A lightweight privacy-preserved spatial and temporal aggregation of energy data

Smart grid provides fine-grained real time energy consumption, and it is able to improve the efficiency of energy management. It enables the collection of energy consumption data from consumer and hence has raised serious privacy concerns. Energy consumption data, a form of personal information that reveals behavioral patterns can be used to identify electrical appliances being used by the user through the electricity load signature, thus making it possible to further reveal the residency pattern of a consumer's household or appliances usage habit. This paper proposes to enhance the privacy of energy consumption data by enabling the utility to retrieve the aggregated spatial and temporal consumption without revealing individual energy consumption. We use a lightweight cryptographic mechanism to mask the energy consumption data by adding random noises to each energy reading and use Paillier's additive homomorphic encryption to protect the noises. When summing up the masked energy consumption data for both Spatial and Temporal aggregation, the noises cancel out each other, hence resulting in either the total sum of energy consumed in a neighbourhood at a particular time, or the total sum of energy consumed by a household in a day. No third party is able to derive the energy consumption pattern of a household in real time. A proof-of-concept was implemented to demonstrate the feasibility of the system, and the results show that the system can be efficiently deployed on a low-cost computing platform.