As organizations struggle with processing vast amounts of information, outsourcing sensitive data to third parties becomes a necessity. To protect the data, various cryptographic techniques are used in outsourced database systems to ensure data privacy, while allowing efficient querying. A rich collection of attacks on such systems has emerged. Even with strong cryptography, just communication volume or access pattern is enough for an adversary to succeed.

In this work we present a model for differentially private outsourced database system and a concrete construction, $\mathcal{E}\text{psolute}$, that provably conceals the aforementioned leakages, while remaining efficient and scalable. In our solution, differential privacy is preserved at the record level even against an untrusted server that controls data and queries. $\mathcal{E}\text{psolute}$ combines Oblivious RAM and differentially private sanitizers to create a generic and efficient construction.

We go further and present a set of improvements to bring the solution to efficiency and practicality necessary for real-world adoption. We describe the way to parallelize the operations, minimize the amount of noise, and reduce the number of network requests, while preserving the privacy guarantees. We have run an extensive set of experiments, dozens of servers processing up to 10 million records, and compiled a detailed result analysis proving the efficiency and scalability of our solution. While providing strong security and privacy guarantees we are less than an order of magnitude slower than range query execution of a non-secure plain-text optimized RDBMS like MySQL and PostgreSQL.

This work has also been presented to Brown Security Reading Group. See presentation video.

In permissioned blockchain systems, participants are admitted to the network by receiving a credential from a certification authority. Each transaction processed by the network is required to be authorized by a valid participant who authenticates via her credential. Use case settings where privacy is a concern thus require proper privacy-preserving authentication and authorization mechanisms.

Anonymous credential schemes allow a user to authenticate while showing only those attributes necessary in a given setting. This makes them a great tool for authorizing transactions in permissioned blockchain systems based on the user's attributes. As in most setups of such systems where there is one distinct certification authority for each organization in the network, the use of plain anonymous credential schemes still leaks the association of a user to her issuing organization. Camenisch, Drijvers and Dubovitskaya (CCS 2017) therefore suggest the use of a delegatable anonymous credential scheme to also hide that remaining piece of information.

In this paper we improve the Camenisch et al. scheme and extend it with revocation and auditability; two functionalities that are necessary for real-world adoption. We present a complete protocol and provide its production-grade open-source implementation including the scheme and the proposed extensions, ready to be integrated with Hyperledger Fabric. Our distributed-setting performance measurements show that the integration of the scheme with Hyperledger Fabric, while incurring an overhead in comparison to the less privacy-preserving solutions, is practical for settings with stringent privacy requirements.

This work has started while on internship at IBM Zurich Research Lab. See official DAC scheme source code on IBM's github.

Database query evaluation over encrypted data has received a lot of attention recently. Order Preserving Encryption (OPE) and Order Revealing Encryption (ORE) are two important encryption schemes that have been proposed in this area. These schemes can provide very efficient query execution but at the same time may leak some information to adversaries. In this paper, we present the first comprehensive comparison among a number of important OPE and ORE schemes using a framework that we developed. We evaluate protocols that are based on these schemes as well. We analyze and compare them both theoretically and experimentally and measure their performance over database indexing and query evaluation techniques using not only execution time but also IO performance and usage of cryptographic primitive operations. Our comparison reveals some interesting insights concerning the relative security and performance of these approaches in database settings. Furthermore, we propose a number of improvements for some of these scheme and protocols. Finally, we provide a number of suggestions and recommendations that can be valuable to database researchers and users.

This work has received the "Most Reproducible Paper" award for exemplary work on experiments, in particular, making them easily reproducible. See award page and the certificate.