Cluster Computing ( IF 4.4 ) Pub Date : 2020-05-08 , DOI: 10.1007/s10586-020-03113-2 Ouafae Kasmi , Amine Baina , Mostafa Bellafkih
Infrastructure interdependency is a bidirectional interconnection between entities of two infrastructures. These Critical Infrastructures (CIs) suffer from several attacks, vulnerabilities, and failures. Indeed a failure in one CI could lead to serious consequences on physical security, economic security, or public health. However, the protection of these infrastructures is essential. The clustering algorithm is considered as one of the best interesting solutions to reduce its impacts. This paper presents a new approach of the Fuzzy Logic-based clustering algorithm to better identify and understand the overall interconnections between entities in CI. The Fuzzy Logic based on the clustering algorithm is split into Cluster heads (CHs) election and Cluster Members formation (CMs) election. The CH is elected by quantifying the degree of dependency of each component and CM is elected by determining their criticality levels using Failure Mode and Effect Analysis method to determine their Number Priority of Risk. The simulation results demonstrate that by adopting our proposed approach, improved management in CIs is gained not only in enhancing the degree of inter/dependency but also in identifying the criticality of interdependencies, minimizing Round Time Trip of failures nodes detection and reduce uncertainty risks.
中文翻译:
基于模糊逻辑的关键基础设施管理聚类算法
基础架构相互依赖性是两个基础架构的实体之间的双向互连。这些关键基础架构(CI)遭受多种攻击,漏洞和故障。实际上,一个配置项的失败可能会对人身安全,经济安全或公共健康造成严重后果。但是,保护这些基础结构至关重要。聚类算法被认为是减少其影响的最有趣的解决方案之一。本文提出了一种基于模糊逻辑的聚类算法的新方法,可以更好地识别和理解CI中实体之间的整体互连。基于聚类算法的模糊逻辑分为簇首(CH)选举和簇成员形成(CM)选举。通过量化每个组件的依赖程度来选择CH,并通过使用故障模式和效果分析方法确定它们的风险优先级来确定它们的关键程度来选择CM。仿真结果表明,采用我们提出的方法,不仅可以提高相互依赖程度,而且可以确定相互依赖的关键程度,最大程度地减少故障节点检测的往返时间,并降低不确定性风险,从而改善了CI的管理。