Research On Privacy-preserving Scheme In Blockchain-based Redundant Shared Storage System

With the popularization of information technology and mobile communication technology,the scale of user data expands rapidly.In order to solve the mass data storage needs of users,cloud storage system has become the basic service facilities of modern information society.However,based on the centralized server to provide file storage service for users,centralized cloud storage system has the problem of single point of failure,which easily makes the user data unavailable and threatens the user privacy security.Blockchain-based distributed shared storage system adopts decentralized architecture,so there is no single point of failure.In addition,the system security is improved and the user's cost is reduced by combining blockchain technology.However,compared with the traditional centralized storage system,blockchain-based distributed shared storage system brings many advantages and also faces two limitations.First of all,in terms of data availability,blockchain-based distributed shared storage system adopts erasable code for data redundancy,which will bring huge network and computing overhead in the process of data recovery.Secondly,the on-chain data are publicly visible and transparent,which enables anyone in the network to obtain/query the transaction information and data on the blockchain.For the problems of data availability and user privacy protection in blockchainbased distributed storage system,we analyze the deficiencies of existing schemes and research work,and propose a blockchain-based distributed shared storage system with fault-tolerant and privacy preserving.The main contributions of this thesis are listed as follows.(1)We propose a redundant storage scheme for blockchain-based distributed shared storage system.In the process of data dividing of erasure code based schemes,the original file is divided into file shards of equal size,then the file shards are allocated to the same data block according to the position in the file.Therefore,there is huge network overhead and computing overhead in the process of reading invalid data blocks.Aiming at the efficiency problem of existing dividing methods,a file redundancy storage scheme based on discrete data dividing is proposed.In the proposed scheme,we divide the original file into file shards of equal size,then we design mapping rules to group the file shards and complete the chunking process.The scheme can obtain several file shards of continuous addresses through one decoding calculation.Therefore,the proposed scheme is more efficient when reading invalid blocks.(2)We propose a privacy-preserving scheme for blockchain-based distributed shared storage system.In the existing blockchain-based distributed shared storage system,the data on the blockchain is open and transparent,and anyone can access the data on the blockchain,which will cause the disclosure of users' private data.In this thesis,we design a privacy-preserving scheme based on stealth transmission scheme and non-interactive zero-knowledge proof protocol.By using stealth transmission scheme,users can use blockchain transactions for data transmission and hide the receiver's address and transaction content during transmission.By using non-interactive zero-knowledge proof protocol,we achieve the concealment of data in Ethereum smart contract,only the data owner can see the plaintext of data,other users can verify the validity of the data,but can not obtain the specific data content,also can not know the data owner's information.Finally,theoretical analysis and experimental analysis show that the privacy-preserving scheme is safe and practicable.(3)In this thesis,we design and implement the blockchain-based distributed shared storage system.Specifically,the prototype system is programmed with Python language,web3.py function library,Ethereum and IPFS.We divide the system into three modular: the front-end,smart contract and data storage.The corresponding functionalities are shown in this thesis.