Privacy-Preserving Edge Processing in Decentralized Citizen-centric Sensor NetworksMaster Thesis

18 September 2022, by Leonie van der Veen

Abstract:
The number of people living in cities is growing every year. This leads to challenges such as scarcity of natural resources, demographic change and ongoing urbanization, to which smart cities promise a solution. With the help of the growing Internet of Things, these visions could become reality. However, there are no free and open platforms that would allow citizens to make their data available for this use case. SANE is a concept for such a platform where citizens can collect and share their data within a decentralized network. One important aspect of this is the citizen’s privacy. Users only want to share the data that is truly necessary for an intended purpose, not all of the data they collect. This work presents an approach to how data can be reduced in SANE. It considers the challenges of being flexible enough to cover different data types and tasks. At the same time, it must be comprehensible enough to be understood by average citizens so that they can make informed decisions. In particular, it must be ensured for citizens that no more of the data is sent out than was specified and consented to. We implement this as a sequence of preprocessing steps similar to MongoDB aggregation pipelines or SQL operations like SELECT and WHERE. In addition, we built a web application in which the step sequences are presented as simply as possible. In the form of a flow chart, users are guided through the sequence with step-by-step explanations. We evaluated the comprehensibility and usability of our approach with a user study. The results show that for the majority of tasks both users with and without computer science backgrounds were able to understand the given preprocessing step sequences. Furthermore, after a short introduction, developers could also define step sequences themselves.

Supervised by:
Prof. Dr. Janick Edinger, Philipp Kisters