With the widespread deployment of smart energy systems, smart metering devices are playing an increasingly critical role in electricity data collection, transmission, and billing. However, existing smart metering systems still face security challenges such as delayed anomaly detection, insufficient intrusion detection capabilities, imperfect privacy protection mechanisms, and weak distributed data storage security. To this end, this paper introduces a big data-driven approach, combining real-time monitoring, user behavior modeling, and multidimensional feature analysis to build a layered defense architecture and dynamic risk assessment mechanism. Furthermore, it integrates homomorphic encryption, differential privacy, and blockchain technologies to achieve end-to-end security for smart metering systems. Encryption and privacy protection mitigate the risk of data leakage, while distributed storage and fine-grained access control strengthen permission management. Experimental results demonstrate that the optimized smart metering system significantly outperforms traditional systems in overall security performance. The optimized system achieved an overall security score of 91.3, an improvement of 18.8 points compared to the traditional system’s 72.5. System availability increased from 95.1% to 97.8%, and average recovery time was shortened from 120 seconds to 45 seconds, significantly improving system stability and emergency response capabilities.
Zhu et al. (Thu,) studied this question.