Privacy-friendly appliance load scheduling in smart grids

Abstract

Abstract-The massive integration of renewable energy sources into the power grid ecosystem with the aim of reducing carbon emissions must cope with their intrinsically intermittent and unpredictable nature. Therefore, the grid must improve its capability of controlling the energy demand by adapting the power consumption curve to match the trend of green energy generation. This could be done by scheduling the activities of deferrable electrical appliances. However, communicating the users' needs about the usage of the electrical appliances leaks sensitive information about habits and lifestyles of the customers, thus arising privacy concerns. This paper proposes a privacypreserving framework to allow the coordination of energy consumption without compromising the privacy of the users: the service requests generated by the domestic appliances are diveded in crypto-shares using Shamir Secret Sharing scheme and collected through an anonymous routing protocol based on Crowds by a set of schedulers, which schedule the requests operating directly on the shares. We discuss the security guarantees provided by our proposed infrastructure and evaluate its performance, comparing it with the optimal scheduling obtained through an Integer Linear Programming formulation.

Privacy-enhanced bi-directional communication in the Smart Grid using trusted computing

“…Rottondi and Verticale [7] have proposed the use of Multi-Party Computation (MPC) to facilitate privacy-friendly appliance load-scheduling. Although it addresses a similar problem, their architecture is designed for collaborative scheduling rather than incentivebased DR.…”

“…However, it has been shown that these bids can be used to infer private information about consumers [2][6] [7]. The magnitude and timing of a particular bid could reveal the use of a particular type of system (e.g.…”

Towards Secure Metering Data Analysis via Distributed Differential Privacy

“…Li et al [8] proposed a secure information aggregation scheme for the smart grid. Rottondi et al [9] proposed a privacy-preserving load scheduling scheme to prevent the NILM attack. However, these schemes [7]- [9] require the use of homomorphic encryption (HE) [10]- [12] such as Paillier's cryptosystem [12], which is computationally expensive and far from practical.…”

“…Rottondi et al [9] proposed a privacy-preserving load scheduling scheme to prevent the NILM attack. However, these schemes [7]- [9] require the use of homomorphic encryption (HE) [10]- [12] such as Paillier's cryptosystem [12], which is computationally expensive and far from practical. Yan et al [13] proposed a symmetric encryption (SE) based secure communication scheme for AMI to protect from eavesdropping.…”

“…As recent works in [4]- [6], [9], [13], we consider a similar adversary model as follows: (1) Honest-but-curious controller where the controller follows the designated protocol specification honestly while it is curious to analyze data in the data collector or the concentrator so as to learn additional information besides those obtained for control policy generation. (2) Honest-but-curious user where the user acts in an 'honest' fashion to answer the query for load scheduling, but in a 'curious' fashion to obtain other users' load profiles by eavesdropping the communications or colluding with the untrusted controller.…”