Fusion Analysis of Intelligent Connected Vehicle Security Events Based on Multidimensional Heterogeneous Data

In response to cybersecurity issues in intelligent connected vehicles, this paper proposes a security event analysis framework based on the fusion of multidimensional heterogeneous data. Multidimensional security event sample library is first built via the integration of automotive security event data from various sources. Moreover, through the application of correlation analysis techniques, security knowledge graphs, and Generative Adversarial Networks (GANs), we have achieved effective identification of complex attack behaviors and anomaly detection in actual vehicle network traffic, significantly enhancing the efficiency and accuracy of vehicle network security monitoring. In addition, this paper employs knowledge graph visualization techniques to create an intuitive representation of the vehicle’s security status and offer early risk warnings.