Semantic model of an exploit’s source code for data protection in automation systems

Abstract

Data protection is especially relevant in automation systems nowadays. The paper proposes a semantic model of exploit’s source code as a basis for unknown attacks detection, forecasting and assessment and therefore for automated data protection from cyber attacks. The proposed model combines a control flow graph with a function call dependencies graph. The paper describes model generation technique that is based on the analysis of exploit’s source code and incorporates code compilation, its decompilation, building the functional semantic models of the exploits, and, finally, comparing these models and generating the standard semantic model of exploits’ source code. Application of the technique for generation of exploit’s model is demonstrated on a case study.