Foundations of decentralised privacy
Distributed ledgers, and specifically blockchains, have been an immensely popular investment in the past few years. The heart of their popularity is due to their novel approach toward financial assets: They replace the need for central, trusted institutions such as banks with cryptography, ensuring no one entity has authority over the system. In the light of record distrust in many established institutions, this is attractive both as a method to combat institutional control and to demonstrate transparency. What better way to manage distrust than to embrace it? While distributed ledgers have achieved great things in removing the need to trust institutions, most notably the creation of fully decentralised assets, their practice falls short of the idealistic goals often seen in the field. One of their greatest shortcomings lies in a fundamental conflict with privacy. Distributed ledgers and surrounding technologies rely heavily on the transparent replication of data, a practice which makes keeping anything hidden very difficult. This thesis makes use of the powerful cryptography of succinct non-interactive zero-knowledge proofs to provide a foundation for re-establishing privacy in the decentralised setting. It discusses the security assumptions and requirements of succinct zero-knowledge proofs atlength, establishing a new framework for handling security proofs about them, and reducing the setup required to that already present in commonly used distributed ledgers. It further demonstrates the possibility of privacy-preserving proof-of-stake, removing the need for costly proofs-of-work for a privacy-focused distributed ledger. Finally, it lays out a solid foundation for a smart contract system supporting privacy – putting into the hands of contract authors the tools necessary to innovate and introduce new privacy features.