“Hello? Is There Anybody in There?” Leakage Assessment of Differential Privacy Mechanisms in Smart Metering Infrastructure

Smart meters provide fine-grained power usage profiles of consumers to utility providers to facilitate various grid functionalities such as load monitoring, real-time pricing, etc. However, information leakage from these usage profiles can potentially reveal sensitive aspects of consumers’ daily routines and their home absence, as state-of-the-art metering strategies lack adequate security and privacy measures. Among various privacy-preserving mechanisms, Differential Privacy (DP) is widely adopted in the literature due to its solid mathematical foundation. Nevertheless, the privacy-utility trade-off problem in smart metering systems limits the amount of privacy protection various instances of DP mechanisms can provide. We demonstrate that the constraints imposed by the privacy-utility trade-off make it possible to launch empirical statistical attacks on the differential private metering data. In this paper, we propose a novel statistical methodology, constructed using the principles of t-test based hypothesis testing, to discover the absence of a consumer in their household upon observing real-time differentially private output traces of sensitive meter readings over successive sampling windows. Additionally, we formally establish that this trade-off is an inherent characteristic of the smart metering problem, implying that any mechanism adhering to this trade-off is susceptible to our attack. We conduct an extensive experimental evaluation using a real-world metering dataset to validate our proposed methodology. We evaluate our scheme against six state-of-the-art DP mechanisms employed in metering infrastructure. Our results demonstrate that the proposed approach attains a success rate exceeding \(90\%\) within a mere six-hour observation interval, highlighting its effectiveness in revealing vulnerabilities within established DP implementations.