Liu Lanying, Du Ning, Sheng Duyong
Liaocheng University Dongchang College, Liaocheng, 252000, Shandong, China.
College of Electrical Engineering and Automation, Shandong University of Science and Technology, Qingdao, 266590, Shandong, China.
Sci Rep. 2025 May 4;15(1):15568. doi: 10.1038/s41598-025-00360-4.
Ensuring network security in complex and dynamic environments has become a critical challenge due to the increasing proliferation of Internet of Things (IoT) devices and decentralized architectures such as Fog Computing. Traditional node identification methods primarily focus on either network centrality measures or security metrics in isolation, which limits their effectiveness in detecting security-critical nodes. In this paper, we propose a novel Security-Centric Node Identification method that integrates multiple centrality measures with security-related metrics and dynamic factors to compute a comprehensive Security Centrality (SC) for each node. Unlike conventional approaches, our method accounts for both structural importance and security vulnerabilities by incorporating degree, betweenness, closeness, and eigenvector centralities with real-time security risk assessments and dynamic network conditions. To achieve this, we develop a mathematical model to compute SC, introduce efficient algorithms for identifying critical nodes, and implement an incremental update mechanism to enhance adaptability in real-time networks. Our experimental evaluation on various network topologies, including random, scale-free, small-world, and real-world networks, demonstrates that the proposed method effectively identifies security-critical nodes with high detection accuracy while maintaining a low false-positive rate. The results show that incorporating dynamic factors significantly improves the robustness of node identification, making our method highly adaptable to real-world network security scenarios.
由于物联网（IoT）设备的不断扩散以及诸如雾计算等去中心化架构的出现，在复杂且动态的环境中确保网络安全已成为一项严峻挑战。传统的节点识别方法主要孤立地关注网络中心性度量或安全指标，这限制了它们在检测安全关键节点方面的有效性。在本文中，我们提出了一种新颖的以安全为中心的节点识别方法，该方法将多种中心性度量与安全相关指标及动态因素相结合，为每个节点计算一个综合的安全中心性（SC）。与传统方法不同，我们的方法通过将度中心性、介数中心性、紧密中心性和特征向量中心性与实时安全风险评估及动态网络条件相结合，兼顾了结构重要性和安全漏洞。为实现这一点，我们开发了一个用于计算SC的数学模型，引入了用于识别关键节点的高效算法，并实现了一种增量更新机制以增强在实时网络中的适应性。我们在各种网络拓扑结构上进行的实验评估，包括随机网络、无标度网络、小世界网络和真实世界网络，表明所提出的方法能够以高检测准确率有效识别安全关键节点，同时保持低误报率。结果表明，纳入动态因素显著提高了节点识别的鲁棒性，使我们的方法高度适用于现实世界的网络安全场景。
