Resilience-based System Importance Measures for System-of-Systems

Abstract Systems-of-systems (SoS) like the air transportation system and missile defense are gaining increasing attention in both the academic and practitioner communities. This research investigates one crucial aspect of SoSs: their ability to recover from disruptions, or their resilience. We develop a family of system importance measures (SIMs) that rank the constituent systems based on their impact on the overall SoS performance. The SIMs address some of the major weaknesses that have prevented researchers from identifying a single resilience metric. While trade-space analyses are standard practice in systems engineering, conducting trades on SoS resilience is difficult because, to date, no reliable and consistent metrics have been developed for SoS resilience. Some metrics have been proposed, but these measures assume homogenous networks, thus ignoring one of the key features of SoSs: the combination of heterogeneous systems (e.g., airports and aircraft) to achieve a common goal (e.g., transport). Instead of focusing on an overall metric, the set of SIMs provides designers with specific information on where an SoS is lacking resilience (or has excess resilience) and hence on where improvements are needed (or where downgrades are possible).