Towards Practical Privacy-Preserving Protocols

Protecting users' privacy in digital systems becomes more complex and challenging over time, as the amount of stored and exchanged data grows steadily and systems become increasingly involved and connected. Two techniques that try to approach this issue are Secure Multi-Party Computation (MPC) and Private Information Retrieval (PIR), which aim to enable practical computation while simultaneously keeping sensitive data private. In this thesis we present results showing how real-world applications can be executed in a privacy-preserving way. This is not only desired by users of such applications, but since 2018 also based on a strong legal foundation with the General Data Protection Regulation (GDPR) in the European Union, that forces companies to protect the privacy of user data by design. This thesis' contributions are split into three parts and can be summarized as follows: MPC Tools Generic MPC requires in-depth background knowledge about a complex research field. To approach this, we provide tools that are efficient and usable at the same time, and serve as a foundation for follow-up work as they allow cryptographers, researchers and developers to implement, test and deploy MPC applications. We provide an implementation framework that abstracts from the underlying protocols, optimized building blocks generated from hardware synthesis tools, and allow the direct processing of Hardware Definition Languages (HDLs). Finally, we present an automated compiler for efficient hybrid protocols from ANSI C. MPC Applications MPC was for a long time deemed too expensive to be used in practice. We show several use cases of real-world applications that can operate in a privacy-preserving, yet practical way when engineered properly and built on top of suitable MPC protocols. Use cases presented in this thesis are from the domain of route computation using BGP on the Internet or at Internet Exchange Points (IXPs). In both cases our protocols protect sensitive business information that is used to determine routing decisions. Another use case focuses on genomics, which is particularly critical as the human genome is connected to everyone during their entire lifespan and cannot be altered. Our system enables federated genomic databases, where several institutions can privately outsource their genome data and where research institutes can query this data in a privacy-preserving manner. PIR and Applications Privately retrieving data from a database is a crucial requirement for user privacy and metadata protection, and is enabled amongst others by a technique called Private Information Retrieval (PIR). We present improvements and a generalization of a well-known multi-server PIR scheme of Chor et al., and an implementation and evaluation thereof. We also design and implement an efficient anonymous messaging system built on top of PIR. Furthermore we provide a scalable solution for private contact discovery that utilizes ideas from efficient two-server PIR built from Distributed Point Functions (DPFs) in combination with Private Set Intersection (PSI)