The Internet of Things (IoT), much more than other fields in computer science (CS), introduces new issues related to inferential and differential privacy. The growing use and sophistication of IoT deployment generates increasing volumes of inferential data about individuals and creates new challenges to privacy law assessment. Further, inferential privacy within IoT ecosystems is actually implied by differential privacy when data is independent, but could also be differential when data correlates.

The authors successfully discover the problem of inferential privacy when IoT infrastructures are in use, especially through various models of smart grid networks. They also challenge the current research on data privacy in smart grid and IoT infrastructures. These studies provide novel mechanisms for protecting the collected data: anonymization and aggregation via a differential privacy approach.

It is interesting how the authors define the utility-privacy tradeoff of the IoT infrastructures using a two-tier approach: model the tradeoff between the data collection process and IoT device performance, and acquire knowledge about the tradeoff between how much data is collected and the use of personal (private) information in the processing of collected data through the IoT infrastructure. Naturally, this is not only a technical area problem; data protection (privacy law) should also be included, defining what data is actually private and what data must not be included in smart grid infrastructure operations providing utility services.

The authors clearly describe these issues through electric utilities based on smart grid technologies. They try to discover scientific principles on which further research could be processed and to provide basic propositions of privacy issues for the real-world deployment of utility services through the IoT infrastructure. Thus, they present a utility-privacy framework with a specific privacy-preserving mechanism, finding the proper relation between the data collection volume and the needed level of functionality of the IoT infrastructure on which utility services are providing to the customer.

The authors also effectively discuss two aspects of data: utility and privacy. These notions are related to the study of inferential privacy, a subject that is more applicable to smart-grid-based utility services. It is broadly defined through a utility-privacy framework, describing direct load control (DLC) programs with the presented model and simulations. The paper includes privacy analysis using a DLC program model based on data collected from real-world utility services deployment.

This study provides a novel approach to the privacy issues within the IoT infrastructure, especially within smart-grid-based utility services. Thus, utility privacy is the focus of this excellent paper. It would be of interest to researchers in the field of privacy who are challenged by the new technology, as well as technicians working on smart-grid-based utility services. It is also a valuable resource for electrical engineering and CS libraries.