| For the financial industry,the development of modern financial business risk control is highly dependent on the support of multisource data.However,to meet the relevant requirements of laws and regulations.The privacy protection of sensitive data can be carried out in the form of a technical form which achieve "the original data does not leave the domain,and the data is not visible but usable".Therefore,it is necessary to design a privacy-preserving computing scheme that supports multi-source data evaluation to break through financial data silos and enable evaluation of data between financial institutions.Homomorphic encryption is a cryptographic technology that can be used to construct a privacy protection risk control system.However,the availability of homomorphic encryption technology is not ideal,and there is a certain gap in the time and resources consumed between ciphertext computing and plaintext computing,so it is difficult to use for data sharing under big data conditions.The design of homomorphic encryption production in practical applications is also worth in-depth study.At present,there is no mechanism related to homomorphic encryption in the field of commercial cryptography.It is necessary to realize the implementation security of algorithms,key management security and device security at the same time,and the applicability improvement in the actual financial cloud environment is also a problem.Aiming at the efficiency of homomorphic encryption algorithm,this paper proposes an efficient homomorphic encryption algorithm,and proves that the scheme has about 100 times the performance improvement compared with the BFV scheme in SEAL.On this basis,a set of general ciphertext calculation extension algorithms is proposed,which realizes integer,floating-point numbers,non-negative and negative number operations of ciphertext addition,subtraction,multiplication and division.Aiming at the application problem of homomorphic encryption algorithm,a cryptographic machine supporting homomorphic encryption was designed and developed,which met the security standards of commercial cryptography secondary modules.The "cryptographic cloud" is designed and implemented,and the concept of "cryptography-as-aservice" is proposed to realize the convenient deployment of homomorphic encryption algorithm products.Aiming at the privacy protection problem in financial data joint operation,a set of multi-source data financial risk control privacy protection scheme is designed,which realizes data security sharing and joint computing between multiple financial institutions,and enriches the data dimension of financial risk control in ensuring data security. |
