SUMMARY

Modern life is enhanced by complex Systems of Systems (SoS), networks that combine constituents such as financial systems, power infrastructure, and transportation networks. Networking constituents increases the possible services and utility than can be provided. The response of a SoS to unexpected subsystem failures, however, undermines its effectiveness and could mitigate the advantages of combining constituent systems. The increased complexity of SoS and multi-layer interaction effects hamper the use of traditional System Engineering approaches (e.g., design by decomposition) to mitigate these unexpected failures. One possible solution to unexpected failures would be to increase the SoS’s resilience. Resilience is an emergent property that describes the ability of a SoS to resist faults, minimize disruption during a fault, and recover from a fault. Current strategies to increase resilience are hampered by the level of complexity within SoS. This dissertation proposes a novel method to increase SoS resilience. We hypothesize that biologically inspired design can increase resilience by informing both the structure of SoS interactions and the agent interactions. This dissertation achieves three research goals which result in two primary contributions to the new field of System of System Engineering (SoSE). The primary tools used in this dissertation are Ecological Network Analysis, System Dynamic Modeling, and Agent-Based Modeling. The result of this research will be two approaches to increase resilience. First, we identify a set of 13 Network Structure Design Heuristics to guide SoS evolution. Secondly, a framework for implementing biologically inspired agent behavior to increase resilience is developed and presented. 14 Agent Interaction Strategy Tactics are developed and tested. Application of these approaches could dramatically increase the sustainability and resilience of modern SoS.