【作者】 杨萱；

【导师】 余昭平；

【作者基本信息】 解放军信息工程大学 ， 应用数学， 2009， 硕士

【摘要】 安全多方计算是指在一个互不信任的多用户网络中,两个或多个用户能够在不泄漏各自输入信息时协同合作执行某项计算任务。自A.C.Yao于1982年首次提出安全两方计算的概念以来,安全多方计算已经成为国际密码学界的研究热点。实际上,如果我们能安全的计算任何函数,我们就掌握了一个强大的工具,因为任何一个密码协议都可以划归为一个特殊的安全多方计算协议。传统的安全多方计算主要关注如何获得一般化的可计算任意函数的协议,而对安全多方计算在具体环境下的应用并没有做深入的研究。本文旨在针对具体的应用问题和安全性要求设计出适合的安全多方计算协议。主要的研究工作如下:1.针对代理签名只能提供授权的认证而不能提供保密性这一问题,研究了安全多方计算的一个典型应用——代理签密。本文利用双线性配对,提出了一个基于身份的代理签密方案,并对方案的安全性和效率进行了分析讨论。2.研究了有关原始签名者的隐私保护问题。代理盲签名是结合盲签名和代理签名的优点所形成的具有特殊用途的数字签名。针对现有大多数代理盲签名方案效率较低的问题,提出了一个高效的代理盲签名方案,方案的安全性基于离散对数困难问题。3.针对现有大多数优化公平交换协议在交换阶段效率较低的问题,运用变色龙哈希函数的特殊性质,给出了一种构造优化公平交换协议的新方法。新的优化公平交换协议在交换阶段不需要复杂的零知识证明系统,在保证安全性及公平性的同时,有效的避免了大量的计算和通信过程。4.对特殊安全多方计算中的保护私有信息的协作线性方程组求解问题和保护私有信息的多项式插值问题进行了研究。通过分析,我们发现,保护私有信息的多项式插值问题可以化归为一个安全两方保护私有信息的协作线性方程组求解问题。基于此,我们首先提出了一个高效的安全两方保护私有信息的协作线性方程组求解协议,并由此设计了一个保护私有信息的多项式插值协议,有效的解决了保护私有信息的多项式插值问题。更多还原

【Abstract】 Secure multiparty computation refers to the problem where two or more parties want to jointly compute a task based on their private inputs, while no party is willing to disclose his privacy to any other one. Since the problem of secure two-party computation was firstly introduced by A. C. Yao in 1982, the research of secure multiparty computation has become one of the focuses in international cryptographic fields. It should be clear that we have a very powerful tool if we can compute any function securely, because virtually all cryptographic protocols are, or can be rephrased to be, special cases of the multiparty computation problems.Though the traditional secure multiparty computation protocols mainly focus on how to acquire the general protocols which can calculate arbitrary functions, the applications of secure multiparty computation in concrete environments have not been deeply researched. In this paper, we will design suitable secure multiparty computation protocols facing concrete applications and security requirements.To sum up, the works and innovations of this thesis could be summarized as follows:1. Since a proxy signature only provides the delegated authenticity and doesn’t provide the confidentiality, we propose a new identity-based proxy signcryption scheme from bilinear pairings and analyze its security and efficiency.2. Research on the problem of proxy signer’s privacy protection. Proxy blind signature, which combines the properties of both proxy signature and blind signature, is useful in many applications. We present a proxy blind signature scheme based on discrete logarithm problem. As compared with existing typical schemes, our scheme achieves higher efficiency.3. Most protocols in current literature for optimistic fair exchange protocols are quite inefficient in the exchange phase. We present a novel method for constructing efficient and secure optimistic fair exchange protocols using a key-exposure-free chameleon hashing scheme. The proposed optimistic fair exchange protocol has no use for verifiably encrypted signature and requires no zero-knowledge proofs in the exchange phase, which will greatly reduce the communication overhead and managing cost.4. Privacy-preserving cooperative linear system of equations problem and privacy-preserving polynomial interpolation problem are studied. We find that the privacy-preserving polynomial interpolation problem can be transformed into a cooperative linear system of equations problem. Therefore, we present a secure and efficient two-party privacy-preserving cooperative linear system of equations protocol and analyze its security. Then, we propose a privacy-preserving polynomial interpolation protocol based on the newly devised secure two-party privacy-preserving cooperative linear system of equations protocol. We also analyze the new protocol and prove its correctness and security.更多还原