Privacy of smart meter systems with an alternative energy source

Smart meter (SM) measurements provide near realtime information on the electricity consumption of a user to the utility provider (UP). This data can be used to extract private information on the energy consumption patterns of the user. Assuming that the user has access to an alternative energy source (AES) in addition to the power grid, SM privacy problem is studied from an information theoretic perspective. The energy requirement of the user (input load) at each time instant can be satisfied either from the power grid (output load) or from the AES. It is assumed that the output load can be perfectly tracked by the UP, and the privacy is measured through the information leakage rate. For given average and peak power constraints on the AES, privacy-power function is defined, and its equivalence to the rate-distortion function with a difference distortion measure is shown. Focusing on continuous input loads, the privacy-power function is characterized when there is only peak power limitation on the AES, while the Shannon lower bound is provided for the general case. The bound is shown to be achievable for the exponential input distribution.