Representing domains and scenarios by means of model replication and composition

We consider a domain as a particular system or a portion of the system, while a scenario is a sequence of effects on the domain, originated by a particular event or condition. We show how it is possible to build first the model of the domain by replication and composition of atomic models, each representing a particular aspect of the domain. Then, the models of the scenarios are obtained from the domain’s model, by composing further atomic models representing the events originating the scenarios. In particular, we take into account the domain consisting of one control centre and a set of substations inside an electrical distribution grid, communicating by means of a network. We consider scenarios originated by threats such as the denial of service attack to the communication network, and the temporary unavailability of substations due to the failure and the repair of the internal components. Stochastic Activity Networks (SAN) are the modelling formalism. The simulation of the models representing the scenarios, estimates the impact of the threats on the communication reliability.