Garg et al. analyzed malicious security in collaborative zk-SNARKs in paper, revealing two traps in existing design templates that could lead to input privacy leaks and proposed countermeasures. Garg等人在论文中分析了协作zk-SNARKs中的恶意安全问题,揭示了现有设计模板中的两个陷阱,可能导致输入隐私泄露,并提出了缓解策略。
Notes
Collaborative zk-SNARK (coSNARKs) design has privacy leakage risks, such as invalid witnesses exposing honest party inputs
Directly applying malicious security compilers (MPC) may render the entire system insecure
Under the assumption of honest majority, traditional compilers' security checks can be omitted to improve efficiency
First implementation of semi-honest MPC protocol to achieve malicious security in non-trivial cases
Optimization scheme reduces proof generation overhead, applicable to various zk-SNARK schemes
Positive results are general and may have applications beyond collaborative zkSNARKs