Keeping Data Private while Computing in the Cloud

Abstract—The cloud offers unprecedented access to compu-tation. However, ensuring the privacy of that computation re-mains a significant challenge. In this paper, we address the prob-lem of distributing computation onto the cloud in a way that pre-serves the privacy of the computation’s data even from the cloud nodes themselves. The approach, called sTile, separates the com-putation into small subcomputations and distributes them in a way that makes it prohibitively hard to reconstruct the data. We evaluate sTile theoretically and empirically: First, we for-mally prove that sTile systems preserve privacy. Second, we de-ploy a prototype implementation on three different networks, in-cluding the globally-distributed PlanetLab testbed, to show that sTile is robust to network delay and efficient enough to signifi-cantly outperform existing privacy-preserving approaches. I