(A) simulation-based behavior analysis for MCI response system of systems

Abstract

System of systems (SoS) vastly differs from conventional systems, both in structure and means of goal achievement. Structurally, SoS contains autonomous systems which are managerially and operationally independent. The autonomous systems composing SoS, commonly referred as constituent systems (CSs), interact each other to achieve common goals. With regard to means to goal achievement, SoS relies on each CSs role and assigned responsibilities. Despite the fact that it may not have full control over their interactions, SoS is entirely dependent on them to fulfill its goal. Due to the complex nature of SoS, which results in its main characteristics including evolutionary development and associated emergent behaviors, there still remains additional efforts to deal and address the challenges posed in the modeling and analysis of SoS’ properties. In this paper, we presented SoS behavior analysis approach via simulation. Some of the main reasons for choosing a simulation approach to this purpose are (1) it is infeasible and impractical to conduct an experiment on real SoS, and (2) simulation gives insight about the unique characteristics of SoS and helps to visualize them. A similar approach is used from the chaos engineering perspective to study and analyze the behavior of complex and distributed systems. Our simulation engine is based on discrete time multi-agent simulation. For our experimentation, to mimic the real-world phenomenon into the simulation, we identified the real-world events that occur in the real-world Mass Casualty Incident (MCI) response SoS. With the identified real-world events, we defined the stimulus which can represent the real-world events and issues in real-world MCI. The identified stimulus, then injected into the developed simulation to mimic the real-life MCI response case in practice. We evaluated the effectiveness of SoS behavior analysis through the MCI scenario.