No peeking: privacy-preserving demand response system in smart grids

Abstract

Demand response (DR) programs are widely used to balance the supply and demand of electricity in a smart grid. This results in a reliable electric power system. Unfortunately, privacy violation becomes a pressing challenge that drastically affects the DR programs. Power usage and operational data can be abused to infer personal information of customers. Without a well-designed privacy preservation mechanism, adversaries can capture, model and divulge customers' behaviour and activities. In this paper, we first investigate the natures of privacy leakages and explore potential privacy threat models. After that, we design and implement a new protocol named privacy-preserving demand response based on the attributed-based encryption, and formally prove its validity. To demonstrate its viability, the protocol is adopted in several types of DR programs on an emulated smart grid platform. Experimental results show substantially lighter overheads while formidable privacy challenges are addressed.