Non-Interactive Zero-Knowledge Proofs for Composite Statements
No ratings
Presented at Crypto 2018 by
The two most common ways to design non-interactive zero-knowledge (NIZK) proofs are based on Sigma protocols and QAP-based SNARKs. The former is highly efficient for proving algebraic statements while the latter is superior for arithmetic representations.Motivated by applications such as privacy-preserving credentials and privacy-preserving audits in cryptocurrencies, we study the design of NIZKs for composite statements that compose algebraic and arithmetic statements in arbitrary ways. Specifically, we provide a framework for proving statements that consist of ANDs, ORs and function compositions of a mix of algebraic and arithmetic components. This allows us to explore the full spectrum of trade-offs between proof size, prover cost, and CRS size/generation cost. This leads to proofs for statements of the form: knowledge of x such that SHA(gx)=y for some public y where the prover’s work is 500 times fewer exponentiations compared to a QAP-based SNARK at the cost of increasing the proof size to 2404 group and field elements. In application to anonymous credentials, our techniques result in 8 times fewer exponentiations for the prover at the cost of increasing the proof size to 298 elements. From Laconic Zero-Knowledge to Public-Key CryptographyItay Berman, Akshay Degwekar, Ron D. Rothblum, Prashant Nalini Vasudevanhttps://youtube.com/watch?v=XAyAP2QguIIhttps://crypto.iacr.org/2018/slides/28803.pdfSince its inception, public-key encryption ( PKE ) has been one of the main cornerstones of cryptography. A central goal in cryptographic research is to understand the foundations of public-key encryption and in particular, base its existence on a natural and generic complexity-theoretic assumption. An intriguing candidate for such an assumption is the existence of a cryptographically hard language Open image in new window.In this work we prove that public-key encryption can be based on the foregoing assumption, as long as the (honest) prover in the zero-knowledge protocol is efficient and laconic. That is, messages that the prover sends should be efficiently computable (given the Open image in new window witness) and short (i.e., of sufficiently sub-logarithmic length). Actually, our result is stronger and only requires the protocol to be zero-knowledge for an honest-verifier and sound against computationally bounded cheating provers.Languages in Open image in new window with such laconic zero-knowledge protocols are known from a variety of computational assumptions (e.g., Quadratic Residuocity, Decisional Diffie-Hellman, Learning with Errors, etc.). Thus, our main result can also be viewed as giving a unifying framework for constructing PKE which, in particular, captures many of the assumptions that were already known to yield PKE .We also show several extensions of our result. First, that a certain weakening of our assumption on laconic zero-knowledge is actually equivalent to PKE , thereby giving a complexity-theoretic characterization of PKE . Second, a mild strengthening of our assumption also yields a (2-message) oblivious transfer protocol.