Abstract
Cloud services provide clients with highly scalable network, storage, and computational resources. However, these service come with the challenge of guaranteeing the confidentiality of the data stored on the cloud. Rather than attempting to prevent adversaries from compromising the cloud server, we aim in this paper to propose a protocol for secure querying in the cloud, while preserving the privacy of the participants and assuming the existence of a passive adversary able to access all data stored in the cloud. In this paper, we address this problem by proposing a network protocol that would allow a third party, such as a health organization, to query privately encrypted data without relying on a trusted entity. The protocol we propose preserves the privacy of the data owners and the querying entity. The protocol relies on homomorphic cryptography, threshold cryptography, differential privacy, and randomization to allow for secure, distributed, and privacy-preserving queries. We evaluate the performance of our protocol and report on the results of the implementation.