Agent-based input-output interdependency model

The modeling and analysis of critical infrastructures and their interdependencies are essential to discovering hidden vulnerabilities and the related threats to national and international security. Over the past few years, several approaches have been proposed to address this problem. The so-called holistic approaches are relatively abstract, but are easily validated using real economic data. Other approaches based on agent-based models provide deeper views of the interdependencies existing between subsystems of different infrastructures. However, agent-based models are often difficult to validate because quantitative data of the appropriate granularity may not be available. This paper presents an agent-based input–output inoperability model designed to overcome the limitations of the holistic and agent-based paradigms. In order to provide a detailed and expressive framework, the exchange of resources between infrastructures is explicitly modeled while inoperability becomes an internal parameter. Nevertheless, the model is easily transformed into a fine-grained, input–output inoperability model whose coefficients can be obtained based on real data.

Questo lavoro presenta una formulazione del modello IIM su base micro. La ratio del lavoro è quella di poter superare i limiti del IIM classico fornendo uno strumento in grado di operare non solo su scala macro-scopica. Un altro vantaggio della formulazione presentata è la possibilità di ricondurre i parametri macro (facilmente reperibili) alle quantità micro del modello e, quindi, utilizzare gli stessi per la validazione del modello stesso, superando in questo modo i ben noti limiti degli approcci agent-based. Sono proposte, infine, alcune condizioni per la esistenza della soluzione di regime del modello IIM e della sue varianti nonché per il calcolo delle stesse.