Elastic security analysis method for multi-source data in 5G kite private network based on integration of communication, sensing and computation

With the advancement of communication technologies toward 6G, the integration of communication, sensing and computation has become a critical characteristic of next-generation power private networks. In scenarios where 5G kite private networks are deployed in power plant production control areas, challenges including difficulties in cross-domain association of multi-source data, poor adaptability in dynamic association rule mining, and insufficient real-time security protection have been identified. To address these issues, an elastic security analysis method for multi-source data oriented to 5G-Advanced dvanced integrated sensing-communication-computation private networks was proposed. This method leveraged the 5G private network as the communication foundation, with perceptual data collected by sensors being integrated to construct a multi-source data fusion processing framework, enabling computational analysis of cross-domain data. By combining Spearman rank correlation analysis with time-series FP-Growth algorithms and incorporating an exponential decay time-weighted model, security risk association rules with time constraints were effectively mined. Simulation verification shows that this method significantly improves rule coverage for existing threats while effectively reducing the false alarm rate. The method well adapts to the dynamic data characteristics and integrated sensing-communication-computation synergy of 5G kite private networks, providing technical support for the secure and continuous operation of core production services in the power industry.