Understanding Differential Privacy in Non-Intrusive Load Monitoring

Abstract

Smart meter devices enable the system operator to better understand the demand at the potential risk of private information leakage. One promising solution to mitigate such risk is to inject noises into the meter data to achieve certain level of differential privacy. In this paper, we cast the non-intrusive load monitoring (NILM) as a compressive sensing problem, and then seek to characterize the physical meaning of the parameters in ϵ-differential privacy in terms of the performance guarantee for NILM inference.