complies with a specification. However verification approaches are limited by the state-space of the system under study. This paper proposes an approach based upon an organizational framework and specifically the capacity concept which enables to abstract a role know-how and to reduce the state space of the system under study. A formal framework based on multi-formalisms language and the specification approach are presented and illustrated through the specification of a part of the contract net protocol.