A Principal Component Analysis Scheme for Security Outsourcing in Cloud Environment Based on Secret Sharing

doi:10.3969/j.issn.1671-1122.2023.04.007

Abstract

Abstract:

The computational overhead of principal component analysis is so high that local devices cannot afford it and often require secure outsourcing of computational tasks. The data security issue in outsourcing computation has gradually become a difficult point for cloud computing security research. This paper proposed a secure outsourcing scheme based on secret sharing in cloud environment, which was based on additive secret sharing technology and designed with secure division and secure square root computation protocols. The PCA secure outsourcing computation could be finished by two cloud servers collaboratively performing the covariance matrix, Lanczos, Householder and other secure protocols. Compared with other secure outsourcing computation schemes, this scheme can better support client offline and multi-party data aggregation with better computational overhead, and the experiments verified the effectiveness of the scheme.

Key words: data security, cloud computing, outsourced computing, secret sharing, principal component analysis

CLC Number:

TP309

MA Min, FU Yu, HUANG Kai. A Principal Component Analysis Scheme for Security Outsourcing in Cloud Environment Based on Secret Sharing[J]. Netinfo Security, 2023, 23(4): 61-71.

Figures/Tables 8

References 30

[1]	LI Fenghua, LI Hui, NIU Ben, et al. Privacy Computing: Concept, Computing Framework, and Future Development Trends[J]. Engineering, 2019, 5(6): 1179-1192. doi: 10.1016/j.eng.2019.09.002
[2]	ZHOU Xiao, GUO Ying, LI Fangshun, et al. Identifying and Assessing Innovation Pathways for Emerging Technologies: A Hybrid Approach Based on Text Mining and Altmetrics[J]. IEEE Transactions on Engineering Management, 2020, 68(5): 1360-1371. doi: 10.1109/TEM.2020.2994049 URL
[3]	FENG Chunmei, XU Yong, LIU Jinxing, et al. Supervised Discriminative Sparse PCA for Com-Characteristic Gene Selection and Tumor Classification on Multiview Biological Data[J]. IEEE Transactions on Neural Networks and Learning Systems, 2019, 30(10): 2926-2937. doi: 10.1109/TNNLS.2019.2893190 pmid: 30802874
[4]	ZHAO Qing, WANG Huiwen, WANG Shanshan. A Principal Component Analysis of Interval Data Based on Center and Log-Radius[J]. Journal of Beijing University of Aeronautics and Astronautics, 2021, 47(7): 1414-1421.
	赵青, 王惠文, 王珊珊. 基于中心-对数半长的区间数据主成分分析[J]. 北京航空航天大学学报, 2021, 47(7):1414-1421.
[5]	SONG Yue, SEBE N, WANG Wei. Batch-Efficient Eigen Decom-position for Small and Medium Matrices[EB/OL]. (2022-07-09)[2022-10-23]. https://arxiv.org/abs/2207.04228v1.
[6]	LI Fenghua, LI Hui, JIA Yan, et al. Privacy Computing: Concept, Connotation and Its Research Trend[J]. Journal on Communications, 2016, 37(4): 1-11. doi: 10.11959/j.issn.1000-436x.2016078
	李凤华, 李晖, 贾焰, 等. 隐私计算研究范畴及发展趋势[J]. 通信学报, 2016, 37(4): 1-11. doi: 10.11959/j.issn.1000-436x.2016078
[7]	ZHANG Yuqing, WANG Xiaofei, LIU Xuefeng, et al. Survey on Cloud Computing Security[J]. Journal of Software, 2016, 27(6): 1328-1348.
	张玉清, 王晓菲, 刘雪峰, 等. 云计算环境安全综述[J]. 软件学报, 2016, 27(6): 1328-1348.
[8]	QIAN Wenjun, SHEN Qingni, WU Pengfei, et al. Research Progress on Privacy-Preserving Techniques in Big Data Computing Environment[J]. Chinese Journal of Computers, 2022, 45(4): 669-701.
	钱文君, 沈晴霓, 吴鹏飞, 等. 大数据计算环境下的隐私保护技术研究进展[J]. 计算机学报, 2022, 45(4): 669-701.
[9]	FENG Dengguo, YANG Kang. Concretely Efficient Secure Multi-Party Computation Protocols: Survey and More[EB/OL]. (2022-06-14)[2022-10-23]. https://sands.edpsciences.org/articles/sands/full_html/2022/01/sands20210001/sands20210001.html.
[10]	LIU Yang, MA Zhuo, LIU Ximeng, et al. Privacy-Preserving Object Detection for Medical Images with Faster R-CNN[J]. IEEE Transactions on Information Forensics and Security, 2022, 17: 69-84. doi: 10.1109/TIFS.2019.2946476 URL
[11]	YING Zuobin, CAO Shuanglong, LIU Ximeng, et al. Privacy Signal: Privacy-Preserving Traffic Signal Control for Intelligent Transportation System[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(9): 16290-16303. doi: 10.1109/TITS.2022.3149600 URL
[12]	BI Renwan, XIONG Jinbo, TIAN Youliang, et al. Achieving Lightweight and Privacy-Preserving Object Detection for Connected Autonomous Vehicles[J]. IEEE Internet of Things Journal, 2022, 10(3): 2314-2329. doi: 10.1109/JIOT.2022.3212464 URL
[13]	LI Minghui, CHOW S S M, HU Shengshan, et al. Optimizing Privacy-Preserving Outsourced Convolutional Neural Network Predictions[J]. IEEE Transactions on Dependable and Secure Computing, 2020, 19(3): 1592-1604. doi: 10.1109/TDSC.2020.3029899 URL
[14]	LIU Xiaoning, ZHENG Yifeng, YUAN Xingliang, et al. Securely Outsourcing Neural Network Inference to the Cloud with Lightweight Techniques[J]. IEEE Transactions on Dependable and Secure Computing, 2022.
[15]	FENG Qi, HE Debiao, LIU Zhe, et al. SecureNLP: A System for Multi-Party Privacy-Preserving Natural Language Processing[J]. IEEE Transactions on Information Forensics and Security, 2020(15): 3709-3721.
[16]	JOLLIFFE I T, CADIMA J. Principal Component Analysis: A Review and Recent Developments[EB/OL]. (2016-04-13)[2022-10-13]. https://royalsocietypublishing.org/doi/10.1098/rsta.2015.0202.
[17]	WANG Xinxin. Improvement and Analysis of Lanczos Algorithm for Solving Eigenvalue Problems of Symmetric Matrices[D]. Harbin:Harbin Institute of Technology, 2008.
	王欣欣. 求解对称矩阵特征值问题的 Lanczos 算法的改进及分析[D]. 哈尔滨: 哈尔滨工业大学, 2008.
[18]	DUBRULLE A A. Householder Transformations Revisited[J]. SIAM Journal on Matrix Analysis and Applications, 2000, 22(1): 33-40. doi: 10.1137/S0895479898338561 URL
[19]	GE J, WANG Zhaoran, WANG Mengdi, et al. Minimax-Optimal Privacy-Preserving Sparse PCA in Distributed Systems[EB/OL]. (2020-02-25)[2022-10-20]. http://proceedings.mlr.press/v84/ge18a.html.
[20]	WANG Di, XU Jinhui. Principal Component Analysis in the Local Differential Privacy Model[J]. Theoretical Computer Science, 2020, 809: 296-312. doi: 10.1016/j.tcs.2019.12.019 URL
[21]	LIU Yingting, CHEN Chaochao, ZHENG Longfei, et al. Privacy Preserving PCA for Multiparty Modeling[EB/OL]. (2020-02-06)[2022-10-23]. https://www2.isye.gatech.edu/-fferdinando3/cfp/PPAI20/papers/paper_16.pdf.
[22]	XIANG Can, TANG Chunming, CAI Yunlu, et al. Privacy-Preserving Face Recognition with Outsourced Computation[J]. Soft Computing, 2016, 20(9): 3735-3744. doi: 10.1007/s00500-015-1759-5 URL
[23]	AL-RUBAIE M, WU Peiyuan, CHANG J M, et al. Privacy-Preserving PCA on Horizontally-Partitioned Data[C]// IEEE. 2017 IEEE Conference on Dependable and Secure Computing. New York: IEEE, 2017: 280-287.
[24]	ZHANG Yushu, XIAO Xiangli, YANG Luxing, et al. Secure and Efficient Outsourcing of PCA-Based Face Recognition[J]. IEEE Transactions on Information Forensics and Security, 2019, 15: 1683-1695. doi: 10.1109/TIFS.10206 URL
[25]	REN Yanli, XU Xiao, FENG Guorui, et al. Non-Interactive and Secure Outsourcing of PCA-Based Face Recognition[EB/OL]. (2021-11-01)[2022-10-23]. https://dl.acm.org/doi/10.1016/j.cose.2021.102416.
[26]	WEI Lu, SARWATE A D, CORANDER J, et al. Analysis of a Privacy-Preserving PCA Algorithm Using Random Matrix Theory[C]// IEEE. 2016 IEEE Global Conference on Signal and Information Processing (GlobalSIP). New York: IEEE, 2016: 1335-1339.
[27]	ANOWAR F, SADAOUI S, SELIM B. Conceptual and Empirical Comparison of Dimensionality Reduction Algorithms (PCA, KPCA, LDA, MDS, SVD, LLE, ISOMAP, LE, ICA, t-SNE)[EB/OL]. (2021-05-01)[2022-10-23]. https://www.sciencedirect.com/science/article/abs/pii/S1574013721000186.
[28]	HUANG Kai, LIU Ximeng, FU Shaojing, et al. A Lightweight Privacy-Preserving CNN Feature Extraction Framework for Mobile Sensing[J]. IEEE Transactions on Dependable and Secure Computing, 2019, 18(3): 1441-1455.
[29]	WANG Qian, DU Minxin, CHEN Xiuying, et al. Privacy-Preserving Collaborative Model Learning: The Case of Word Vector Training[J]. IEEE Transactions on Knowledge and Data Engineering, 2018, 30(12): 2381-2393. doi: 10.1109/TKDE.69 URL
[30]	LINDELL Y. How to Simulate It-A Tutorial on the Simulation Proof Technique[J]. Tutorials on the Foundations of Cryptography, 2017(6): 277-346.

数据集	特征	明文			SPCA
数据集	特征	准确度	召回率	F1-score	准确度	召回率	F1-score
Diabetes	8	83%	84%	0.84	83%	88%	0.85
winequality(red)	11	95%	96%	0.95	95%	97%	0.96
Ionosphere	34	93%	78%	0.85	92%	80%	0.85
sonar	60	78%	85%	0.82	82%	91%	0.86

协方差矩阵维度	Al-Rubaie	SPCA
协方差矩阵维度	服务器	服务器
8	28.3 s	1.58 s
10	49.6 s	2.36 s
34	43.2 min	13.92 s
50	126.7 min	47.21 s

	文献[24]方案		SPCA
协方差矩阵维度	客户端 /s	服务器/s	主成分个数	客户端/s	服务器/s
1000	1.4597	6.9114	3	0.0151	3.4871
1000	1.4597	6.9114	6	0.0151	12.6735
1500	2.2959	23.6918	3	0.0299	5.5616
1500	2.2959	23.6918	6	0.0299	16.1527
2000	3.7349	49.1470	3	0.0511	7.635
2000	3.7349	49.1470	6	0.0511	22.1756
2500	4.7115	103.0026	3	0.0795	10.524
2500	4.7115	103.0026	6	0.0795	27.3319