Semantic modeling of cyber threats in the energy sector using Dynamic Cognitive Maps and Bayesian Belief Network

Abstract

The article discusses the use of semantic modeling in the analysis of threats to energy security (ES). Semantic modeling is proposed to be applied at a qualitative level, followed by quantitative assessment of the ES level in studies of energy security. Exercise of traditional software systems provides a quantitative assessment, which is characterized by the duration of information preparation, and the formation and adjustment of large enough models for computational experiments. At the first level, a decision maker selects options for which a detailed rationale is required based on the results of semantic modeling. These options are calculated at the second level. The article presents basic notions of Dynamic Cognitive Maps (DCM) and Bayesian Belief Network (BBN). The paper presents the information model that is suggested to use in analyzing cyber threats in the energy sector. Exemplification of the impact of cyber threats on an energy facility carried out under Dynamic Cognitive Maps and Bayesian Belief Network is given in this article. The advantages of each tool and their role in analyzing cyber threats in the energy sector are presented.

A Systematic Review of Smart City Infrastructure Threat Modelling Methodologies: A Bayesian Focused Review

“…One of these is Gaber et al [55] threat modelling methodology for joint-estimation detection of cyberattacks within the smart grid, using both Bayesian and Neyman-Pearson optimum test methods, expanding upon a previous research methodology by Tajer et al [42], with the advantages of checking the decision rules regarding the BN model providing absolute values. Another proposal is Gaskova and Massel [36] which applies dynamic cognitive maps alongside Bayesian networks for analysing cyber-threats to understand the relationships across the identified weights in respective time-scales. Expanding upon this by including DBN application after designing the cognitive maps could help maintain the flow of time throughout the system modelling and a more powerful metric transition.…”

“…They break down systems different from BN approaches and are structured much more different, such as Zografopoulos et al [2], Zhou et al [5], Caviglione and Coccoli [29], which demonstrate smart grids and other critical infrastructures different towards similar goals. These alternative options can be applied alongside a Bayesian-based approach to creating a new structured methodology to assist in understanding these systems and compensate and improve all aspects regarding the ToE [36]. Cyber threats are used corresponding within the model to test cyber-attacks on how the proposed framework could best understand the systems and walk through these attacks on how they influence the modelled nodes and values.…”

Dynamic Fuzzy Cognitive Models as a Means of Visual Analytics and Cognitive Graphics

“…A dynamic cognitive map differs from a regular cognitive map in that it is a bunch of cognitive maps that reflect the model at different points in time [11] (1):…”