In response to the dynamic and ever-evolving landscape of network attacks and cybersecurity, this study aims to enhance network security by identifying critical nodes and optimizing resource allocation within budget constraints. We introduce a novel approach leveraging node centrality scores from four widely-recognized centrality measures. Our unique contribution lies in converting these centrality metrics into actionable insights for identifying network attack probabilities, providing an unconventional yet effective method to bolster network robustness. Additionally, we propose a closed-form expression correlating network robustness with node-centric features, including importance scores and attack probabilities. At the core of our approach lies the development of a nonlinear optimization model that integrates predictive insights into node attack likelihood. Through this framework, we successfully determine an optimal resource allocation strategy, minimizing cyberattack risks on critical nodes while maximizing network robustness. Numerical results validate our approach, offering further insights into network dynamics and improved resilience against emerging cybersecurity threats.

为了应对动态且不断变化的网络攻击和网络安全形势，本研究旨在通过识别关键节点并在预算限制内优化资源分配来增强网络安全。我们引入了一种利用四种广泛认可的中心性度量中的节点中心性得分的新颖方法。我们的独特贡献在于将这些中心性指标转化为可操作的见解，以识别网络攻击概率，提供一种非常规但有效的方法来增强网络的稳健性。此外，我们提出了一种将网络鲁棒性与以节点为中心的特征相关联的封闭式表达式，包括重要性得分和攻击概率。我们方法的核心在于开发非线性优化模型，该模型将预测洞察集成到节点攻击可能性中。通过这个框架，我们成功地确定了最佳的资源分配策略，最大限度地减少了关键节点上的网络攻击风险，同时最大限度地提高了网络的稳健性。数值结果验证了我们的方法，提供了对网络动态的进一步见解，并提高了针对新兴网络安全威胁的抵御能力。