Achieving continuous privacy-preserving histogram query in smart grid communications
University of New Brunswick
Privacy has been taken very seriously in recent times with the introduction of the General Data Protection Regulation (GDPR) by the European Union (EU) and the addition of new rules to the existing Personal Information Protection and Electronic Documents Act (PIPEDA act) by the Canadian government. Governments are strengthening their stance on privacy by ensuring that organizations respect individuals' privacy rights. As such privacy within the smart grid in terms of usage data of customers must be treated with utmost importance. It is in this vein that this research is embarked on, a thesis which involves achieving a continuous histogram query in smart grid communications in a privacy preserving manner. Specifically, this research first gives a brief description of the smart grid from the aspects of characteristics for smart grid design, architecture of the smart grid, advantages and challenges of smart grid, current research focus in smart grid, security and privacy issues in smart grid communications and related works on privacy-preserving smart grid. Then, we employ Paillier Homomorphic Encryption to propose a continuous privacy-preserving histogram query scheme for secure smart grid communications, which can generate a histogram for a user-specified time period while preserving the privacy of residential users. The proposed scheme presents residential users' electricity usage data to the control center without violating their privacy. It does this by presenting all the users' electricity data into two forms of histogram data. The first form is the sum of each class of data, which sums up all the electricity usage data within a particular range and presents it to the control center. The second is the count of each class of data, which counts all the electricity usage data that has been added within a particular range and presents to the control center. Our scheme contains three phases, i.e., Report Generation, Report Aggregation and Array Recovery phase. We analyze the security of each phase and evaluate its performance. The results show that our scheme is privacy-preserving and efficient. Especially, the average time consumption for each of the phases is less than 20 ms in our evaluation.