An attack-defense game model in infrastructure networks under link hiding

Area of Science:

• Network Security
• Game Theory
• Complex Systems

Background:

• Infrastructure networks are increasingly vital, raising concerns about their protection.
• Existing strategies combine complex network and game theory but overlook hidden link effects.
• A novel approach is needed to address the information asymmetry in attack-defense scenarios.

Purpose of the Study:

• To develop a novel attack-defense game model incorporating link hiding strategies.
• To investigate the impact of node properties and cost constraints on link hiding effectiveness.
• To enhance the resilience of infrastructure networks against sophisticated attacks.

Main Methods:

• Introduced a rule for link hiding based on node properties with a maximum limit.
• Developed a Stackelberg game model for attack-defense dynamics with link hiding.
• Conducted experiments on scale-free networks and a real-world power grid.

Main Results:

• Defender strategies combining first-mover advantage and link hiding improve outcomes across various node costs.
• Hiding 50% of links in a power grid reduced damage by 22.8% on average under equal resource investment.
• Link hiding proved more effective when attackers possess greater resources and against high-degree attack strategies.

Conclusions:

• Link hiding is a viable strategy to mislead attackers and improve network defense.
• The proposed model offers a more realistic simulation of network attack-defense scenarios.
• This approach significantly enhances infrastructure network resilience, particularly in power grids.