隐私计算关键技术与创新*

doi:10.12267/j.issn.2096-5931.2021.06.004

信息通信技术与政策 ›› 2021, Vol. 47 ›› Issue (6): 27-37.doi: 10.12267/j.issn.2096-5931.2021.06.004

隐私计算关键技术与创新*

Key technology and innovation of privacy preserving computing

符芳诚1,2,侯忱1,2,程勇2,陶阳宇2

1. 北京大学信息科学技术学院高可信软件技术重点实验室,北京 100871;
2. 腾讯数据平台部数据中心,北京 100083

出版日期:2021-06-15 发布日期:2021-07-15
作者简介:
符芳诚:北京大学信息科学技术学院高可信软件技术重点实验室在读博士,腾讯数据平台部数据中心研究员,主要研究领域为机器学习、隐私计算、分布式计算等
侯忱:北京大学信息科学技术学院高可信软件技术重点实验室在读硕士,腾讯数据平台部数据中心研究员,主要研究领域为机器学习、隐私计算、分布式计算等
程勇:腾讯数据平台部数据中心专家研究员,主要研究领域为隐私计算、联邦学习、多方安全计算、分布式计算等
陶阳宇:通信作者,腾讯数据平台部数据中心AI 平台总监,主要研究领域为机器学习、隐私计算、大数据、分布式系统等

FU Fangcheng1,2, HOU Chen1,2, CHENG Yong2, TAO Yangyu2

1. Department of Computer Science & Key Lab of High Confidence Software Technologies (MOE), Peking University,
Beijing 100871, China;
2. Department of Data Platform, TEG, Tencent Inc. , Beijing 100083, China

Online:2021-06-15 Published:2021-07-15

摘要/Abstract

摘要： 数据已经成为数字经济中重要的生产要素和战略资源。在实际应用中,数据分散在各个机构,由于隐私保护等原因,这些数据很难整合到一起,形成数据孤岛,从而阻碍大数据与人工智能产业的发展。隐私计算可以破解数据孤岛难题,在保障用户隐私的同时赋能多方数据协同应用,助力释放数据融合价值。对隐私计算的关键技术与创新成果,包括隐私集合求交技术、斜向联邦学习、异步并行计算、消息压缩机制、单向通信连接方案、可信执行环境、联合数据分析等进行了研究和分析,并对这些技术与创新在Angel PowerFL 通用隐私计算平台中的应用进行了介绍。

关键词: 隐私计算, 联邦学习, 联合数据分析, Angel PowerFL

Abstract: Big Data has become one of the factors of production and a strategic resource in the digital economy.However, data are usually scattered in different organizations and cannot be integrated in a centralized manner due to privacy concerns in practice, hindering the performance of many real-world big data applications. In order to tackle this problem, privacy preserving computing has been developed to enable the application of data from different parties for federated learning and analysis with privacy guarantees. In this paper, we provide a survey of the key techniques and advances in privacy preserving computing, including private set intersection, diagonal federated learning, asynchronous parallel computation, message compression protocols, unidirectional connections, trusted execution environments, and federated data analysis. Finally, we introduce the applications and techniques in Angel PowerFL, which is a general and industry-grade privacy preserving computing platform.

Key words: privacy preserving computing, federated learning, federated data analysis, Angel PowerFL

符芳诚, 侯忱, 程勇, 陶阳宇. 隐私计算关键技术与创新*[J]. 信息通信技术与政策, 2021, 47(6): 27-37.

FU Fangcheng, HOU Chen, CHENG Yong, TAO Yangyu. Key technology and innovation of privacy preserving computing[J]. Information and Communications Technology and Policy, 2021, 47(6): 27-37.

导出引用管理器 EndNote|Ris|BibTeX

链接本文:

http://ictp.caict.ac.cn/CN/10.12267/j.issn.2096-5931.2021.06.004

http://ictp.caict.ac.cn/CN/Y2021/V47/I6/27

参考文献 25

[1]	闫树, 袁博. 隐私计算———实现数据价值释放的突破口[ EB / OL ]. [ 2021-04-10]. https:/ / mp. weixin. qq.
	com / s / 2jL3x-m-IX5Cc8ZRSxpdVw.
[2]	闫树. 隐私计算产品测试及行业发展观察[EB / OL].[2021-04-10]. https:/ / mp. weixin. qq. com / s / 472o6CBMYbh2TPvX3rWv9g.
[3]	袁博. 隐私计算应用发展现状与趋势[ EB / OL].[ 2021-04-10 ]. https:/ / mp. weixin. qq. com / s / ETVL2
	QGL4Nxl9NT84TZqZQ.
[4]	中国信息通信研究院. 大数据白皮书(2020 年)[R], 2021.
[5]	Yang Qiang, Liu Yang, Chen Tianjian, et al. Federatedmachine learning: Concept and applications [C]. ACMTransactions on Intelligent Systems and Technology(TIST ) 10, 2 ( 2019) . https:/ / dl. acm. org / doi / 10.1145/ 3298981.
[6]	Jakub Konecny, Brendan McMahan H, Yu Felix X, et al. Federated learning: strategies for improving communication efficiency [ C ]. [ 2021-03-10 ]. https:/ / arxiv. org /abs / 1610. 05492.
[7]	Li Tian, Anit Kumar Sahu, Ameet Talwalkar, et al. Federated learning: challenges, methods, and future
	directions[C]. IEEE Signal Processing Magazine , 2020.https:/ / arxiv. org / abs / 1908. 07873.
[8]	Jiang Jie, Yu Lele, Jiang Jiawei, et al. Angel: a new large-scale machine learning system[J]. National Science
	Review, 2018,5(2): 216-236.
[9]	Zhang Zhipeng, Cui Bin, Shao Yingxia, et al. PS2:parameter server on spark[C]. SIGMOD’19, 2019(6):376-388. https:/ / dl. acm. org / doi / 10. 1145/ 3299869. 3314038.
[10]	Andrew Yao Chi-Chih, “Protocols for secure computations ( extended abstract )”, 23rd annual symposium on foundations of computer science [ Z / OL ]. [ 2021-03-
20	]. https:/ / ieeexplore. ieee. org / document / 4568388.[11] Evans David, Vladimir Kolesnikov, Rosulek Mike.
	Pragmatic introduction to secure multi-party computation[Z / OL ]. [ 2021-03-20]. https:/ / securecomputation.org / .
[12]	Pascal Paillier. Public-key cryptosystems based on
	composite degree residuosity classes[Z / OL]. [2021-03-
20	]. https:/ / link. springer. com / chapter / 10. 1007/ 3-
54	0-48910-X_16.
[13]	Ronald L Rivest, Adi Shamir, Leonard M, et al. A method for obtaining digital signatures and public-key
	cryptosystems [ Z / OL ]. [ 2021-03-10 ]. https:/ /dl. acm. org / doi / 10. 1145/ 359340. 359342.
[14]	Whitfield Diffie, Martin Hellman E. New directions in cryptography [ J ]. IEEE Transactions on Information
	Theory, 2014,22(6): 644 – 654.
[15]	Emiliano De Cristofaro, Gene Tsudik. Practical private set intersection protocols with linear computational and bandwidth complexity[Z / OL]. [2021-03-10]. https:/ /eprint. iacr. org / 2009/ 491. pdf.
	[ 16 ] Emiliano De Cristofaro, Gene Tsudik. On the performance of certain Private Set Intersection protocols[C], 2012. https:/ / eprint. iacr. org / 2012/ 054. pdf.
[17]	Michael J Freedman, Kobbi Nissim, Benny Pinkas.Efficient Private Matching and Set Intersection [ C ],EUROCRYPT ’ 2004. https:/ / iacr. org / archive /
	eurocrypt2004/ 30270001/ pm-eurocrypt04-lncs. pdf.
[18]	Yang Liu, Xiong Zhang, Libin Wang. Asymmetrical Vertical Federated Learning [ C ], 2020. https:/ /arxiv. org / abs / 2004. 07427.
[19]	Cheng Kewei, Fan Tao, Jin Yilun, et al. SecureBoost: a lossless federated learning framework [ EB / OL ].
	[2021-03-10]. http:/ / arxiv. org / abs / 1901. 08755.
[20]	Fu Fangcheng, Shao Yingxia, Yu Lele, et al. VF 2Boost: very fast vertical federated gradient boosting forcross-enterprise learning[Z]. SIGMOD, 2021.
[21]	Stephen Hardy, Wilko Henecka, Hamish Ivey-Law, et al. Private federated learning on vertically partitioned
	data via entity resolution and additively homomorphicencryption[Z / OL]. [2021-03-10]. http:/ / arxiv. org /
	abs / 1711. 10677.
[22]	Yang Shengwen, Ren Bing, Zhou Xuhui, et al. Paralleldistributed logistic regression for vertical federated
	learning without third-party coordinator[Z / OL]. [2021-03-10]. http:/ / arxiv. org / abs / 1911. 09824.
[23]	Martin Abadi, Chu Andy, Ian J, et al. Deep learningwith differential privacy [ Z / OL ]. [ 2021-03-10 ].
	https:/ / dl. acm. org / doi / 10. 1145/ 2976749. 2978318.
[24]	Song Shuang, Kamalika Chaudhuri, Anand D Sarwate.Stochastic gradient descent with differentially private
	updates [ Z / OL ]. [ 2021-03-10]. https:/ / ieeexplore.ieee. org / document / 6736861.
[25]	McKeen Frank, Ilya Alexandrovich, Alex Berenzon, etal. Innovative instructionsand software model for isolatedexecution [ Z / OL ]. [ 2021-03-10]. https:/ / dl. acm.org / doi / 10. 1145/ 2487726. 2488368.

隐私计算关键技术与创新*

Key technology and innovation of privacy preserving computing

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献 25

相关文章 9

编辑推荐

Metrics

本文评价

[1]	闫树, 吕艾临. 隐私计算发展综述[J]. 信息通信技术与政策, 2021, 47(6): 1-11.
[2]	袁博, 王思源. 隐私计算产品评估体系[J]. 信息通信技术与政策, 2021, 47(6): 12-18.
[3]	夏家骏, 鲁颖, 张子扬, 张钰婷, 张佳辰. 基于秘密共享与同态加密的纵向联邦学习方案研究[J]. 信息通信技术与政策, 2021, 47(6): 19-26.
[4]	徐潜, 章庆, 喻博, 于文青, 贺伟. 基于中间件与区块链的异构隐私计算平台互通系统研究[J]. 信息通信技术与政策, 2021, 47(6): 38-49.
[5]	黄翠婷, 张帆, 孙小超, 卞阳. 隐私集合求交技术的理论与金融实践综述[J]. 信息通信技术与政策, 2021, 47(6): 50-.
[6]	强锋, 薛雨杉, 相妹. 隐私计算在金融领域的合规性分析[J]. 信息通信技术与政策, 2021, 47(6): 57-62.
[7]	程强, 刘姿杉. 电信网络智能化发展现状与未来展望[J]. 信息通信技术与政策, 2020, 46(9): 16-22.
[8]	李鉴, 　邵云峰, 　卢燚, 　吴骏. 联邦学习及其在电信行业的应用[J]. 信息通信技术与政策, 2020, 46(9): 35-41.
[9]	梁伟, 张政, 冯明, 何志强. 基于区块链的可信数据交换技术与应用[J]. 信息通信技术与政策, 2020, 46(4): 91-96.