Online Distributed Interdependency Estimation for Critical Infrastructures

The paper deals with the problem to provide a global estimation of a scenario composed by several and interdependent infrastructures. Due to the increasing presence of interdependencies, to correctly manage any infrastructure it is more and more relevant to have information about the "state" of the others, especially in critical situations. However, it is unfeasible both any centralized solution and to exchange detailed data about the state of each infrastructure, due to their huge volumes and in order to avoid the disclosure of critical information. To overcome such problems in this paper a fully distributed approach is proposed, attesting a copy of the whole model inside each one of the infrastructures' control room; each copy directly receives information coming from its own field and shares with the others only aggregated and non-sensible data. Moreover, in order to provide consistent outputs, the different copies have to synchronize. To this end a specific procedure has been developed, and some conditions for the convergence have been identified. In this way the proposed system is able to operate as an Online Distributed Interdependency Estimation (ODIE) System. The approach is under experimental tests on a national wide electric and telecommunication networks inside the European Project MICIE.