A blockchain consensus method with security perception and response strategy

Abstract

This invention discloses a blockchain consensus method with security awareness and response strategies, comprising two parts: data consistency confirmation and establishing a credibility judgment model. Data consistency confirmation includes: S1. Electing packaging nodes within the blockchain shards. These packaging nodes are responsible for packaging block information and sending it to the consensus committee for verification; S2. The consensus committee nodes verify the block information and generate aggregate opinions, then use the credibility judgment model to determine the aggregate opinions and form a judgment result; S3. Each blockchain node verifies the validity of the judgment result and updates the status of its local blockchain shard. Establishing the credibility judgment model includes: S11. Collecting historical blockchain data and constructing a dataset based on the consensus committee's reputation, aggregate opinions, and block information as feature inputs for the credibility judgment model; S12. Constructing the credibility judgment model; S13. Calculating the accuracy evaluation index of the credibility judgment model to complete the evaluation of the credibility judgment model.

Description

Technical Field

[0001] This invention relates to the field of blockchain technology, and in particular to a blockchain consensus method with security awareness and response strategies. Background Technology

[0002] With the rapid development of IoT technology, the time-series data generated by IoT devices is also growing explosively. However, many security issues exist in the storage and sharing of IoT data, such as data leakage and data tampering. Blockchain technology, with its decentralized and immutable characteristics, provides a promising solution for the storage and sharing of IoT data and is gradually becoming an important infrastructure in the era of the Internet of Everything.

[0003] In the Industrial Internet of Things (IIoT), device stability often varies across different time periods and business responses. Therefore, consensus mechanism design should prioritize addressing security risks arising from dynamic changes in node behavior while simultaneously improving throughput. In hybrid consensus, some nodes are elected to form a committee representing all nodes in the consensus process. Blocks are verified by the committee, broadcast to other nodes, and then directly added to the blockchain, increasing consensus speed and system throughput. However, while the security of this consensus process relies on the committee's trustworthiness, a widespread denial-of-service attack or attack leading to erroneous verification can paralyze the consensus. Therefore, establishing a secure, efficient, and reliable consensus mechanism is a major research focus in blockchain today. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and optimize blockchain sharding while maintaining a certain degree of decentralization. It provides a blockchain consensus method with security awareness and response strategies, implementing a consensus security awareness mechanism through a neural network binary classification model and optimizing the block verification method. Based on the consensus security awareness results, a dynamic consensus switching method is designed to achieve a consensus mechanism that balances security and efficiency.

[0005] The objective of this invention is achieved through the following technical solution:

[0006] A blockchain consensus method with security awareness and response strategies includes two parts: data consistency confirmation and establishing a trustworthiness determination model. The data consistency confirmation includes:

[0007] S1. According to the Randhound protocol, the blockchain nodes in the blockchain system are randomly divided into different blockchain shards. Then, within each blockchain shard, a packaging node is elected. The packaging node is responsible for packaging the block information and sending it to the consensus committee for verification.

[0008] S2. The consensus committee nodes verify the block information and generate aggregate opinions, then use a credibility judgment model to judge the aggregate opinions and form a judgment result;

[0009] S3. Each blockchain node verifies the validity of the judgment result and updates the status of its local blockchain shard;

[0010] The establishment of the credibility determination model includes:

[0011] S11. Collect historical blockchain data and construct a dataset based on the consensus committee's reputation, aggregated opinions, and block information as feature inputs for the credibility determination model;

[0012] S12. Construct a credibility determination model, divide the dataset into training and test sets, and optimize the hyperparameters based on the results of k-fold cross-validation.

[0013] S13. Verify the effectiveness of the credibility judgment model through the test set, construct a confusion matrix based on the output results of the credibility judgment model and the real labels, calculate the accuracy evaluation index of the credibility judgment model, and complete the evaluation of the credibility judgment model; the real labels are obtained based on historical blockchain data, and the accuracy evaluation index of the credibility judgment model includes accuracy, recall, and F1-Score.

[0014] Furthermore, blockchain nodes include ordinary nodes, mining nodes, block packaging nodes, consensus nodes, committee nodes, and leadership nodes within the committee.

[0015] Furthermore, step S1 specifically includes:

[0016] (101) Blockchain nodes within a blockchain shard elect consensus nodes through voting. Each blockchain node has one vote to elect a representative or become a representative to run for election. The top n representatives with the most votes are selected as consensus nodes. The consensus nodes form a consensus committee and verify the latest block generated by the blockchain shard. A leader node is randomly generated within the consensus committee to collect and broadcast the consensus results of the consensus committee. n is the number of committee nodes in the consensus committee, and it is an odd number.

[0017] (102) Within a blockchain shard, mining nodes compete for the right to package the latest block through competitive or delegated consensus mechanisms, becoming block packaging nodes. A Competitive consensus mechanisms include PoW and PoS consensus mechanisms, while delegated consensus mechanisms include DPoS consensus mechanism.

[0018] (103) Block Packaging Node A Broadcast a request message <request, Block> to the consensus committee N`sign>` is used to transmit the latest block and notify the start of a new round of consensus, where `request` is the request message content, and `Block` is the block. N As the latest block, it needs to be verified by the committee nodes; sign is the Node. A Signature of the request message.

[0019] Furthermore, step S2 specifically includes:

[0020] (201) Committee nodes within the consensus committee i For the latest block N Verify the correctness and generate a verification opinion msg. i And send it to its leader node, 0 < i ≤ n, where n is the number of committee nodes in the consensus committee;

[0021] (202) The leader node will connect all committee nodes in the consensus committee. i For Block N Verification comments msg i Integrate and form a consensus opinion.<MSG,hasg,aggregatesign> And integrate aggregated opinions into the Block N In the aggregated opinions, MSG represents the summation of validation opinions from all nodes in the committee, where MSG = [msg1, msg2, ... msg]. i ,...,msg n ], msg i For committee node i For Block N The verification opinion takes values ​​of 0, 1, and X, representing support, opposition, and absence, respectively. aggregatesign is the aggregated signature generated by the leader node after aggregating the signatures of the consensus committee nodes.

[0022] (203) The leader node broadcasts the aggregated opinion to other blockchain nodes in the blockchain shard;

[0023] (204) The credibility judgment model judges the aggregated opinions and forms a judgment result. All blockchain nodes verify the validity of the judgment result.

[0024] Furthermore, step S3 specifically includes:

[0025] (301) If other nodes Node other If the verification result is valid, then the latest block will be added. NAdd directly to the blockchain and update the reputation value of the committee nodes; if Node other If the verification result is invalid, then the latest block will be... N The verification is performed, and the verification result is broadcast to all blockchain nodes. An impeachment motion is then initiated against the consensus committee.

[0026] (302) If impeachment succeeds, a high-security consensus mechanism is switched, all nodes within the blockchain shard verify the block and update the blockchain based on the result, triggering the consensus committee iteration mechanism; if impeachment fails, the Block... N As a valid block, and the Block N It is added to the blockchain, and the reputation value of the committee nodes is updated simultaneously; the highly secure consensus mechanism is PoW consensus.

[0027] Furthermore, step S11 specifically includes:

[0028] (1101) Construct the dataset D = {(x (1) ,y (1) ),(x (2) ,y (2) ),...,(x (N) ,y (N) )}, input vector Let m be the feature vector of the m-th round, and reputation be... i Represents a committee node i Reputation value, msg i For committee node i For Block m Verification opinion; tag y (N) This indicates the validity of the consensus result in the Nth round, with a value of either valid or invalid.

[0029] (1102) Use the MinMax normalization method to normalize the reputation values ​​of all consensus committees, map the reputation values ​​to decimals between 0 and 1, and convert the dimensional expressions into dimensionless expressions.

[0030] Furthermore, step S12 specifically includes:

[0031] (1201) Initialize the credibility determination model. The credibility determination model is divided into an input layer, an output layer, and a hidden layer. The input layer and the output layer are 1 layer, and the hidden layer is 3 layers.

[0032] (1202) The training set is divided into k equal-sized subsets using k-fold cross-validation. During training, these k subsets are traversed sequentially, with the current subset used as the validation set and all other subsets used as the training set for training and evaluating the credibility determination model. The hyperparameters are adjusted based on the performance on the current subset, thereby improving the generalization ability of the credibility determination model. Performance is defined by evaluation metrics such as accuracy and recall. If the training set performs well but the validation set performs poorly, it indicates poor generalization ability, and the regularization parameters need to be adjusted. If both the training and validation sets perform poorly, the size of the training set is increased.

[0033] Furthermore, step S13 specifically includes:

[0034] (1301) Taking the features of each dimension of the test set as input, the credibility judgment model will output the consensus security judgment result;

[0035] (1302) Construct a confusion matrix by combining the output of the credibility judgment model with the true label values ​​of the test set, and calculate the accuracy, recall and F1-Score typical indicators through the confusion matrix to complete the evaluation of the credibility judgment model.

[0036] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor, when executing the program, implements the steps of the blockchain consensus method with security awareness and response strategies described above.

[0037] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the blockchain consensus method with security awareness and response strategies.

[0038] Compared with the prior art, the beneficial effects of the technical solution of the present invention are:

[0039] 1. Compared to hybrid consensus, where the security of the consensus process is guaranteed by the trustworthiness of the committee, consensus paralysis can occur when the committee experiences widespread denial of service or is attacked, leading to erroneous verification. This invention, however, analyzes the attributes and behaviors of consensus nodes, perceives the consensus security situation, and adaptively switches consensus mechanisms. This allows for timely detection of attacks on the consensus process, preventing consensus paralysis and better protecting the security of smaller consensus groups after sharding.

[0040] 2. The consensus security awareness of the present invention can be combined with the self-organizing sharding strategy, which enables nodes to quickly start the re-sharding mechanism when consensus fails, reselect trusted domain nodes to construct shards, avoid long-term consensus failure, and improve the security of the consensus mechanism.

[0041] 3. Compared with the traditional consensus algorithm PoW, this invention uses a consensus committee instead of all blockchain nodes to reach consensus, greatly improving system throughput while maintaining a certain degree of decentralization. Furthermore, the consensus mechanism of this invention will not cause the blockchain system to fork and can effectively resist double-spending attacks.

[0042] 4. When committee nodes verify blocks, this invention uses aggregated signatures to broadcast committee opinions, which effectively reduces communication and storage overhead during the consensus process and improves the scalability of the blockchain system.

[0043] 5. This invention uses the MinMax normalization method to normalize the reputation values ​​of all consensus committees. Normalization can effectively improve the convergence speed and accuracy of the credibility determination model. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of the data consistency verification process.

[0045] Figure 2 A schematic diagram of the process for building a credibility assessment model. Detailed Implementation

[0046] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only for explaining the present invention and are not intended to limit the present invention.

[0047] This embodiment provides a highly reliable and scalable blockchain consensus method with security awareness and response strategies, comprising two parts: data consistency confirmation and trustworthiness determination model establishment. Data consistency confirmation includes three parts: block generation, block verification, and block confirmation. Details are as follows:

[0048] Step 1: Within each blockchain shard, a packaging node is elected. This packaging node is responsible for packaging block information and sending it to the consensus committee for verification. The specific process is as follows:

[0049] (101) The blockchain system randomly divides blockchain nodes into different blockchain shards according to the Randhound protocol. The Randhound protocol is a scalable and secure multi-party computation protocol that can provide unbiased distributed randomness in the Byzantine protocol. Blockchain nodes include ordinary nodes, mining nodes, block packaging nodes, consensus nodes, committee nodes, and leader nodes within the committee.

[0050] (102) Within the blockchain shard, the blockchain nodes elect consensus nodes through voting. Each blockchain node has one vote to elect a representative or become a representative to participate in the election. The top n representatives with the most votes are selected as consensus nodes. The consensus nodes form a consensus committee and verify the latest block generated by the blockchain shard. A leader node is randomly generated within the consensus committee to collect and broadcast the consensus results of the consensus committee. n is the number of committee nodes in the consensus committee, which is usually an odd number.

[0051] (103) Within a blockchain shard, mining nodes compete for the right to package the latest block through competitive or delegated consensus mechanisms, becoming block packaging nodes. A Competitive consensus mechanisms include PoW and PoS, while delegated consensus mechanisms include DPoS.

[0052] (104) Block Packaging Node A Broadcast a request message to the consensus committee. <request, Block N `sign>` is used to transmit the latest block and notify the start of a new round of consensus, where `request` is the request message content, and `Block` is the block name. N As the latest block, it needs to be verified by the committee nodes; sign is the Node. A Signature of the request message.

[0053] Step 2: The consensus committee nodes verify the block information and generate aggregate opinions. The specific process is as follows:

[0054] (201) Committee nodes within the consensus committee i For the latest block N Verify the correctness and generate a verification opinion msg. i And send it to its leader node (0 < i ≤ n, where n is the number of committee nodes in the consensus committee).

[0055] (202) The leader node will connect all committee nodes in the consensus committee. i For Block N Verification comments msg i Perform integration and form an aggregated opinion = <MSG, hash, aggregatesign>, and integrate the aggregated opinion into the Block. N In the aggregated opinions, MSG represents the summary result of the validation opinions of all nodes in the committee, in the form of MSG = [msg1, msg2, ... msg]. i ,...,msg n ], where n is the number of committee nodes in the consensus committee. msgi For committee node i For Block N The verification opinion takes values ​​of 0, 1, and X, representing support, opposition, and absence, respectively. aggregatesign is the aggregated signature generated by the leader node after aggregating the signatures of the consensus committee nodes.

[0056] (203) The leader node broadcasts the aggregated opinion to other blockchain nodes in the blockchain shard.

[0057] (204) The credibility judgment model judges the aggregated opinions and forms a judgment result. All blockchain nodes verify the validity of the judgment result.

[0058] Step 3: Each blockchain node verifies the validity of the judgment result and updates the status of its local blockchain shard; the specific process is as follows:

[0059] (301) If other nodes Node other If the verification result is valid, then the latest block will be added. N This is done by directly adding the data to the blockchain and updating the reputation values ​​of committee nodes, thereby simplifying the block verification process, improving consensus efficiency and system throughput, and ensuring the scalability of the blockchain system. If Node... other If the verification result is invalid, then the latest block will be... N The verification is performed, and the verification results are broadcast to all blockchain nodes, and an impeachment motion is initiated against the consensus committee.

[0060] (302) If impeachment succeeds, a high-security consensus mechanism is switched, all nodes within the block verify the block and update the blockchain based on the result, triggering the consensus committee iteration mechanism. If impeachment fails, then the Block... N As a valid block, and the Block N Added to the blockchain, simultaneously updating the reputation values ​​of committee nodes. High-security consensus mechanisms such as PoW consensus.

[0061] The credibility determination model treats the validity determination of consensus security as a classification problem. A binary classifier is constructed using a neural network model. The binary classifier verifies the validity of the determination result and outputs the results as valid or invalid.

[0062] Step 11: Collect historical blockchain data and construct a dataset based on the consensus committee's reputation, aggregated opinions, and block information as feature inputs for the credibility determination model.

[0063] (1101) Construct the dataset D = {(x (1) ,y(1) ),(x (2) ,y (2) ),...,(x (N) ,y (N) )}, input vector Let m be the feature vector of the m-th round, and reputation be... i Represents a committee node i Reputation value, msg i For committee node i For Block m Verification opinion. Tag y (N) This indicates the validity of the consensus result in the Nth round, with a value of either valid or invalid.

[0064] (1102) For data feature preprocessing, the MinMax normalization method is used to normalize the reputation values ​​of all consensus committees. First, the reputation values ​​are mapped to decimals between 0 and 1, and at the same time, the dimensional expressions are converted into dimensionless expressions. Normalization can effectively improve the convergence speed and accuracy of the credibility determination model.

[0065] Step 12: Construct a credibility determination model, divide the dataset into training and test sets, and optimize the hyperparameters based on the results of k-fold cross-validation.

[0066] (1201) Initialize the credibility determination model, which consists of an input layer, an output layer, and a hidden layer. The input and output layers are 1 layer each, and the hidden layers are 3 layers.

[0067] (1202) The training set is divided into k equal-sized subsets using k-fold cross-validation. During training, these k subsets are traversed sequentially. Each time, the current subset is used as the validation set, and all other samples are used as the training set for training and evaluating the credibility judgment model. The hyperparameters are adjusted based on the performance on the current subset to improve the generalization ability of the credibility judgment model. If the training set performs well but the validation set performs poorly, it indicates poor generalization ability, and the regularization parameter needs to be adjusted. If both the training set and validation set perform poorly, the size of the training set is increased.

[0068] Step 13: Verify the effectiveness of the credibility determination model using a test set. Construct a confusion matrix based on the output of the credibility determination model and the real labels. Calculate the accuracy evaluation index of the credibility determination model to complete the evaluation of the credibility determination model. The real labels are obtained based on historical blockchain data. The accuracy evaluation index of the credibility determination model includes precision, recall, and F1-Score.

[0069] (1301) The effectiveness of the credibility judgment model needs to be evaluated by unknown new data. The test set plays this role. The credibility judgment model will output a consensus security judgment result by taking the features of the test set as input.

[0070] (1302) Construct a confusion matrix by combining the output of the credibility judgment model with the real label values ​​of the test set, and calculate typical indicators such as accuracy, recall and F1-Score through the confusion matrix to complete the evaluation of the credibility judgment model.

[0071] Finally, it should be noted that the above examples are only used to illustrate the calculation process of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing examples, those skilled in the art should understand that modifications can still be made to the calculation process described in the foregoing examples, or equivalent substitutions can be made to some of the parameters. Such modifications or substitutions do not cause the essence of the corresponding calculation method to deviate from the spirit and scope of the calculation method of the present invention.

[0072] This invention is not limited to the embodiments described above. The above description of specific embodiments is intended to illustrate and explain the technical solutions of this invention. The specific embodiments described above are merely illustrative and not restrictive. Without departing from the spirit and scope of the claims, those skilled in the art can make many specific modifications based on the teachings of this invention, and these modifications all fall within the scope of protection of this invention.

Claims

1. A blockchain consensus method with security awareness and response strategies, characterized in that, It includes two parts: data consistency verification and establishing a credibility determination model. The data consistency verification includes: S1. According to the Randhound protocol, the blockchain nodes in the blockchain system are randomly divided into different blockchain shards. Then, within each blockchain shard, a packaging node is elected. The packaging node is responsible for packaging the block information and sending it to the consensus committee for verification. S2. The consensus committee nodes verify the block information and generate aggregate opinions, then use a credibility determination model to determine the aggregate opinions and form a determination result; including: (201) Committee nodes within the consensus committee For the latest block Verify the correctness and generate verification opinions. And send it to its leader node, This represents the number of committee nodes in the consensus committee. (202) The leader node will connect the various committee nodes in the consensus committee. right Verification opinion Integrate and form a consensus. ， And aggregate opinions Integrated into In the aggregated opinions, MSG represents the summarized results of the validation opinions of all nodes in the committee. , For committee nodes right The verification opinion takes values ​​of 0, 1, and X, representing support, opposition, and absence, respectively. The aggregate signature is generated by the leader node after aggregating the signatures of the consensus committee nodes; (203) The leadership node will aggregate opinions. Broadcast to other blockchain nodes in the blockchain shard; (204) The credibility judgment model judges the aggregated opinions and forms the judgment result. All blockchain nodes verify the judgment results. Validity; S3. Each blockchain node verifies the validity of the judgment result and updates the status of its local blockchain shard; including: (301) If other nodes Verification and judgment results If valid, then the latest block will be... Add directly to the blockchain and update the reputation value of the committee nodes; if Verification and judgment results If invalid, then for the latest block The verification is performed, and the verification result is broadcast to all blockchain nodes. An impeachment motion is then initiated against the consensus committee. (302) If impeachment is successful, a high-security consensus mechanism is switched, all nodes within the blockchain shard verify the block and update the blockchain based on the result, triggering the consensus committee iteration mechanism; if impeachment fails, then For a valid block, and will Added to the blockchain, simultaneously updating the reputation value of committee nodes; the highly secure consensus mechanism is PoW consensus; The establishment of the credibility determination model includes: S11. Collect historical blockchain data and construct a dataset based on the consensus committee's reputation, aggregated opinions, and block information as feature inputs for the credibility determination model; S12. Construct a credibility determination model, dividing the dataset into training and test sets, to... k The hyperparameters were optimized based on the results of cross-validation. S13. Verify the effectiveness of the credibility judgment model through the test set, construct a confusion matrix based on the output of the credibility judgment model and the real labels, calculate the accuracy evaluation index of the credibility judgment model, and complete the evaluation of the credibility judgment model; the real labels are obtained based on historical blockchain data, and the accuracy evaluation index of the credibility judgment model includes accuracy, recall, and F1-Score.

2. The blockchain consensus method with security awareness and response strategy according to claim 1, characterized in that, Blockchain nodes include ordinary nodes, mining nodes, block packaging nodes, consensus nodes, committee nodes, and leadership nodes within the committee.

3. The blockchain consensus method with security awareness and response strategy according to claim 1, characterized in that, Step S1 specifically includes: (101) Within a blockchain shard, blockchain nodes elect consensus nodes through voting. Each blockchain node has one vote to elect a representative or become a representative to run for election. Before receiving a vote... Representatives are selected as consensus nodes. These nodes form a consensus committee, which verifies the latest blocks generated by blockchain sharding. A leader node is randomly selected within the consensus committee to collect and broadcast the committee's consensus results. The number of committee nodes in the consensus committee is an odd number. (102) Within a blockchain shard, mining nodes compete for the right to package the latest block through competitive or delegated consensus mechanisms, thus becoming block packaging nodes. Competitive consensus mechanisms include PoW and PoS consensus mechanisms, while delegated consensus mechanisms include DPoS consensus mechanism. (103) Block Packaging Node Broadcast a request message to the consensus committee. Used to transmit the latest block and notify the start of a new round of consensus, among which It is the request message content. As this is the latest block, it needs to be verified by the committee nodes. yes Signature of the request message.

4. The blockchain consensus method with security awareness and response strategy according to claim 1, characterized in that, Step S11 specifically includes: (1101) Constructing the dataset Input vector , For the first The eigenvectors of the wheel Indicates committee node Reputation value For committee nodes right Verification opinion; label Indicates the first The validity of the consensus result is determined by whether it is valid or invalid. (1102) Use The normalization method normalizes the reputation values ​​of all consensus committees, mapping the reputation values ​​to decimals between 0 and 1, while converting dimensional expressions into dimensionless expressions.

5. The blockchain consensus method with security awareness and response strategy according to claim 1, characterized in that, Step S12 specifically includes: (1201) Initialize the credibility determination model. The credibility determination model is divided into an input layer, an output layer, and a hidden layer. The input layer and the output layer are 1 layer, and the hidden layer is 3 layers. (1202) adopted k The cross-validation method divides the training set into... k The training process iterates through k subsets of equal size, using the current subset as the validation set and all other subsets as the training set for training and evaluating the credibility determination model. The hyperparameters are then adjusted based on the performance on the current subset, thereby improving the generalization ability of the credibility determination model.

6. The blockchain consensus method with security awareness and response strategy according to claim 1, characterized in that, Step S13 specifically includes: (1301) Taking the features of each dimension of the test set as input, the credibility judgment model will output the consensus security judgment result; (1302) Construct a confusion matrix by combining the output of the credibility judgment model with the true label values ​​of the test set, and calculate the accuracy, recall and F1-Score typical indicators through the confusion matrix to complete the evaluation of the credibility judgment model.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the blockchain consensus method with security awareness and response strategy as described in any one of claims 1 to 6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the blockchain consensus method with security awareness and response strategies as described in any one of claims 1 to 6.

Images