A Randomized Response Model for Privacy Preserving Smart Metering Academic Article uri icon

abstract

  • The adoption of smart meters may bring new privacy concerns to the general public. Given the fact that metering data of individual homes/factories is accumulated every 15 minutes, it is possible to infer the pattern of electricity consumption of individual users. In order to protect the privacy of users in a completely de-centralized setting (i.e., individuals do not communicate with one another), we propose a novel protocol, which allows individual meters to report the true electricity consumption reading with a pre-determinted probability. Load serving entities (LSE) can reconstruct the total electricity consumption of a region or a district through inference algorithm, but their ability of identifying individual users' energy consumption pattern is significantly reduced. Using simulated data, we verify the feasibility of the proposed method and demonstrate performance advantages over existing approaches.

author list (cited authors)

  • Wang, S., Cui, L., Que, J., Choi, D., Jiang, X., Cheng, S., & Xie, L. e.

citation count

  • 51

publication date

  • May 2012