CN111402055A - Decentralized data transaction method based on block chain - Google Patents

Abstract

The invention discloses a decentralized data transaction method based on a block chain, which is applied to a big data transaction environment consisting of a key generation center, the block chain and an intelligent contract, and is used for realizing data summary information acquisition of a data transaction market through the block chain, recording evaluation information of data use conditions and transaction records in data transaction; the intelligent contract can release data requirements, match and screen the data required by the data demand provider, and automatically complete the distribution of data rewards. The invention can solve the problem that the data transaction center is not credible, so as to ensure the ownership and the right of knowing the personal data in the transaction of the data provider, improve the intellectualization and the high efficiency of the data transaction process, and provide safe and reliable big data transaction service for both transaction parties.

Description

A Decentralized Data Transaction Method Based on Blockchain

technical field

The invention relates to the technical field of information security, in particular to a big data platform transaction method based on a block chain and a smart contract.

Background technique

Blockchain is a decentralised technology, and its non-tampering properties of information, distributed storage of data and a series of features make it gradually applied in more fields. With the advent of the era of big data, some service providers need to mine and analyze massive amounts of big data in order to improve more accurate customer service. Data owners are also more aware of the value of data and pay more attention to data privacy. At present, the big data trading market provides a market environment for data trading, but there are many defects in the privacy protection of user data and the value reward of data. Therefore, in this case, a safe and reliable trading platform is provided to allow Better utilization of the value of big data is an important direction.

A smart contract is a digital agreement that uses algorithms and programs to formulate contract terms, deploys on the blockchain, and executes automatically according to the rules. When the blockchain is applied to more fields, more practical contracts are required to meet the needs of the system.

At present, the big data trading platform adopts a third-party transaction mode. The data provider uploads the data to the third-party platform and formulates an initial price. When the data demander wants to use the data, it pays the data provider the corresponding price The reward of , that is, the data can be accessed. In this transaction mode, there are the following defects:

1. In the third-party data trading platform, the data transaction will charge an expensive handling fee. At the same time, the third-party center also has the problem of untrustworthiness. It may tamper with the provider's data without the permission of the data provider, or even resell the provider's data. data.

2. Once a third-party trading center encounters a network attack, it is easy to face the problem of a single point of failure, resulting in a large amount of data leakage. In this process, it is easy to cause user data and identity privacy issues.

3. Data demanders need to retrieve their own data from a large amount of data. The quality of data is mixed, and there is a lack of a relatively complete evaluation mechanism for big data. The degree of intelligence in data matching and screening is low, so the efficiency is relatively low, and the data mainly comes from comparison. single data provider.

4. In the data transaction, the data provider submits the data to the third-party trading platform, and entrusts the third-party platform to be responsible for the data transaction. Therefore, to a certain extent, the data ownership and the right to know of the data provider rely more on the trading center. reliability.

5. In the value reward of data, the reward model for third-party data transactions is relatively simple. The data provider presets the initial price of the data set, and the data demander pays the price corresponding to the reward, which cannot reflect the data quality according to the data. value.

SUMMARY OF THE INVENTION

The present invention aims to provide a blockchain-based decentralized data transaction method, aiming at the defects and deficiencies of the existing big data transaction platform, such as uneven quality of process data, untrustworthy third-party data centers, and easy data and identity privacy security issues. In order to solve the problem of untrustworthy data centers, ensure data providers' ownership and right to know personal data in transactions, improve the intelligence and efficiency of the data transaction process, and provide a safe and fair big data transaction model for both parties. .

The present invention is achieved by adopting the following technical solutions:

The feature of a blockchain-based decentralized data transaction method of the present invention is that it is applied in a transaction environment composed of a key generation center, a blockchain and a smart contract, and is implemented in the transaction environment according to the following steps A peer-to-peer transaction method between several data providers, one data demander, and several data broker nodes:

Step 1. Any data provider i obtains an anonymous identity public-private key pair (PK _i , SK _i ) and an account address α _i through registration, and applies to the key generation center for a public-private key pair (pk i , SK i ) for encryption and decryption. _i ,ski ₎ ;

Step 2, the data provider i obtains the data ciphertext Epk _i (D _i ) after encrypting its own data Di, wherein, D _i is the plaintext data that the data provider _i is used for the big data transaction, Epk _i ( ) indicates that the encryption operation is performed using the public key pk _i of the data provider i;

The data provider i provides a data digest about the plaintext data D _i , including: a description Des(D _i ) of the plaintext data D _i , a Hash value Hash(D _i ) of the plaintext data D _i , data Type Type(D _i ), number of data entries DT _i (D _i ), signature information Sig _Ski , and account address α _i for receiving data rewards;

Step 3, the data provider i selects any agent node B, and sends the data ciphertext Epk _i (D _i ) and the data digest to the agent node B;

Step 4. After the agent node B receives the data ciphertext Epk _i (D _i ) and the data digest of the data provider i, it uses formula (1) and formula (2) to generate the unique identifier UID _i and store it respectively. Prove SP(UID _i ):

UID _i =Hash(α _i ,Hash(D _i )) (1)

In formula (1), Hash( ) represents hash calculation;

是代理商节点B利用自身公私钥对(SK_B，PK_B)对存储证明SP(UID_i)的签名，AL_UIDi是所述数据提供商i的唯一标识UID_i的数据密文Epk_i(D_i)的存储位置；In formula (2),

is the signature of the storage proof SP ( _{UID i} ) by the agent node B using its own public and private key pair (SK _B , PK _{B ) . i} ) the storage location;

Step 5. The agent node B uses formula (3) to generate data summary information T _brief (UID _i ):

Step 6. The agent node B obtains the data summary information generated by all agent nodes in the Nth time period, uses the obtained data summary information as the main part of the block, and adds the block header and packages it as the Nth summary. block, the information of the block header in the Nth digest block includes: timestamp, the Merkle Tree of the data digest information and the Hash value of the N-1th digest block;

After the agent node B reaches a consensus with other agent nodes, it links the Nth summary block to the N-1th summary block to form the latest data summary chain. When N=0, the N-1th summary block is formed. The hash value of each digest block is the set hash value;

其中，

分别是第N个摘要区块上的第n个评价区块中唯一标识UID_i的所对应数据的完整性、一致性和准确性的评分；当n＝1时，

为所设定的完整性、一致性和准确性的初始评分；Step 7. Elect the Nth master node from the agent nodes corresponding to the data in the Nth digest block in the latest data digest chain, and the Nth master node generates the Nth master node on the Nth digest block. The nth evaluation block of the evaluation side chain, the block body part of the nth evaluation block contains the quality description and quality score of the data in the Nth digest block in the nth evaluation block

in,

are the scores of the integrity, consistency and accuracy of the corresponding data uniquely identifying UID _i in the nth evaluation block on the nth summary block; when n=1,

An initial score for completeness, consistency and accuracy as set;

The information of the block header of the nth evaluation block includes: timestamp, MerkleTree of the data evaluation information and the Hash value of the n-1th evaluation block; when n=1, the n-1th evaluation block The Hash value is the hash value of the Nth digest block;

Step 8. Any data demander j obtains an anonymous identity public and private key pair (PK _j , SK _j ) through registration, and applies to the key generation center for a public and private key pair (pk _j , sk _j ) for encryption and decryption ;

Step 9, the data demander j publishes a data demand contract through the smart contract, the data demand contract includes: demand data type D _S , data demand item number D _A , data demand period dl, data demand provider encryption public key pk _j , data quality requirement β, and data reward distribution rules;

After the data demander j deploys the data demand contract, it sends the data transaction budget τ to the contract account generated by the data demand contract;

Step 10, all data providers interact with the data demander j through the data demand contract, thereby responding to the data demand;

Step 11. Assuming that the data provider i responds to the data request, the data request contract reads the Nth digest block and the Nth digest block on the corresponding data digest chain according to the unique identifier UID _i of the data provider i that responds to the data request. After the first n evaluation blocks on the data evaluation side chain, according to the function in the data demand contract, a number of data of the data provider i responding to the data demand is matched and checked, and the data that meets the requirements of the data demand provider j will be checked. The data is added to the candidate data pool candidateDataList{};

Step 12, the data demander j checks the summary information and evaluation information of the data in the corresponding blockchain according to the unique identifier UID _i in the candidate data pool candidateDataList{}, and then the data demander j decides whether to purchase the corresponding data, if If you choose to buy, the corresponding data will be added to the transaction data pool dataUidList{};

Step 13, the data demand contract records the number of entries DR added to the data in the transaction data pool _{dataUidList {} }, when the entry demand DA of the data demand quotient is satisfied, or when the time limit of the data demand is reached, it indicates the data demand matching completed;

Step 14, the data demander j sends a data request to the data provider i of the data in the transaction data pool dataUidList{}, and the corresponding data provider _i decrypts the private key ski and the data in the data demand contract through its own. The encrypted public key pk _{j of the demander j} generates a proxy re-encryption key rk _i→j and sends it to the proxy node B that stores the data of the data provider i;

The corresponding agent node B uses the agent re-encryption key rk _i→j to re-encrypt the data ciphertext Epk _i (D _i ) of the data added by the data provider i to the transaction data pool dataUidList{}, and obtain the re-encrypted The ciphertext Epk′ _i (D _i ) is sent to the data demander j, so that the data demander j can use its own private key sk _j to decrypt the re-encrypted ciphertext Epk′ _i (D _i ), and obtain The data provider i joins the plaintext data D _i of the data in the transaction data pool dataUidList{};

Step 15, the data demand quotient j calculates the corresponding hash value Hash(D _i ) according to the plaintext data D _i obtained by decryption, if it is the same as the hash value of the corresponding plaintext data D _i in the data digest chain, it means The decrypted plaintext data D _i is the original data provided by the data provider i;

After using the corresponding plaintext data D _i , the data demand quotient j provides corresponding data evaluation information according to the data usage effect and sends it to the Nth master node, where the data evaluation information includes: the Nth digest block The quality description and quality score of the data corresponding to the unique identification UID _i in the n+1th evaluation block;

Step 16, according to the received data evaluation information and the unique identifier UID _i , all the Nth master nodes record the data evaluation information to the n+1th of the evaluation side chain of the Nth digest block in the corresponding data digest chain. in the evaluation block;

Step 17: The data demand contract allocates the budget τ in the contract account to the account address α _i of the data provider i corresponding to the data in the transaction data pool dataUidList{} according to the data reward allocation rules in the contract.

The feature of the decentralized data transaction method of the present invention is that the functions in the data demand contract in the step 11 include:

The type contract function matchStyle() is used to check whether the data types match;

The quality contract function checkQuality() is used to check whether the quality score of the data meets the quality requirements of the data demand quotient. If the quality score of the data is greater than the data quality requirement β, it means that the quality requirements are met.

Compared with the prior art, the beneficial effects of the present invention are reflected in:

1. The decentralized big data trading system proposed by the present invention fundamentally solves the untrustworthy problem of traditional third-party trading centers and the single point of failure problem of the center, which greatly ensures the reliability of data providers in data transactions. Data Ownership and Right to Know.

2. The present invention uses blockchain technology to record the summary information and evaluation information of the data, as well as the transaction information, which not only ensures the verifiability of the data, but also improves the credibility of the data evaluation, and provides a more comprehensive service for all parties involved in big data transactions. immutable data transaction records.

3. The present invention uses smart contract technology to conduct data transactions, improves the intelligence and efficiency of transactions, and effectively improves the efficiency of data retrieval, matching and transactions, and the contract design provides data demanders with more diversified data solutions. Program.

4. The present invention utilizes the proxy re-encryption technology and the smart contract technology, and the encrypted public key of the data demander is released through the contract, which improves the correctness of the key conversion, prevents the data from being illegally decrypted during the transmission process, and ensures the security of the transaction data. .

5. The data evaluation mechanism and value reward rules designed in the present invention can encourage the high-quality development of data and further promote the sharing and circulation of big data.

Description of drawings

Fig. 1 is a system user diagram of the present invention;

Fig. 2 is a schematic diagram of block chain block information of the present invention;

FIG. 3 is a schematic diagram of the blockchain chain structure of the present invention;

Detailed ways

In this embodiment, as shown in Figure 1, an application scenario of a blockchain-based decentralized data transaction method includes three types of users: data providers, data demanders, and data agent nodes. The data provider stores the data At the data agent node; the agent node stores and forwards user data and is responsible for maintaining the blockchain for big data transactions, and writing data information into the blockchain. Data buyers obtain the required data by issuing smart contracts, give corresponding rewards to data providers through smart contracts, and feed back data evaluation information. The data summary information of the data provider is recorded through the blockchain, and the data demander can understand the basic information of the data through the data summary information. The smart contract can interact with the information in the blockchain to help the data demander match the corresponding data. data set. Specifically, the method is carried out as follows:

Step 1. Any data provider i obtains an anonymous identity public and private key pair (PK _i , SK _i ) and account address α _i through registration. The anonymous user scheme is adopted. The user only needs to pass a series of custom strings to The corresponding private key SK _i and public key PK _i are obtained, and the address α _i that accepts the reward can be generated through the public key. At the same time, according to the public key, the user can apply to the key generation center for a public-private key pair for proxy re-encrypted data. Therefore, for each user, there are two pairs of key pairs, one pair is the user's anonymous identity key pair (PK _i , SK _i ), and the other pair is the key pair used for encrypting and decrypting data (ski _i , pk ) _i ), a data provider can have multiple data sets for transactions. This anonymous registration method makes it impossible for data demanders to establish a connection between data and data users, which further enhances the protection of user privacy.

Step 2, data provider i obtains data ciphertext Epk _i (D _i ) after encrypting its own data Di, wherein, D _i is the plaintext data that data provider _i uses for big data transaction, Epk _i ( ) represents use the public key pk _i of the data provider i to perform the encryption operation;

The data provider i provides a data abstract about the plaintext data Di, including: the description Des(D _i ) of the plaintext data _{Di ,} the Hash value Hash(D _i ) of the plaintext data _Di , the data type Type(D _i ) , the number of data entries DT _i (D _i ), the signature information Sig _Ski , and the address α _i of the account receiving the data reward;

Step 3, the data provider i selects any agent node B, and sends the data ciphertext Epk _i (D _i ) and the data digest to the agent node B;

Step 4. After the agent node B receives the data ciphertext Epk _i (D _i ) and the data digest of the data provider i, it calculates a unique ID number UID _i by formula (1). This number UID _i can be used in big data Unique identification of the data in the trading system;

UID _i =Hash(α _i ,Hash(D _i )) (1),

In formula (1), Hash( ) represents hash calculation;

The proxy node is a data transaction node with certain storage and computing capabilities. In addition to storing the data of the data provider, it is also responsible for data re-encryption and maintenance of the blockchain system. The identity public and private key pair of the agent node B is represented as (SK _B , P _B ), after the agent node B receives the data of the data provider, it will generate the corresponding storage proof field SP (UID _i ) through formula (2);

是代理商节点B利用自身公私钥对(SK_B，PK_B)对存储证明SP(UID_i)的签名，AL_UIDi是数据提供商i的唯一标识UID_i的数据密文Epk_i(D_i)的存储位置，通过存储证明，数据提供商可以验证自己数据的存储有效性。In formula (2),

AL _UIDi is the data ciphertext Epk _i (D _i ) that uniquely identifies UID _i of the data provider i, which is the signature of the agent node B using its own public and private key pair (SK _B , PK _B ) to the storage proof SP (UID _i ) The storage location, through the storage proof, the data provider can verify the storage validity of its own data.

Step 5. The agent node B uses the formula (3) to generate the summary information T _brief (UID _i ) of the data UID _i :

Step 6. The agent node B obtains the data summary information generated by all agent nodes in the Nth time period, uses the obtained data summary information as the main part of the block, and adds the block header and packages it as the Nth summary block. , the information of the block header in the Nth digest block includes: the timestamp, the Merkle Tree of the data digest information, and the Hash value of the N-1th digest block;

After the agent node B reaches a consensus with other agent nodes, it links the Nth summary block to the N-1th summary block to form the latest data summary chain. When N=0, the N-1th summary block is formed. The Hash value of each digest block is the set Hash value; the digest block of the data digest chain is shown in Figure 2.

Step 7. Elect the N-th master node in the agent node corresponding to the data in the N-th summary block in the latest data summary chain, and the N-th master node generates the N-th evaluation on the N-th summary block. The nth evaluation block of the side chain, the link form of the data summary chain and the data evaluation chain are shown in Figure 3.

其中，

分别是第N个摘要区块上的第n个评价区块中唯一标识UID_i的所对应数据的完整性、一致性和准确性的评分；当n＝1时，

为所设定的完整性、一致性和准确性的初始评分；数据评价链的评价区块如图2所示。The block body part of the nth evaluation block contains the quality description and quality score of the data in the nth summary block in the nth evaluation block

in,

are the scores of the integrity, consistency and accuracy of the corresponding data uniquely identifying UID _i in the nth evaluation block on the nth summary block; when n=1,

It is the initial score of the set integrity, consistency and accuracy; the evaluation block of the data evaluation chain is shown in Figure 2.

The information of the block header of the nth evaluation block includes: timestamp, MerkleTree of the data evaluation information, and the Hash value of the n-1th evaluation block; when n=1, the Hash of the n-1th evaluation block The value is the hash value of the Nth digest block;

Step 8. Any data demander j obtains an anonymous identity public and private key pair (PK _j , SK _j ) through registration, and applies to the key generation center for a public and private key pair (pk _j , sk _j ) for encryption and decryption;

Step 9. The data demander j publishes the data demand contract through the smart contract. The data demand contract includes: the demand data type D _S , the number of data demand entries D _A , the data demand period dl, the data demand business encryption public key pk _j , and the data quality requirements β, and data reward distribution rules. The reward distribution rules are formulated by data demand quotients and can be allocated according to the rules in combination with the quality and quantity of demand quotient data.

After the data demander j deploys the data demand contract, it sends the data transaction budget τ to the contract account generated by the data demand contract. By prepaying the reward, the malicious behavior of the data demander who refuses to pay the reward after obtaining the data can be effectively avoided, and sending the reward to the contract account can realize the smart contract to automatically distribute the reward to the account of the data provider.

Step 10. All data providers interact with data demander j through data demand contracts to respond to data demand; the mode of data demander responding through contracts ensures the data provider's control over data transactions, and each time they participate in a transaction, All need to be confirmed by contract interaction with the data provider.

Step 11. Assuming that the data provider i responds to the data request, the data request contract reads the Nth digest block and the Nth data on the corresponding data digest chain according to the unique identifier UID _i of the data provider i that responds to the data request After evaluating the first n evaluation blocks on the side chain, according to the function in the data demand contract, a number of data of the data provider i responding to the data demand is matched and checked, and the data that meets the needs of the data demand provider j is added to the candidate data pool candidateDataList{};

The functions in the data request contract include:

The type contract function matchStyle() is used to check whether the data types match;

The quality contract function checkQuality() is used to check whether the quality score of the data meets the quality requirements of the data demand quotient. If the quality score of the data is greater than the data quality requirement β, it means that the quality requirements are met. The candidate data pool candidateDataList{} is to more accurately improve the matching degree of the data requirements of the data demander. It is expanded on the basis of the data entry of the data demander's demand, so that more data can become candidate data, such as the data of the candidate data pool. The number of entries is three times that of the data demander's demand data entry, so that the data demander can refer to more data information for the next selection;

Step 12. The data demander j checks the summary information and evaluation information of the data in the corresponding blockchain according to the unique identifier UID _i in the candidate data pool candidateDataList{}. By viewing the data summary information, it can help the demander understand the relevant attributes of the data itself , check the evaluation side chain information, the data demander can see the feedback information of each user before the data, and then the data demander j decides whether to buy the corresponding data, and if it chooses to buy, the corresponding data will be added In the transaction data pool dataUidList{}.

Step 13: _The data demand contract records the number of entries DR of data added to the transaction data pool _dataUidList {}. When the entry demand DA of the data demand quotient is satisfied, or the time limit of the data demand is reached, it indicates that the data demand matching is completed. Through the design of the candidate data pool and transaction data pool for data demanders, the data matching process adopts the mode of automatic matching of smart contracts and manual selection of data demanders. Smart contracts help to filter out the data that matches the needs of data demanders. Further selection favors the selection of higher quality data.

Step 14: The data demander j sends a data request to the data provider i of the data in the transaction data pool dataUidList{}, and the corresponding data provider _i uses its own decryption private key ski and the data demander j in the data demand contract. The encryption public key pk _j generates the proxy re-encryption key rk _i→j and sends it to the proxy node B that stores the data of the data provider i;

The corresponding agent node B uses the agent re-encryption key rk _i→j to re-encrypt the data ciphertext Epk _i (D _i ) of the data added by the data provider i to the transaction data pool dataUidList{} to obtain the re-encrypted ciphertext Epk′ _i (D _i ) is sent to data demander j, so that data demander j can decrypt the re-encrypted ciphertext Epk′ _i (D _i ) with its own private key sk _j , and obtain data provider i to join the transaction The plaintext data D _i of the data in the data pool dataUidList{}. In this process, only the data provider and data demander can obtain plaintext data, and the data is always kept in ciphertext state to the proxy node, thus reducing the risk of data tampering and theft;

The detailed execution steps of proxy re-encryption are:

Step 9.1, Enc(D _i , pk _i )→Epk _i (D _i ): the data provider i executes the data encryption algorithm Enc(), inputs the data plaintext D _i and the public key pk _i , and outputs the data ciphertext Epk _i (D i ) _i );

Step 9.2, Delegate(ski, pk _j )→rk _{i→j :} The data provider executes the re-encryption authorization algorithm Delegate(), inputs the private key ski and the public key pk _j of the data demander _j , and outputs the proxy re-encryption key rk _i→j ;

Step 9.3, ReEnc(rk _i→j , Epk _i (D _i ))→Epk′ _i (D _i ): the agent node executes the re-encryption algorithm ReEnc(), and inputs the data ciphertext encrypted by the public key p _i of the data provider Epk _i (D _i ) and the proxy re-encryption key rk _i→j , output the re-encrypted ciphertext Epk′ _i (D _i );

Step 9.4, Dec(Epk′ _i (D _i ), sk _j )→D _i : the data demand quotient executes the decryption algorithm Dec(), and inputs the re-encrypted ciphertext Epk′ _i (D _i ) and the data demand quotient to decrypt the private key sk _j , get the data plaintext D _i .

Step 15: The data demand quotient j calculates the corresponding Hash value Hash(D _i ) according to the plaintext data D _i obtained by decryption. If it is the same as the hash value of the corresponding plaintext data D _i in the data digest chain, it means that the decrypted plaintext is obtained. The data is the original data Di provided by the data provider _i .

After using the corresponding plaintext data D _i , the data demand quotient j gives the corresponding data evaluation information according to the data usage effect and sends it to the Nth master node. The data evaluation information includes: the unique identifier UID in the Nth abstract block The quality description and quality score of the data corresponding to _i in the n+1th evaluation block.

其中

表示第N个摘要区块中标识为UID_i的数据对应评价侧链第n个评价区块中的数据完整性、一致性和准确性的评分，t是单次数据评价分数的阈值，为了保证单次评价对数据的质量影响过大，这个反馈评价阈值t应小于初始评分，这样随着更多的数据提供商给出评价，数据的质量优劣将逐渐明显。The quality description is the use feedback in the process of data use, such as whether the feedback data is consistent with the data summary information provided by the demander, and whether the data meets its own data needs, etc.; the quality score is the score of the different data quality evaluation dimensions of the data. Symbols are expressed as

in

Indicates the score of the data integrity, consistency and accuracy in the nth evaluation block of the evaluation side chain corresponding to the data identified as UID _i in the Nth digest block, t is the threshold of the single data evaluation score, in order to ensure A single evaluation has a great impact on the quality of the data. The feedback evaluation threshold t should be smaller than the initial score, so that as more data providers give evaluations, the quality of the data will gradually become obvious.

Due to its uniqueness, big data has some differences from commonly used commodities. Therefore, this method reflects the data quality through multi-dimensional scoring. The big data transaction scheme evaluates the three dimensions of data integrity, consistency and accuracy. , Integrity refers to whether the data of the data itself is valid data. For example, for a structured database, the integrity of the data includes domain integrity, entity integrity and referential integrity. Consistency refers to the integrity of the data and the blockchain. Whether the data summary information is consistent, accuracy is whether the data will have a significant impact in the process of use, such as data used for machine learning, the results may have a significant impact due to inaccurate data.

Step 16, according to the received data evaluation information and the unique identifier UID _i , all the Nth master nodes record the data evaluation information to the n+1th of the evaluation side chain of the Nth digest block in the corresponding data digest chain. in the evaluation block;

Step 17: The data demand contract allocates the budget τ in the contract account to the account address α _i of the data provider i corresponding to the data in the transaction data pool dataUidList{} according to the data reward allocation rules in the contract.

The following scheme provides a reward distribution rule, and the total quality score of the data uniquely identified as UID _i is calculated according to formula (4) as:

When the smart contract distributes the reward, the sum of the three dimensions is used as the total quality score DL _i of the data. The smart contract distributes the rewards for the data of the data provider. The reward distribution is calculated by the number of items of the comprehensive data and the data quality score. The data D _i of the data provider A should be rewarded by formula (5):

In formula (5), τ is the data reward paid in advance by the data demander, and this reward is stored in the contract address before the contract execution of the big data transaction. DT _i is the number of entries in the data Di. If the corresponding data set is a picture, then DT _i represents how many pictures there are. If the corresponding data is a traditional database, then DT _i represents the number of tuples. This allocation rule can allocate relatively higher data rewards to high-quality data, encourage data providers of high-quality data to participate in big data transactions, and further promote the sharing and trading of big data.

Example:

The data demand quotient j requires the data type D _S to be the monitoring data about running of the smart bracelet, the entries of the demand data are D _A =1000, the data quality score requires β≥6.0, and the size of the candidate pool is N ₌ 3×DA . The data demander j publishes the data demand through a smart contract, the contract contains the above fields, and pays the reward τ=1000 for the payment data set to the contract account in advance.

In the blockchain-based big data trading platform, multiple data providers own this type of data. After the data demand contract is released, the data provider can summarize the blockchain data through the contract address of the smart contract and the provided interface. The information is sent to the smart contract to realize the authorization of the data transaction, agreeing to the data to participate in the transaction, and the smart contract will match the data. Table 1 represents the data information of the data providers authorized to participate in this data transaction.

Table 1

The smart contract first filters the data into the candidate data pool. The provider public key number 1 is PK ₁ and the data UID is HG89K8JHKJK7, which does not meet the data quality score requirements of the data demand provider; the provider public key number 8 is PK 8 and the data UID is PK ₈ and the data UID The data type of 3HJH9JHG6DJJ does not meet the requirements of the data demand quotient type. The size of the candidate pool is 3000, so the data numbered 2, 3, 4, 5, 6, and 7 enter the candidate data pool. The data demand quotient j is based on the corresponding UID of the data By querying the data summary chain and data evaluation chain, you can learn the detailed description and feedback of the data. The demander j selects the data numbered 3, 4, and 7. The smart contract monitors the number of data entries to meet the demand, and the execution of the demand contract ends.

The data demander j sends the public key and data access request to the proxy nodes B2, B3, and B6 where the data numbered 3, 4, and 7 are stored. These proxy nodes will request the data provider for re-encryption authorization, and the corresponding data provider uses its own The private key and the public key of the data demander generate the proxy re-encryption key, the proxy node performs the re-encryption operation, and sends the data numbered 3, 4, and 7 to the data demander j, and the smart contract will pre-store the reward in the contract address. According to the distribution rules, the distributed rewards are calculated as follows:

The data demander j decrypts the data with its own public key, and gives feedback and evaluation on the data.

Claims (2)

1. A blockchain-based decentralized data transaction method, characterized in that it is applied in a transaction environment composed of a key generation center, a blockchain and a smart contract, and in the transaction environment, the steps are as follows Implement a peer-to-peer transaction method between several data providers, one data demander, and several data agent nodes: Step 1. Any data provider i obtains an anonymous identity public-private key pair (PK _i , SK _i ) and an account address α _i through registration, and applies to the key generation center for a public-private key pair (pk i , SK i ) for encryption and decryption. _i ,ski ₎ ; Step 2, the data provider i obtains the data ciphertext Epk _i (D _i ) after encrypting its own data Di, wherein, D _i is the plaintext data that the data provider _i is used for the big data transaction, Epk _i ( ) indicates that the encryption operation is performed using the public key pk _i of the data provider i; The data provider i provides a data digest about the plaintext data D _i , including: a description Des(D _i ) of the plaintext data D _i , a Hash value Hash(D _i ) of the plaintext data D _i , data Type Type(D _i ), number of data entries DT _i (D _i ), signature information Sig _Ski , and account address α _i for receiving data rewards; Step 3, the data provider i selects any agent node B, and sends the data ciphertext Epk _i (D _i ) and the data digest to the agent node B; Step 4. After the agent node B receives the data ciphertext Epk _i (D _i ) and the data digest of the data provider i, it uses formula (1) and formula (2) to generate the unique identifier UID _i and store it respectively. Prove SP(UID _i ): UID _i =Hash(α _i ,Hash(D _i )) (1) In formula (1), Hash( ) represents hash calculation;

In formula (2),

is the signature of the storage proof SP ( _{UID i} ) by the agent node B using its own public and private key pair (SK _B , PK _{B ) . i} ) the storage location; Step 5. The agent node B uses formula (3) to generate data summary information T _brief (UID _i ):

Step 6. The agent node B obtains the data summary information generated by all agent nodes in the Nth time period, uses the obtained data summary information as the main part of the block, and adds the block header and packages it as the Nth summary. block, the information of the block header in the Nth digest block includes: timestamp, the Merkle Tree of the data digest information and the Hash value of the N-1th digest block; After the agent node B reaches a consensus with other agent nodes, it links the Nth summary block to the N-1th summary block to form the latest data summary chain. When N=0, the N-1th summary block is formed. The hash value of each digest block is the set hash value; Step 7. Elect the Nth master node from the agent nodes corresponding to the data in the Nth digest block in the latest data digest chain, and the Nth master node generates the Nth master node on the Nth digest block. The nth evaluation block of the evaluation side chain, the block body part of the nth evaluation block contains the quality description and quality score of the data in the Nth digest block in the nth evaluation block

in,

are the scores of the integrity, consistency and accuracy of the corresponding data uniquely identifying UID _i in the nth evaluation block on the nth summary block; when n=1,

An initial score for completeness, consistency and accuracy as set; The information of the block header of the nth evaluation block includes: timestamp, Merkle Tree of the data evaluation information and the Hash value of the n-1th evaluation block; when n=1, the n-1th evaluation area The hash value of the block is the hash value of the Nth digest block; Step 8. Any data demander j obtains an anonymous identity public and private key pair (PK _j , SK _j ) through registration, and applies to the key generation center for a public and private key pair (pk _j , sk _j ) for encryption and decryption ; Step 9, the data demander j publishes a data demand contract through the smart contract, the data demand contract includes: demand data type D _S , data demand item number D _A , data demand period dl, data demand provider encryption public key pk _j , data quality requirement β, and data reward distribution rules; After the data demander j deploys the data demand contract, it sends the data transaction budget τ to the contract account generated by the data demand contract; Step 10, all data providers interact with the data demander j through the data demand contract, thereby responding to the data demand; Step 11. Assuming that the data provider i responds to the data request, the data request contract reads the Nth digest block and the Nth digest block on the corresponding data digest chain according to the unique identifier UID _i of the data provider i that responds to the data request. After the first n evaluation blocks on the data evaluation side chain, according to the function in the data demand contract, a number of data of the data provider i responding to the data demand is matched and checked, and the data that meets the requirements of the data demand provider j will be checked. The data is added to the candidate data pool candidateDataList{}; Step 12, the data demander j checks the summary information and evaluation information of the data in the corresponding blockchain according to the unique identifier UID _i in the candidate data pool candidateDataList{}, and then the data demander j decides whether to purchase the corresponding data, if If you choose to buy, the corresponding data will be added to the transaction data pool dataUidList{}; Step 13, the data demand contract records the number of entries DR added to the data in the transaction data pool _{dataUidList {} }, when the entry demand DA of the data demand quotient is satisfied, or when the time limit of the data demand is reached, it indicates the data demand matching completed; Step 14, the data demander j sends a data request to the data provider i of the data in the transaction data pool dataUidList{}, and the corresponding data provider _i decrypts the private key ski and the data in the data demand contract through its own. The encrypted public key pk _{j of the demander j} generates a proxy re-encryption key rk _i→j and sends it to the proxy node B that stores the data of the data provider i; The corresponding agent node B uses the agent re-encryption key rk _i→j to re-encrypt the data ciphertext Epk _i (D _i ) of the data added by the data provider i to the transaction data pool dataUidList{}, and obtain the re-encrypted The ciphertext Epk _i '(D _i ) is sent to the data demander j, so that the data demander j can use its own private key sk _j to decrypt the re-encrypted ciphertext Epk _i '(D _i ), and obtain The data provider i joins the plaintext data D _i of the data in the transaction data pool dataUidList{}; Step 15, the data demand quotient j calculates the corresponding hash value Hash(D _i ) according to the plaintext data D _i obtained by decryption, if it is the same as the hash value of the corresponding plaintext data D _i in the data digest chain, it means The decrypted plaintext data D _i is the original data provided by the data provider i; After using the corresponding plaintext data D _i , the data demand quotient j provides corresponding data evaluation information according to the data usage effect and sends it to the Nth master node, where the data evaluation information includes: the Nth digest block The quality description and quality score of the data corresponding to the unique identification UID _i in the n+1th evaluation block; Step 16, according to the received data evaluation information and the unique identifier UID _i , all the Nth master nodes record the data evaluation information to the n+1th of the evaluation side chain of the Nth digest block in the corresponding data digest chain. in the evaluation block; Step 17: The data demand contract allocates the budget τ in the contract account to the account address α _i of the data provider i corresponding to the data in the transaction data pool dataUidList{} according to the data reward allocation rules in the contract. 2. The decentralized data transaction method according to claim 1, wherein the function in the data demand contract in the step 11 comprises: The type contract function matchStyle() is used to check whether the data types match; The quality contract function checkQuality() is used to check whether the quality score of the data meets the quality requirements of the data demand quotient. If the quality score of the data is greater than the data quality requirement β, it means that the quality requirements are met.

Images