A novel multi-agent game-theoretic model for cybersecurity strategies in EV charging networks: Addressing risk propagation and budget constraints

This paper focuses on cybersecurity strategies for Electric Vehicle Charging Service Providers (EVSPs) within interconnected Electric Vehicle Charging Networks, considering cyber risk propagation and budget constraints. A novel muti-agent game-theoretic model is developed to capture competitive interactions among EVSPs and their strategic decision-making regarding cybersecurity investments and charging transactions. To solve for the Nash equilibrium, an adapted extra-gradient algorithm based on variational inequality theory is developed, achieving 33–50 % faster convergence than traditional methods such as Euler and projection-based algorithms as the network scales. Numerical analysis validates the model and investigates the impact of various factors on cybersecurity investments, cyber risk levels, and economic performance of EVSPs. Key findings include: (1) higher network density (more EVSPs) intensifies competition and cyber risk propagation, consequently reducing individual EVSP profits and cybersecurity investments; (2) increased risk propagation probability induces divergent investment strategies, with low-vulnerability EVSPs increase their investment while high-vulnerability ones decrease theirs, revealing free-riding behavior among EVSPs and emphasizing the need for targeted cybersecurity intervention; (3) heightened cybersecurity awareness among EV users leads to increased EVSP profits and cybersecurity investments across all providers, highlighting positive spillover effects where enhanced cybersecurity investment by one EVSP can indirectly benefit the entire network. These results offer actionable insights for designing regulatory frameworks, market-based incentives, and collaborative cybersecurity mechanisms that can strengthen the resilience of interconnected EVCNs.