| dc.description.abstract | Smart meter data can be used by the utility for billing the consumers timely along with power theft detection, demand side management, distribution network planning, consumer segmentation, load forecasting, fault detection, outage management and restoration, etc. Owing to many possible applications of smart meter data, the utilities are replacing traditional energy meters with smart meters worldwide. However, mining of smart meter data may result in leakage of private information of consumers. As a result, privacy concerns may inhibit consumers from adopting smart meters and sharing smart meter data with the utility. To address consumers' privacy concerns, privacy preserving algorithms have been proposed in the literature. Although privacy preserving algorithms can reduce private information leakage of a consumer, they cannot preserve the data usability. Privacy preserving algorithms tend to reduce the usefulness of the data by limiting the information that can be extracted from the data. So, there is a need for a smart meter data sharing method that can preserve data usability.
Data usability can be preserved if consumers share their data without applying any privacy preserving techniques. However, such a way of sharing data may result in privacy loss of consumers. Offering monetary compensation to the consumers for their privacy loss can be an alternate approach to encourage smart meter data sharing. The advantage of this approach is that it does not reduce the data usability. Hence, in this thesis, a novel privacy pricing framework is proposed that can be used by the utility to determine the price of privacy of smart meter data. The determined price can be paid to the consumers to compensate them for their privacy loss.
Estimation of price of privacy requires quantification of the private information content of smart meter data of the consumers. Hence, a mathematical privacy model is proposed in this thesis for the quantification of privacy of smart meter energy data of a consumer. At first, the possible applications of privacy model of smart meter data are presented. Then, the notion of smart meter data privacy is discussed. Based on this notion, a mathematical model is proposed to quantify the privacy of a time series of energy data of a consumer. The price of privacy of consumers' smart meter data is determined under a given budget of the utility by solving a nonlinear convex optimization problem formulated using the proposed privacy model. A constraint is incorporated in the optimization problem to ensure that the privacy price does not exceed a certain portion of the electricity bill paid by the consumer. Mathematical analysis is performed to shed light on the choice of various hyperparameters associated with the proposed privacy pricing framework (including the utility's budget). The robustness of the proposed privacy pricing framework is illustrated mathematically. The proposed privacy pricing framework can incentivize the consumers to give up their privacy right over their smart meter data. As a result, usability of smart meter data will be maintained.
Performance of the proposed privacy pricing framework is evaluated using practical smart meter data. At first, the privacy of practical smart meter data is quantified using the proposed privacy model. The impacts of metering interval and data length on the privacy content of practical smart meter data are analyzed. Then, the quantified privacy information is used for finding the privacy price. Results show that the proposed privacy pricing framework tries to pay a higher compensation to a consumer incurring a higher privacy loss. In addition, it guarantees that the privacy price will not exceed a certain portion of the energy bill paid by a consumer. Thus, the pricing framework is fair to both the utility and the consumers. The impacts of the utility's budget (allocated for privacy pricing), the metering interval of smart meters and the number of participating consumers on the price of privacy are also demonstrated using practical smart meter data. | en_US |
