Using differential privacy for the internet of things

Abstract

In this paper we propose a hybrid privacy-protection model for the Internet of Things (IoT) with the ultimate purpose of balancing privacy restrictions and usability in data delivery services. Our model uses traditional de-identification methods (such as k-anonymity) under low-privacy requirements, but allows for the transmission of aggregate statistical results (calculated with a privacy-preserving method such as Differential Privacy) as an alternative if the privacy requirements exceed a threshold. We show a prototype implementation for this model, and present a small step-by-step example.