Privacy-Preserving Data Aggregation for Smart Meters under a Mixed Adversary Model

Abstract:
Smart meters are increasingly deployed in modern power systems to enable fine-grained monitoring, demand response, and efficient grid management. While smart meters benefit both utilities and consumers, the fine-grained insights they provide into household behavior raise significant privacy concerns. To mitigate these risks, we introduce a protocol for privacy-preserving aggregation of high-resolution usage data across multiple smart meters. In particular, we seek to use the smart meters themselves to aggregate data rather than making them available to a central authority. At the same time, our protocol accounts for unresponsive or malicious smart meters, introducing mechanisms for exception handling and detecting anomalous reports while preserving competitive performance.