Research on Dynamic Anomaly Detection Method of Edge Devices Based on Attack Chain Knowledge Graph

As power distribution IoT technology advances, traditional security measures are increasingly inadequate. This study proposes a dynamic anomaly detection method for edge devices using attack chain knowledge graph technology. By analyzing IoT attack paths and constructing an attack chain model, the method establishes a hierarchical network security knowledge extraction model. It automatically builds a security knowledge graph based on the attack chain, enabling effective dynamic anomaly detection and improving security performance. The approach combines graph structure storage and algorithms for fast attack detection and threat traceability. Additionally, it incorporates atlas similarity and critical path analysis for dynamic detection. Experimental results show that this method effectively identifies potential security threats, enhances system security, and meets the complex protection needs of power distribution IoT systems.