Research On Efficient Data Processing Technology In Industrial Blockchain Data Collaborative Circulation

With its advantages of immutability,openness and transparency,blockchain can solve problems such as difficulty in confirming data rights and complex data transaction processes in industrial data circulation.For processing large data file,the existing solution is to store its hash value into the blockchain.However,the traditional hash value calculation method is difficult to meet the need of efficiently processig large-scale industrial data.On the other hand,since only the hash value is stored into the blockchain,the data demander can not confirm whether the source data itself is complete when only the hash value is received during the transaction.At the same time,the TPS(Transactions Per Second)of the current blockchain is generally low,which makes it difficult to meet the need of storing a large amount of industrial data into the blockchain.This paper studies the above problems,and the specific research contents are as follows:1.Aiming at the low efficiency of processing large-scale industrial data by the traditional hash value calculation method,this paper designs a data parallel processing method based on Vitis on the heterogeneous computing platform of CPU-FPGA.The method accelerates the calculation speed of large-scale industrial data hash value and improves the processing efficiency of large-scale industrial data through reasonable source data block,multi-stage data block allocation and SM3 computing unit scheduling,SM3 computing unit optimization etc.In addition,the method performs operations such as data block signature based on bilinear mapping,which provides basic support for the construction of data integrity verification between the two parties of the data transaction.2.For the problem that the data demander can not confirm whether the source data is complete when only the hash value is received during the transaction,this paper designs a two-party data integrity verification method based on smart contracts on the basis of the data parallel processing method.Through this method,the data demander can effectively verify the integrity of the source data according to the proof information without obtaining the source data during the transaction,and avoid paying a large communication cost of transmitting corrupted or tampered data.3.In view of the low TPS of the blockchain system,this paper has analyzed the transaction process of the BCOS blockchain.For the situation that the transaction pool is very busy and the transaction signature verification takes a lot of time,this paper designs a parallel transaction verification method based on Vitis.The method improves the TPS of the blockchain system by means of transaction signature preprocessing,batch caching and transmission of transaction signatures,optimization of elliptic curve point addition and point multiplication,parallel computing architecture of multiple point multiplication modules,and efficient point multiplication module scheduling.The simulation and experimental results show that,compared with the traditional data serial processing method,the data parallel processing method proposed in this paper can correspondingly improve the processing efficiency of 101 MB to 5GB data by about 2 to 9 times.It can be seen from the security analysis that in the special case where the data owner holds the signature private key,the data demander can effectively verify the integrity of the source data during the transaction.For the BCOS blockchain,the transaction parallel verification method improves the TPS of the blockchain system by nearly 1 times.