Research And Implementation Of Data Encoding For Homomorphic Encryption

With the rapid development of information technology,human beings are increasingly inseparable from the computer's powerful data processing capabilities.People enjoy a convenient life brought by information technology such as new generation distributed storage,computing service outsourcing,and big data analysis,and they also face enormous risks of privacy leakage.How to protect user privacy is an important issue in the promotion of these technologies.Fully homomorphic encryption can satisfy any function operation on ciphertext without revealing any plaintext information,and the function operation effect is reflected on the decrypted plaintext.Its characteristics can be applied to the above information technology,ciphertext calculation,and privacy protection.Therefore,fully homomorphic encryption has become a hot topic in cryptography circles in recent years.The characteristics of fully homomorphic encryption have broad application prospects,but they cannot avoid the problems that exist.Research shows that efficiency problems make fully homomorphic encryption impossible to use widely in business and everyday life.Encoding converts application data into a specific string that can be manipulated by homomorphic encryption.This is an essential part of ciphertext computation,and different encoding schemes can lead to different homomorphic encryption computational efficiency.Therefore,the most important issue facing fully homomorphic encryption applications is how to encode data and how to perform arithmetic operations on the encoded string.In response to the above issues,the research work in this paper mainly includes the following aspects:1 Comparing common data encoding methods,and using the basic operations of symmetric ternary coding,we designed the arithmetic operations of symmetric ternary coding.2 Analyze the process of encryption and decryption and homomorphic operation of BGV12 scheme,apply symmetric ternary coding in BGV12 scheme,and complete the design of homomorphic encryption arithmetic operation of symmetric ternary coding.3 Using software engineering idea,the detailed design of the class is given,the complexity of the important algorithm is analyzed and optimized,the pseudo code of the algorithm is given,and based on the HElib open-source code base,the homomorphic encryption arithmetic operations of symmetrical ternary coding are implemented.4 The homomorphic encryption arithmetic operations of the symmetric ternary coded code are tested,and the experimental results are recorded and compared with the binary coded homomorphic encryption arithmetic.The results show that symmetrical ternary coding has advantages in time efficiency of homomorphic multiplication arithmetic operations and improves space utilization.5 Design a fully homomorphic encryption scheme to implement the privacy protection application model in machine learning and make a validation experiment on the model.The experimental results prove the accuracy of machine learning based on gradient descent method under homomorphic encryption.Finally,the optimization method of the model is given.Through the above-mentioned work,the homomorphic encryption scheme of symmetric ternary coding is designed,and the connection between application data and arithmetic operations under homomorphic encryption is established,and the correctness and effectiveness of the scheme are verified through experiments.