Novel Temporal Perturbation-Based Privacy-Preserving Mechanism for Smart Meters

Smart meters provide strong data support for the intelligent construction of power grids but expose users' privacy-sensitive information to adversaries. Thus, several mechanisms have been proposed for smart meter privacy protection. However, these existing mechanisms either lack consideration of the protection for customer power consumption mode or focus only on data security and ignore the data availability for intelligent services of smart meters. In this study, we propose a temporal perturbation-based privacy-preserving mechanism to achieve a balance between privacy security and data availability. The time disturbance model improves privacy security by staggering the acquisition and release time of smart meter data, destructs the load characteristics hidden in the power waveform, and realizes the fuzzification of real consumption events. Moreover, data availability for metering and billing, electronic scheduling and management, and value-added services is guaranteed. We analyze data availability and privacy security from theoretical and experimental perspectives, deduce a data availability analysis model, and introduce a non-intrusive load monitoring algorithm as a testing method for evaluating security performance. Evaluation results show that the proposed scheme performs well in protecting users' consumption pattern and is resistant to attacks by load detection; it also ensures data availability when used for intelligent services.