Paper 2003/232 Original: eprint.iacr.org/2003/232 ↗

The Statistical Zero-knowledge Proof for Blum Integer Based on Discrete Logarithm

Abstract

Blum integers (BL), which has extensively been used in the domain of cryptography, are integers with form $p^{k_1}q^{k_2}$, where $p$ and $q$ are different primes both $\equiv 3\hspace{4pt}mod\hspace{4pt}4$ and $k_1$ and $k_2$ are odd integers. These integers can be divided two types: 1) $M=pq$, 2) $M=p^{k_1}q^{k_2}$, where at least one of $k_1$ and $k_2$ is greater than 1.\par In \cite{dbk3}, Bruce Schneier has already proposed an open problem: {\it it is unknown whether there exists a truly practical zero-knowledge proof for $M(=pq)\in BL$}. In this paper, we construct two statistical zero-knowledge proofs based on discrete logarithm, which satisfies the two following properties: 1) the prover can convince the verifier $M\in BL$ ; 2) the prover can convince the verifier $M=pq$ or $M=p^{k_1}q^{k_2}$, where at least one of $k_1$ and $k_2$ is more than 1.\par In addition, we propose a statistical zero-knowledge proof in which the prover proves that a committed integer $a$ is not equal to 0.\par