Cascading failures in complex networks occur when local node or edge failures propagate to trigger large-scale collapse. Traditional pairwise network models cannot adequately capture group coordination and multi-agent higher-order interactions. Higher-order networks incorporating simplicial structures more accurately represent group and multi-node interactions, providing a new framework to study cascading failures and network robustness. The paper proposes a higher-order coupled map lattice (CML) model to characterize cascading failures in simplicial complex networks and analyze the influence of higher-order structures on network robustness. Further experiments on fourth-order simplicial networks investigate robustness differences under various topologies and attack strategies. Results indicate that fourth-order simplicial networks are vulnerable to targeted attacks but robust against random failures, regardless of network type. Furthermore in single-order networks, the higher simplex dimensions, the greater robustness. The theoretical perturbation thresholds for third-order networks show a negative correlation between the critical perturbation and the sum of network coupling parameters. These results are validated by analysis of simplices added to ordinary networks, destructive experiments, and empirical networks. This study deepens the understanding of cascading failure mechanisms and robustness in higher-order networks, and provides theoretical guidance for designing resilient networks based on higher-order structures.
In reality, infectious diseases rarely spread in isolation; instead, multiple diseases often spread concurrently. The infection of one disease may influence an individual’s susceptibility or transmissibility of another disease through mechanisms such as immunosuppression or symptom superposition. Furthermore, population contact structures are not limited to simple pairwise interactions, but also involve group events with simultaneous exposure, such as family gatherings. Traditional network models, which are based on pairwise interactions, are difficult to accurately capture these higher-order interaction structures and the coupling mechanisms among multiple pathogens. Therefore, this paper presents a hypergraph SIS transmission model based on dynamic thresholds. It systematically investigates the transmission dynamics of multiple diseases on uniform and non-uniform hypergraph structures in BA and ER networks. According to co-infection scenarios, three coupling mechanisms are proposed, positive coupling which promotes co-infection, negative coupling which suppresses co-infection, and no coupling where transmission occurs independently. To account for variations in initial disease intensity, three comparison groups are designed: high-low, same-high and same-low. This paper analyses the combined effects of coupling mechanisms, threshold variations and network structural characteristics on the transmission thresholds, propagation rates and infection scales of multiple diseases. These findings are validated in a dengue fever transmission network.