Research On Privacy Protection Of Blockchain Based On Lexical Analysis And Fully Homomorphic Encryption

Smart contracts are executable codes deployed in blockchain that do not rely on trusted third parties to automate the execution of transaction protocols.However,because all transactions such as transaction data are publicly shared in the blockchain nodes,there is a threat of transaction privacy leakage in the blockchain system.Although some scholars have proposed solutions,it is still a challenge to find a balance between maintaining transaction throughput and protecting transaction privacy.In view of this,this paper focuses on the smart contract of blockchain,and carries out the following research:(1)The lexical rules of Solidity,the smart contract language,are analyzed and the formal definitions of lexical units are given.Solidity's lexical analyzer SLA is designed and its automaton model is established.An automatic separation algorithm based on SLA for sensitive transaction data of the smart contract is given to make the privacy protection object more targeted.(2)In order to improve the efficiency of single encryption and single cipher transmission,an improved:NTRU-FHE17 scheme is proposed.The plaintext space of NHTU-FHE17 scheme is extended from Z2,of single bit range to Zp,and the correctness,homomorphism and security of the improved scheme are proved.Based on Ubuntu 18.04 and NTL library,the correctness and homomorphism of the improved scheme are verified and its efficiency is analyzed.The improved scheme is used to encrypt sensitive transaction data into corresponding cipher.In order to ensure that the data types of sensitive transactions remain unchanged before and after encryption in the smart contract,integralization algorithm of cipher vector and its inverse algorithm are designed to complete transaction verification.(3)Based on the lexical analyzer SLA and improved NTRU-FHE17 scheme,the smart contract privacy protection protocol SCPPP is designed and its verifiability and non-repudiation are analyzed.A block chain transaction model based on SCPP protocol is established.Based on Ubuntu 18.04?Remix?Metamask and Ropsten,deployment of multi-commodity transaction smart contract MTSC without privacy protection function is realized and transaction is completed;sensitive transaction data of MTSC are separated and separation efficiency of SLA is analyzed..(4)Theoretical and experimental results show that the lexical analyzer SLA has higher efficiency in separating sensitive transaction data of smart contract,and it does not affect the throughput of blockchain transactions;the improved NHTU-FHE17 scheme has higher security and efficiency;compared with the existing similar schemes,the proposed scheme combines lexical analysis and homomorphic encryption with smart contract of blockchain for the first time,which solves The problem of blockchain privacy protection to a certain extent.