Visible to the public Security Assessment of Enterprise Networks Based on Analytic Network Process and Evidence Theory

TitleSecurity Assessment of Enterprise Networks Based on Analytic Network Process and Evidence Theory
Publication TypeConference Paper
Year of Publication2021
AuthorsLv, Huiying, Zhang, Yuan, Li, Huan, Chang, Wenjun
Conference Name2021 3rd International Conference on Artificial Intelligence and Advanced Manufacture (AIAM)
Date Publishedoct
Keywordsanalytic network process, Analytical models, enterprise network, evidence theory, expert systems, Fuses, Human Behavior, Knowledge engineering, Network analyzers, Network security, Physical layer, pubcrawl, resilience, Resiliency, Scalability, security, security assessment
Abstract

Network security has always been the most important of enterprise informatization construction and development, and the security assessment of network system is the basis for enterprises to make effective security defense strategies. Aiming at the relevance of security factors and subjectivity of evaluation results in the process of enterprise network system security assessment, a security assessment method combining Analytic Network Process and evidence theory is proposed. Firstly, we built a complete security assessment index system and network analysis structure model for enterprise network, and determined the converged security index weights by calculating hypermatrix, limit hypermatrix and stable limit hypermatrix; then, we used the evidence theory on data fusion of the evaluation opinions of multiple experts to eliminate the conflict between evidences. Finally, according to the principle of maximum membership degree, we realized the assessment of enterprise network security level using weighted average. The example analysis showed that the model not only weighed the correlation influence among the security indicators, but also effectively reduced the subjectivity of expert evaluation and the fuzziness and uncertainty in qualitative analysis, which verified the effectiveness of the model and method, and provided an important basis for network security management.

DOI10.1109/AIAM54119.2021.00069
Citation Keylv_security_2021