Research on network robustness analysis method of combat system-of-systems considering collaborative reconfiguration

The increasing complexity of the intelligent combat system-of-systems (CSoS) necessitates advanced robustness mechanisms that leverage collaborative reconfiguration. However, existing studies face a key challenge: the lack of quantitative evaluation of the robustness of the CSoS under coordinated node failure, with a particular focus on the interaction between system reconfiguration strategies and adversarial attack patterns. To address this challenge, three innovative collaborative reconfiguration strategies are proposed: random node collaborative (RNC) reconfiguration, maximum structural similarity node collaborative (MSSNC) reconfiguration, and maximum functional similarity node collaborative (MFSNC) reconfiguration. To explore the impact of different attack methods on network robustness, these strategies are systematically tested in five attack scenarios: random node attack (RA), preferential mixed node attack (PMA), preferential sensor node attack (PSA), preferential decision-making node attack (PDA), and preferential influence node attack (PIA). The findings demonstrate that collaborative reconfiguration strategies improve the robustness of the CSoS and that the MFSNC collaborative strategy has the best effect. In addition, the analysis based on node contribution rates reflects the importance of nodes and indirectly indicates the internal mechanism by which collaborative reconfiguration strategies enhance robustness. This study is important for CSoS robustness analysis.