Fine-grained smart-meter consumption data are essential for real-time-tariff billing, but their collection can reveal sensitive information about customers' consumption patterns. To address this privacy challenge, we propose a lightweight privacy-preserving smart metering protocol for real-time tariff billing with dynamic tariff policy adjustment. Our scheme employs a data perturbation mechanism with a tariff-weighted zero-sum property, allowing the utility provider to compute the exact customer bill from perturbed consumption readings. The protocol further supports tariff policy adjustments after the consumption readings have been reported. For proportional (Type I) adjustments, previously reported noisy readings are reused without any additional smart-meter report. For non-proportional (Type II) adjustments, the first $L-1$ noisy readings (where L is the number of intervals) are reused and only the final corrected noisy consumption value is updated, reducing additional communication, computation, and storage resources. The number of accepted Type II adjustments is bounded to limit additional exact algebraic information introduced by repeated non-proportional tariff changes. We evaluate the scheme in terms of computational, memory, communication, and privacy characteristics. The protocol requires approximately $3.94540$ seconds of execution time for a complete year, and results confirm exact billing to numerical precision. Privacy is evaluated through statistical characterization, reconstruction-based analysis, tariff-adjustment analysis, and Jensen--Shannon divergence. Increasing the evaluated noise scale reduces the target-specific reconstruction value available to the considered attack while increasing the distributional difference between the original and perturbed consumption data.

翻译：暂无翻译