A model analysis of a distributed monitoring system using a multi-formalism approach

Automated applications for environmental monitoring are an important aid for security and safety of buildings and for civil (domotic) and industrial use. The relevance of such applications is sound, and ensuring a correct fulfilling of the mission goals implies the need for performability and dependability requirements. For reasons of cost, flexibility and efficiency, a good candidate architecture is based on mobile agents capable of real time operation on a network of intelligent sensors, in addition of normal monitoring tasks. Design and tuning of these systems is a very hard task. In this paper a multiformalism approach based on model-checking techniques for Timed Automata and Timed Petri Nets is presented to model a real-time-mobile-agents based monitoring system. With this approach, different component of the system may be modeled by using the most suitable modeling formalism to cope with the need for modeling agent behaviors, real-time constraints and the load of the overall system.