$ABC^+$

Abstract

The system of systems (SoS) is a large-scale and complex system composed of autonomous and independent constituent systems (CSs). It deals with complex requirements as SoS-level goals that are not able to be satisfied by a single CS. To analyze SoS-level behaviors and SoS-level goal achievements considering CSs&amp

rsquo

autonomy and independence, SoS-level engineers and managers need to model and simulate an SoS properly. Action-Benefit-Cost (ABC) modeling provides an effective and efficient way to model and simulate the autonomous and independent behaviors of CSs by focusing on their external actions, benefits, and costs. However, ABC modeling does not support internal state of a CS so that the CS cannot update its cost-benefit table and makes problem in their decision-making. In addition, ABC modeling does not support communications between CSs so that the CSs cannot share their information during the simulation. SoS engineers also have difficulty in reflecting all possible interaction into models in advance of simulation. In this paper, we propose $ABC^+$ modeling, which supports the three-phased stateful decision-making model and the message-based interaction model. The case study shows that $ABC^+$ modeling improves the expressiveness of SoS models and lets SoS engineers simulate and analyze the SoS-level behaviors by different interaction patterns, which cannot be even represented in the existing ABC modeling. The simulation execution time is examined about both ABC and $ABC^+$ modeling, including its relationship with message counts varying from different interaction models.