Privacy-Preserving Storage in the Fog

Abstract

In recent years cloud storage services have gained much attention and become a commodity to companies and private users. Nevertheless, cloud storage services have some limitations, especially related to privacy, latency, and availability. In this work we propose a distributed storage system which tackles the major limitations of classical cloud storage services. To this end, we design and implement a storage system which combines classical cloud storage services with the approach of fog computing by using resources at the edge of the network. At the core of our system lies a placement strategy which distributes the data to different storage components. Our implementation is based on well-established methods and techniques from information theory and cryptography. Our empirical analysis shows that our system preserves privacy, provides low latency, and offers high availability. Most notably, we reduce the latency by up to 42% in Upload Mode and even by up to 76% in Download Mode compared to a cloud-only solution.