【作者】 吕猷；

【导师】 何大可；

【作者基本信息】 西南交通大学 ， 密码学， 2010， 硕士

【摘要】 随着网络技术的不断进步,以及高性能计算机、网格等日益强大的计算环境的出现,计算的方式正在发生极大地改变,用户可以利用网络上强大的计算资源来完成自己的计算任务。因此,在这种环境中,如何确保用户数据的安全,是需要考虑的一个最基本问题。安全多方计算(Secure Multiparty Computation, SMC)是指一组互不信任的参与者,在不泄露各自私有信息的前提下进行的多方合作计算。自图灵奖得主A. C. Yao于上世纪80年代提出安全多方计算的概念以来,已得到许多学者的研究,其在密码学上的地位也日渐重要,它是电子选举、电子拍卖等密码学协议的基础。安全多方计算拓展了传统的分布式计算以及信息安全的范畴,为网络协作计算开创了一种新的计算模式,对解决网络环境下的信息安全问题具有重要的现实意义。安全多方计算的研究是一个极其复杂的工作,尤其对于特殊问题的研究远远没有达到理想的程度。在本文中,笔者将重点讨论有关科学计算领域的安全两方及多方计算协议的问题,其内容包括多项式的求值及乘法运算、初等函数的运算(除反三角函数外)、行列式求值、向量及矩阵的相关运算、解线性方程组等。文章对一些已有的半诚实模型下的安全两方计算协议进行推广,提出了一些的安全多方计算协议,并对各协议的正确性、安全性及复杂性进行了讨论。此外,本文还对恶意模型下的安全两方及多方计算进行了探讨,提出了几个可容忍恶意行为模型下的基础及扩展协议。更多还原

【Abstract】 With the continuous development of network technology, and with the appearance of high quality computers and net grid, the meanings and mode of computations are mostly changed. Users can make use of the powerful computation resources to complete their computation tasks. Therefore in this environment, the safety of users’data become very important and necessary.Secure Multi-party Computation (SMC) is a computation among a set of participants who do not have confidence in one another, in which participants collaborate on desired tasks but do not reveal private information of their own. Since it was initially suggested by A.C.Yao in the 1980s, SMC have been a major part of modern cryptography and attracted numerous researchers’interest. It’s the basis of many distributed cryptographic protocols such as threshold cryptosystem, electronic voting and electronic auction, etc. Secure multi-party computation has extended the traditional distributed computation and information security areas. A new computation mode was created for network collaborative computation. It is very significant to solve the information security problems in the network environment.Researches on the SMC are very complicated tasks. And researches on the special issues are far from ideal. In this paper, secure two-party and multiparty computation protocol problems were discussed which including polynomial evaluation and multiplication, elementary function operation (except inverse trigonometric function), determinant evalua-tion, vector and matrix related operation, solving linear equation and so on. Some existing secure two-party computation protocols of semi-honest models were generalized. A series of secure multiparty computation protocols were advanced. Correctness, security and complexi-ty of these protocols were discussed. In addition, it was discussed the secure two-party and multiparty computations of malicious models. Some foundations and extended protocols were advanced which can be tolerated under the malicious behavior models.更多还原