信息安全研究 ›› 2022, Vol. 8 ›› Issue (10): 956-.

• 数据安全与隐私计算专题 • 上一篇    下一篇

多方安全计算框架下的智能合约方法研究

凡航1徐葳2,3王倩雯3王云河3   

  1. 1(清华大学五道口金融学院北京100083)
    2(清华大学交叉信息研究院北京100084)
    3(华控清交信息科技(北京)有限公司北京100084)
  • 出版日期:2022-10-25 发布日期:2022-10-24
  • 通讯作者: 凡航 博士,助理研究员.主要研究方向为金融科技、机器学习和隐私计算. fanhang@tsinghua.edu.cn
  • 作者简介:凡航 博士,助理研究员.主要研究方向为金融科技、机器学习和隐私计算. fanhang@tsinghua.edu.cn 徐葳 博士,副教授,华控清交首席科学家.主要研究方向为分布式系统、隐私计算和金融科技. xuwei@tsinghua.edu.cn 王倩雯 博士,高级工程师.主要研究方向为密码学与信息安全. wangqianwen@tsingj.com 王云河 博士,高级工程师.主要研究方向为隐私计算技术与标准化、数据安全与合规. yunhe@tsingj.com

Research on Smart Contract Method in the Framework of  Secure Multi-party Computation

  • Online:2022-10-25 Published:2022-10-24

摘要: 数据是国家新兴基础性战略资源,构建一个新兴的数据要素市场引发各界的高度关注,但随之而来的数据隐私泄露问题、数据滥用等问题也被引起重视.隐私计算解决了“数据可用不可见”的问题,但仍然缺乏数据使用过程的有效监管技术和手段,所以数据滥用的风险仍然存在.构建健康有序的数据要素市场需要建立在坚固的信任基础和智能化的监管措施之上.以去中心化的、多方互为监督的技术思路为指引,结合隐私计算中的多方安全计算技术,研究设计了一种新型的智能合约——计算合约.计算合约的内容由多方互相确认,可自动化执行,算法公开可验证,实现了数据流通使用的安全可控.利用所设计的智能合约,将其应用到北京大数据交易所中,为某银行融合多方数据测算贷款额度提供了支撑,演示了计算合约全生命周期的执行过程,相关流程可追溯、可监管.研究成果具有较好的示范性和可推广性,可为数字经济的发展提供有力支撑.

关键词: 大数据;计算合约;多方安全计算, 区块链;贷款风控

Abstract: Data is a new foundational and strategic resource of our country. It has attracted great attention from the whole society to construct an emerging data capitalization market, but the ensuing data privacy leakage problems, data abuse and other issues have also been paid attention to.  Privacypreserving computation has solved the problem by “using data without seeing data”, but there is still a lack of effective regulatory technology and means for data usage. The risk of dataabuse still exists. To build a healthy and orderly data capitalization market, it is necessary to construct a solid trust foundation and a series of intelligent supervision measures. Guided by the technical idea of decentralized and multiparty mutual supervision as well as combined with secure multiparty computing technology in computing, this paper studies and designs a new smart contract named computation contract. The content of the computational contract is mutually confirmed by multiple parties and can be executed automatically. The algorithm is open and verifiable, and the whole process of execution is traceable and supervisable, which realizes the secure control of data circulation and use. The designed smart contract has been applied to the Beijing Big Data Exchange to provide support for a bank to calculate the loan amount by integrating data from multiple parties, demonstrating the execution process of the whole life cycle of the computation contract and the related process is traceable and supervisable. The research results have good demonstration effect and can provide strong support for the development of digital economy.

Key words: big data, computation contract, secure multiparty computation, blockchain, loan risk management