Knowledge-oriented secure multiparty computation
Knowledge-oriented secure multiparty computation
复制标题
面向知识的安全多方计算
DOI：
10.1145/2336717.2336719
复制
发表时间：
2012
期刊：
影响因子：
--
通讯作者：
M. Srivatsa
中科院分区：
文献类型：
--
作者：
Piotr (Peter) Mardziel;M. Hicks;Jonathan Katz;M. Srivatsa
用于安全多方计算(SMC)的协议允许一组相互不信任的各方计算他们的私有输入的函数f，而除了结果所暗示的以外，不显示关于他们的输入的任何信息。然而，根据f的不同，结果本身可能会透露出比各方满意的更多信息。几乎所有以前关于SMC的工作都认为f是已知的。没有回答的问题是，各方应该如何决定计算f对他们来说是否“安全”。我们在这里提出了一种将信念跟踪应用于SMC的方法，以准确地解决这个问题。在我们的方法中，每个参与方都能够对其他各方通过计算f而获得的知识增长进行推理，并可以选择不参与(或只部分参与)，以限制知识的增长。我们开发了两种技术-信念集方法和SMC信念跟踪方法-证明了它们的可靠性，并通过一系列实验讨论了它们的精度/性能权衡。
Protocols for secure multiparty computation (SMC) allow a set of mutually distrusting parties to compute a function f of their private inputs while revealing nothing about their inputs beyond what is implied by the result. Depending on f, however, the result itself may reveal more information than parties are comfortable with. Almost all previous work on SMC treats f as given. Left unanswered is the question of how parties should decide whether it is "safe" for them to compute f in the first place. We propose here a way to apply belief tracking to SMC in order to address exactly this question. In our approach, each participating party is able to reason about the increase in knowledge that other parties could gain as a result of computing f, and may choose not to participate (or participate only partially) so as to restrict that gain in knowledge. We develop two techniques---the belief set method and the SMC belief tracking method---prove them sound, and discuss their precision/performance tradeoffs using a series of experiments.