Smart meter privacy using a rechargeable battery: Minimizing the rate of information leakage

Abstract

A rechargeable battery may be used to partially protect the privacy of information contained in a household's electrical load profile. We represent the system as a finite state model to make tractable the computation of the rate of information leakage. Specifically, we use a trellis algorithm to estimate the mutual information rate between the battery's input and output loads. We show that stochastic battery policies can leak 26% less information than a so-called best-effort algorithm (that holds the output load constant whenever possible). We finally describe the extension of the technique to more realistic models of the battery system.