Design and Implementation of Big Data Analysis Algorithm in Ciphertext Domain Based on Homomorphic Encryption

Yun Fan Huang,Dian Hua Tang,Wei Zhao,Juan Ren,Xiao Peng Yu
DOI: https://doi.org/10.1109/icaml54311.2021.00038
2021-01-01
Abstract:As a new technology emerging in the era of Internet big data, cloud computing has huge storage space and tough computing power, providing a solution for the analysis and processing of big data. However, with the concern of privacy disclosure, data analysis in plaintext can not meet the needs of privacy protection. Therefore, the data security issues in the process of cloud computing need to be solved. Homomorphic encryption technology has the characteristics of computing on encrypted data, which can provide security protection for privacy data in cloud computing. In this paper, combined with CKKS homomorphic encryption scheme, using two different packing encoding methods based on row encoding and column encoding, we design different ciphertext data analysis algorithm schemes, including ciphertext mean calculation, ciphertext variance calculation, ciphertext covariance matrix calculation, etc., to achieve privacy protection in the process of big data analysis. In the order of millions of data sets, the ciphertext data analysis can achieve second order response. At the same time, the ciphertext analysis algorithm proposed in this paper supports multi-dimensional sample data sets. The experimental results show that the ciphertext data analysis under 16-dimensional data sets reaches 23.49 seconds.
What problem does this paper attempt to address?