Greedy control of cascading failures in interdependent networks

Area of Science:

• Complex Systems Science
• Network Science
• Control Theory

Background:

• Complex systems exhibit nonlinear responses and feedback, making them difficult to control.
• Interdependent systems face increased systemic risk due to cross-system propagation of failures.
• Cascading failures in networks are a significant challenge in various domains.

Purpose of the Study:

• To explore optimal control strategies for cascading failures in interdependent networks.
• To analyze the dynamics of a modified Bak-Tang-Wiesenfeld sandpile model with control mechanisms.
• To understand how network topology influences control strategy effectiveness.

Main Methods:

• Utilizing the Bak-Tang-Wiesenfeld sandpile model with an integrated control mechanism.
• Simulating sandpile-like dynamics near the critical state (supercritical and subcritical regimes).
• Investigating the impact of topological coupling on control settings and failure propagation.

Main Results:

• Coupling a supercritical layer with a subcritical layer reduces avalanche probability, but benefits only one layer (parasitic relation).
• A mutualistic configuration, where both layers adopt the same control strategy, is another optimal solution.
• Interdependence necessitates coordinated control strategies across connected networks.

Conclusions:

• Optimal control of cascading failures in interdependent networks involves strategic coupling of layers with different dynamics.
• Both parasitic and mutualistic configurations offer benefits, highlighting the importance of control strategy selection.
• Findings offer insights into managing systemic risk in interconnected complex systems.