Practical Privacy-Enhancing Technologies and Applications for Distributed Systems

Abstract

This thesis explores practical privacy-enhancing technologies and applications for various distributed systems. The body of the thesis consists of two parts. The first part focuses on the two Sigma-protocol zero-knowledge arguments of knowledge in the discrete logarithm setting:

In Chapter 3, we propose a new type of gas-efficient zero-knowledge rang arguments with a transparent setup, allowing a prover to convince a verifier that a committed value lies in a specified range. Our range arguments achieve sublinear efficiency in both communication and verification for the range [0, 2^N - 1], where N indicates the bit length.

In Chapter 4, we present zero-knowledge polynomial evaluation arguments with a transparent setup, allowing a prover to convince a verifier that two committed values satisfy a public polynomial evaluation. We depict two zero-knowledge protocols, which are optimised for the polynomials y=P(x;D) of lower-degree (3 <= D <= 512) and higher-degree (D > 512), respectively, where D is the polynomial degree. Based on the techniques of Bayer & Groth, the polynomial evaluation arguments yield a non-trivial improvement in the overall efficiency.

The second part is dedicated to three secure distributed systems and applications:

In Chapter 5, we propose a new scheme for cross-chain atomic swap confidential transactions with a transparent setup. As a black-box decentralised mixer, our scheme can offer confidentiality and anonymity of atomic swaps of all kinds of cryptocurrencies across different smart-contract blockchain platforms. The most distinguishing feature from traditional solutions is its modularity. An array of modularised components and elements are devised to cohesively and elegantly work together to achieve atomic swaps in a privacy-preserving manner. The modularity design brings greater flexibility and extensibility, which lays a foundation for more advanced features.

In Chapter 6, we put forward a practically efficient model for securely computing rank-based statistics, e.g., median, percentiles and quartiles, over distributed datasets without leaking individual data privacy. Based on the binary search technique of Aggarwal et al., we present an interactive protocol and a non-interactive protocol in a client-server communication model, where a group of users outsource computations to a set of servers by leveraging their powerful computational and communication resources.

In Chapter 7, we create a generic blockchain-based collaborative consumption system to achieve secure decentralised collaborative consumption against a dishonest majority of adversaries. We propose secure MPC protocols based on the SPDZ secret-sharing framework to achieve off-chain computations of aggregate usage and on-chain cost-sharing for collaborative consumption scenarios, where a group of users can share a service and pay the costs based on their demands in a privacy-preserving manner. Our core technique is a collaborative zero-knowledge argument that allows multiple users to jointly generate single arguments over distributed secrets. The argument aims to prove that a committed secret is consistent with a set of distributed shares among multiple parties. Show more