Floating-point number processing method and device, terminal and storage medium

Abstract

The invention provides a floating-point number processing method and device, a terminal and a storage medium capable of efficiently encrypting and calculating a floating-point number based on a homomorphic encryption algorithm of a classical password, and the method is characterized by comprising the following steps: S1, obtaining a to-be-encrypted first floating-point number; s2, representing the first floating-point number as a score, and obtaining a corresponding to-be-encrypted numerator and a to-be-encrypted denominator; s3, encrypting the to-be-encrypted numerator and the to-be-encrypted denominator based on a homomorphic encryption algorithm to obtain a first numerator and a first denominator corresponding to the first floating-point number; s4, based on the first numerator and the first denominator, completing homomorphic operation for the floating-point number to form a second numerator and a second denominator; and S5, decrypting the second numerator and the second denominator according to a predetermined key, and dividing to obtain a second floating-point number after homomorphic operation is performed on the first floating-point number.

Description

technical field [0001] The invention belongs to the field of data security, and in particular relates to a floating point number processing method, device, terminal and storage medium. Background technique [0002] Since the mathematical foundation of cryptography is based on finite groups, rings, and fields of integers, the processing of floating-point numbers is naturally not supported in the context of cryptography. In recent years, with the gradual popularity of lattice ciphers, there have been some academic innovations in the approximate expression of floating-point numbers, but they still cannot meet the requirements of engineering applications. [0003] From the perspective of privacy computing, the data needs to be encrypted with a cryptographic algorithm, and then the ciphertext is used for operation. In practical engineering, the operation of floating-point numbers must be indispensable. Therefore, this causes the following two problems: [0004] 1. Floating-point...

AI Technical Summary

Problems solved by technology

[0004] 1. Floating-point numbers cannot use classical cryptography technology for privacy protection, let alone calculations under the background of classical cryptography

[0005] 2. The floating-point numbers processed by the lattice cipher have a large loss of precision. The current best solution, after several floating-point number ciphertext multiplications, the precision of the floating-point numbers cannot become very significant, resulting in unusable calculation results

Images