Data space consensus method based on improved DPoS

By dividing data space in the DPoS system and establishing a multi-level consensus mechanism, the problems of election results manipulation and misbehavior in the DPoS system are solved, the fairness and stability of the system are enhanced, and the decentralization and security of the data space are ensured. .

CN119030845AInactive Publication Date: 2024-11-26SHANGHAI JINRON DIGITS TECHNOLOGY CO LTD
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411035291.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-31
Publication Date
2024-11-26
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

There are problems in the DPoS system with manipulation and misconduct of election results, and the lack of effective supervision mechanisms, resulting in centralization risks and inefficiency, affecting the decentralization and security of the system.

Method used

By dividing the data space into subspaces, each subspace is managed and supervised by super nodes, select candidate nodes based on the initial trust value and random perturbation, and establish a multi-level consensus mechanism, including campaign rules, voting behavior, reward mechanism and supervision mechanism, to ensure that Fairness and supervisibility.

Benefits of technology

Enhance the fairness, immutability and supervisibility of the data space, avoid data blowout during the voting period, ensure the stability and continuity of node work, and ensure the data space through the supervision mechanism tolerate deviations under a certain threshold. working stability and continuity.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119030845A_ABST
    Figure CN119030845A_ABST
Patent Text Reader

Abstract

The invention discloses a data space consensus method based on an improved DPoS, and relates to the technical field of block chains, on the basis of an original DPoS, data space division is carried out, a multi-level consensus mechanism is constructed, the multi-level consensus mechanism comprises an election rule, a voting behavior, a reward mechanism, a block-out probability and a supervision mechanism, a super node is further selected through an elected candidate node, and a consensus result is obtained. And the whole data space nodes are supervised according to a supervision mechanism, and the candidate nodes and the super nodes are updated according to a period, so that the working stability and continuity of the data space nodes are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of deep learning and intelligent resources, and in particular to a power transmission component fault identification system based on deep learning. Background Art

[0002] Since the advent of Bitcoin, blockchain technology has quickly become the focus of the technology and financial circles. As a decentralized distributed ledger technology, blockchain ensures the immutability and transparency of data through encryption algorithms and distributed storage. This feature makes blockchain show great potential in many fields such as finance, Internet of Things, and supply chain management. The core of blockchain is the consensus mechanism, which determines how each node in the network reaches consensus on transactions and data. The consensus mechanism is not only related to the security of the blockchain, but also affects its performance and degree of decentralization. In early blockchain systems, Proof of Work (PoW) was the most common consensus mechanism. PoW competes to generate new blocks by calculating complex mathematical problems. Although this mechanism greatly enhances the security and anti-attack capabilities of the system, it also brings serious resource waste and high energy consumption problems. In addition, the high computational difficulty of PoW requires participants to have a large amount of computing power resources, leading to the emergence of mining pools, which in turn exacerbates the problem of centralized computing power and violates the original intention of blockchain decentralization. In order to overcome the shortcomings of PoW, Proof of Stake (PoS) came into being. PoS determines the node's block generation rights by the number and time of coins held, reducing the waste of computing resources. However, PoS also has some problems, such as the rich get richer effect, which makes it easier for nodes with more coins to get more block generation opportunities, further exacerbating the risk of centralization. In order to achieve a better balance between decentralization, security and efficiency, DPoS (Delegated Proof of Stake) was proposed as an improved version of PoS.

[0003] DPoS is a delegated proof of stake mechanism that selects a group of representatives through an election mechanism to verify transactions and generate blocks. In the DPoS system, all token holders can vote to elect block representatives, who are authorized by the holders to exercise power, thereby improving the efficiency and scalability of the system. Compared with PoW and PoS, DPoS has significant performance advantages, can support higher transaction throughput, and reduce network latency.

[0004] However, DPoS is not perfect. Since the weight of votes is directly related to the number of tokens held, large holders in the DPoS system may still influence the entire network by manipulating the election results. In addition, after obtaining votes, block producers may behave improperly or be inefficient if there is a lack of an effective supervision mechanism. Therefore, how to improve DPoS to solve these potential risks and problems has become the focus of researchers.

[0005] Traditional DPoS focuses on tokens and votes, while the data space consensus method focuses more on the distribution and storage of data, improving the overall performance and security of the system through more efficient data management and transmission mechanisms. This method can not only improve the access speed of data, but also further enhance the decentralization of the system by optimizing data storage strategies. Summary of the invention

[0006] In view of the above problems, the present invention aims to provide a data space consensus method based on improved DPoS to solve the problems raised in the above background technology.

[0007] The purpose of the present invention is achieved through the following technical solutions:

[0008] A data space consensus method based on improved DPoS, characterized by comprising the following steps:

[0009] Step 1: Divide the entire data space into R subspaces, each subspace contains several node sets;

[0010] Step 2: The I-th subspace includes N I Group nodes, each node has an initial trust value;

[0011] Step 3: Select the initial candidate nodes based on the initial trust value and random perturbation;

[0012] Step 4: Establish a multi-level consensus mechanism.

[0013] Furthermore, in step 1, the entire data space is divided into R subspaces. For the entire data space Z, there is: Z = {Z1, Z2, ... Z I ,…Z R}, where Z1 is the first subspace, Z2 is the second subspace, and Z I is the I-th subspace, Z R is the Rth subspace, each subspace is processed by a group of nodes, in, is the node group responsible for the first subspace, is the node group responsible for the second subspace, is the node group responsible for the I-th subspace, is the node group responsible for the Rth subspace.

[0014] Further, in step 1, each subspace includes a plurality of node sets, and the I-th subspace is defined to include N I Group node set, Nth I The group node set is in, For N I The first node of the group node set, For N I The second node of the group node set, For N I The Qth node in the group node set.

[0015] Further, in step 2, define the Nth I The i-th node of the group node set The workload is History records The contribution is For nodes The initial trust value is

[0016] Among them, w L , W H , W C They are the weights of workload, history record, and contribution respectively.

[0017] Furthermore, in step three, the initial candidate nodes are selected based on the initial trust value and the random disturbance, and the rules for selecting the initial candidate nodes are as follows:

[0018]

[0019] Among them, SORT() is a sorting function, sorting from high to low. N I The initial trust value, GAUSS (0, σ) is the Gaussian perturbation, σ is the standard deviation of the Gaussian perturbation, A k is the amplitude, f k is the frequency, φ k As the phase, through the SORT() sorting function, Z I The set of all nodes in the subspace is sorted from high to low.

[0020] Furthermore, in step 3, in order to avoid accidental shocks, the rule

[0021] V1, the second iteration is V2, and after the LOOP iteration is V LOOP,calculate in, is phase addition, which means adding the corresponding items before sorting. The node set corresponding to the maximum number of times the maximum value appears in the LOOP iterations is The node set corresponding to the maximum trust value after summing in LOOP iterations is Two sets of nodes compete to become initial candidate nodes.

[0022] Furthermore, in step 4, a multi-level consensus mechanism is established, including campaign rules, voting behavior, reward mechanism, block generation probability and supervision mechanism.

[0023] Furthermore, the election rules are as follows:

[0024] No. Z I subspaces, including N I Group nodes, the number of node groups that can participate in voting is N I -2;

[0025] The voting behavior is as follows:

[0026] Each node can vote and choose or As the initial candidate node, it can only cast one vote in each node set, and the one vote is selected according to the maximum value method, that is: Among them, ticket A represents voting Ticket B means voting

[0027] The reward mechanism is as follows:

[0028] For voting nodes, there are: Where η is the benchmark coefficient of reward allocation, λ is the reward allocation adjustment parameter, and r i is the reward for winning, rewardpool is the reward pool, is voting node n i voting weight;

[0029] The block probability is as follows:

[0030]

[0031] in, for The probability of being a candidate node to generate blocks, for The probability of being a candidate node to generate blocks, For not choosing As a reward for candidate nodes, For not choosing As a reward for candidate nodes, It is in Z I In the word space, all nodes are cast to The total number, as The adjustment parameters of the block probability, As The adjustment parameter of the block probability.

[0032] The supervision mechanism is as follows:

[0033] Candidate nodes are classified by probability and Elected super node N * , build a supervision mechanism to and To conduct supervision:

[0034]

[0035] in, for The supervision score of θ is the weight of the initial trust value. for The initial trust value of , ε is the block quality weight coefficient, for The amount of blocks produced, for The effective block output is μ, μ is the malicious behavior weight coefficient, For the detected The number of malicious behaviors, v is the weight coefficient of the number of blocks produced, for The number of blocks produced, for The supervision score, for The initial trust value of for The amount of blocks produced, for The effective block output, For the detected Number of malicious behaviors, for The number of blocks produced;

[0036] Super Node N * The supervision score cannot be lower than the lower boundary value L - , if it is lower than the lower boundary value L - , it is necessary to re-elect candidate nodes, that is

[0037] Allow candidate nodes The supervision score is temporarily lower than The supervision score of The supervision score is temporarily lower than supervision rating.

[0038] Furthermore, the term “temporary” is defined in two ways: supervision score definition: or As a candidate node, if The current super node will not be replaced; time period limit: after time period T1, reselect the candidate node and Participate in the calculation, where the node set corresponding to the maximum number of maximum values ​​in the LOOP iterations after the time period T1 is The node set corresponding to the maximum trust value after summing in LOOP iterations after time period T1 is And will re-accord with the probability and Elected super node N *

[0039] Beneficial effects of the present invention:

[0040] (1) The present invention is oriented to the data space, and uses blockchain technology to record the behavior operations of big data, divides the data space, and uses a super node as the main responsible node in each subspace to supervise and manage the actions of each node in the subspace. At the same time, each node in the subspace also supervises the super node. The record of each node will affect its own status and the status of the super node.

[0041] (2) The present invention selects initial candidate nodes based on initial trust values ​​and random disturbances. In the data space, it is inevitable that the node will be disturbed by data noise, and this disturbance cannot be ignored for the record of the data space. The present invention adds disturbance interference on the basis of selecting the initial candidate nodes based on the initial trust values, thereby enhancing the activity of the candidate nodes to be elected. In order to avoid accidental impacts of disturbances, the trust values ​​of the elected nodes are accumulated in an iterative manner, and the nodes corresponding to the maximum values ​​appearing the most times in the iteration and the nodes corresponding to the trust values ​​after summing up in the iteration are selected as candidate nodes to participate in the super node election, thereby expanding the election space.

[0042] (3) The present invention establishes a multi-level consensus mechanism, including election rules, voting behavior, reward mechanism, block generation probability and supervision mechanism. Under the action of the reward mechanism, the nodes that participate in the correct voting will be rewarded. Each node can vote to select the node corresponding to the maximum number of times the maximum value appears and the node corresponding to the maximum trust value after summing up in iterations as the candidate node, but can only cast one vote in each node set, avoiding the data blowout phenomenon during voting; the super node is affected by the supervision mechanism. After a cycle, the super node may continue to be elected or may be replaced. Under the influence of the supervision mechanism, the fairness, immutability and supervision of the entire data space are enhanced. In order to enhance the stability and continuity of node operation, the supervision mechanism includes supervision score limitation and time period limitation. Under the condition of not exceeding the threshold, the method tolerates the deviation of the supervision value under a certain threshold in exchange for the stability and continuity of the operation of the entire data space. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The invention is further illustrated by the accompanying drawings, but the embodiments in the accompanying drawings do not constitute any limitation to the invention. A person skilled in the art can obtain other drawings based on the following drawings without creative work.

[0044] Figure 1 It is a structural schematic diagram of the present invention. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] See also Figure 1 The present invention will be further described in conjunction with the following examples.

[0047] See also Figure 1 The present invention aims to provide a data space consensus method based on improved DPoS, characterized in that it includes the following steps:

[0048] Step 1: Divide the entire data space into R subspaces, each subspace contains several node sets;

[0049] Step 2: The I-th subspace includes N I Group nodes, each node has an initial trust value;

[0050] Step 3: Select the initial candidate nodes based on the initial trust value and random perturbation;

[0051] Step 4: Establish a multi-level consensus mechanism.

[0052] Specifically, in step 1, the entire data space is divided into R subspaces;

[0053] Specifically, for the entire data space Z, we have: Z = {Z1, Z2, … Z I ,…Z R}, where Z1 is the first subspace, Z2 is the second subspace, and Z I is the I-th subspace, Z R is the Rth subspace, each subspace is processed by a group of nodes, in, is the node group responsible for the first subspace, is the node group responsible for the second subspace, is the node group responsible for the I-th subspace, is the node group responsible for the Rth subspace.

[0054] Specifically, in step 1, each subspace includes a plurality of node sets, and the I-th subspace is defined to include N I Group node set, Nth I The group node set is in, For N I The first node of the group node set, For N I The second node of the group node set, For N I The Qth node in the group node set.

[0055] Specifically, in step 2, define the Nth I The i-th node of the group node set The workload is History records The contribution is For nodes The initial trust value is

[0056]

[0057] Among them, w L 、w H 、w C They are the weights of workload, history record, and contribution respectively.

[0058] Specifically, in step 3, the initial candidate node is selected based on the initial trust value and the random disturbance, and the rules for selecting the initial candidate node are as follows:

[0059]

[0060] Among them, SORT() is a sorting function, sorting from high to low. N I The initial trust value, GAUSS (0, σ) is the Gaussian perturbation, σ is the standard deviation of the Gaussian perturbation, A k is the amplitude, f k is the frequency, φ k As the phase, through the SORT() sorting function, Z I The set of all nodes in the subspace is sorted from high to low.

[0061] Specifically, in step 3, in order to avoid accidental shocks, remember the rule is V1, the second iteration is V2, and after the LOOP iteration is V LOOP ,calculate in, is phase addition, which means the corresponding items before sorting are added.

[0062] Specifically, the node set corresponding to the maximum number of maximum values ​​in the LOOP iterations is The node set corresponding to the maximum trust value after summing in LOOP iterations is Two sets of nodes compete to become initial candidate nodes.

[0063] Specifically, in step 4, a multi-level consensus mechanism is established, including campaign rules, voting behavior, reward mechanism, block generation probability and supervision mechanism.

[0064] Specifically, the election rules are as follows:

[0065] No. Z I subspaces, including N I Group nodes, the number of node groups that can participate in voting is N I -2;

[0066] Specifically, the voting behavior is as follows:

[0067] Each node can vote and choose or As the initial candidate node, it can only cast one vote in each node set, and the one vote is selected according to the maximum value method, that is: Among them, ticket A represents voting Ticket B means voting

[0068] Specifically, the reward mechanism is as follows:

[0069] For voting nodes, there are: Where η is the benchmark coefficient of reward allocation, λ is the reward allocation adjustment parameter, and r i is the reward for winning, rewardpool is the reward pool, is voting node n i voting weight;

[0070] Specifically, the block generation probability is as follows:

[0071]

[0072] in, for The probability of being a candidate node to generate blocks, for The probability of being a candidate node to generate blocks, For not choosing As a reward for candidate nodes, For not choosing As a reward for candidate nodes, It is in Z I In the word space, all nodes are cast to The total number, as The adjustment parameters of the block probability, As The adjustment parameter of the block probability.

[0073] Specifically, the supervision mechanism is as follows:

[0074] Candidate nodes are classified by probability and Elected super node N * , build a supervision mechanism to and To conduct supervision:

[0075]

[0076] in, for The supervision score of θ is the weight of the initial trust value. for The initial trust value of , ε is the block quality weight coefficient, for The amount of blocks produced, for The effective block output is μ, μ is the malicious behavior weight coefficient, For the detected The number of malicious behaviors, v is the weight coefficient of the number of blocks produced, for The number of blocks produced, for The supervision score, for The initial trust value of for The amount of blocks produced, for The effective block output, For the detected Number of malicious behaviors, for The number of blocks produced;

[0077] Specifically, super node N * The supervision score cannot be lower than the lower boundary value L - , if it is lower than the lower boundary value L - , it is necessary to re-elect candidate nodes, that is

[0078] Specifically, allow candidate nodes The supervision score is temporarily lower than The supervision score of The supervision score is temporarily lower than supervision rating.

[0079] Specifically, the term "temporarily" is defined from two aspects:

[0080] (1) Supervision rating limit: or As a candidate node, if The current super node will not be replaced;

[0081] (2) Time period limitation: After time period T1, reselect candidate nodes and Participate in the calculation, where the node set corresponding to the maximum number of maximum values ​​in the LOOP iterations after the time period T1 is The node set corresponding to the maximum trust value after summing in LOOP iterations after time period T1 is And will re-accord with the probability and Elected super node N * .

[0082] Beneficial effects of the present invention:

[0083] (1) The present invention is oriented to the data space, and uses blockchain technology to record the behavior operations of big data, divides the data space, and uses a super node as the main responsible node in each subspace to supervise and manage the actions of each node in the subspace. At the same time, each node in the subspace also supervises the super node. The record of each node will affect its own status and the status of the super node.

[0084] (2) The present invention selects initial candidate nodes based on initial trust values ​​and random disturbances. In the data space, it is inevitable that the node will be disturbed by data noise, and this disturbance cannot be ignored for the record of the data space. The present invention adds disturbance interference on the basis of selecting the initial candidate nodes based on the initial trust values, thereby enhancing the activity of the candidate nodes to be elected. In order to avoid accidental impacts of disturbances, the trust values ​​of the elected nodes are accumulated in an iterative manner, and the nodes corresponding to the maximum values ​​appearing the most times in the iteration and the nodes corresponding to the trust values ​​after summing up in the iteration are selected as candidate nodes to participate in the super node election, thereby expanding the election space.

[0085] (3) The present invention establishes a multi-level consensus mechanism, including election rules, voting behavior, reward mechanism, block generation probability and supervision mechanism. Under the action of the reward mechanism, the nodes that participate in the correct voting will be rewarded. Each node can vote to select the node corresponding to the maximum number of times the maximum value appears and the node corresponding to the maximum trust value after summing up in iterations as the candidate node, but can only cast one vote in each node set, avoiding the data blowout phenomenon during voting; the super node is affected by the supervision mechanism. After a cycle, the super node may continue to be elected or may be replaced. Under the influence of the supervision mechanism, the fairness, immutability and supervision of the entire data space are enhanced. In order to enhance the stability and continuity of node operation, the supervision mechanism includes supervision score limitation and time period limitation. Under the condition of not exceeding the threshold, the method tolerates the deviation of the supervision value under a certain threshold in exchange for the stability and continuity of the operation of the entire data space.

[0086] Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A data space consensus method based on improved DPoS, characterized in that: The following steps are involved: Step 1: Divide the entire data space into R subspaces, each subspace contains several node sets; Step 2: The I-th subspace includes N I Group nodes, each node has an initial trust value; Step 3: Select the initial candidate nodes based on the initial trust value and random perturbation; Step 4: Establish a multi-level consensus mechanism.

2. According to claim 1, a data space consensus method based on improved DPoS is characterized in that: In step 1, the entire data space is divided into R subspaces. For the entire data space Z, there is: Z = {Z1, Z2, ... Z I ,…Z R }, where Z1 is the first subspace, Z2 is the second subspace, and Z I is the I-th subspace, Z R is the Rth subspace, each subspace is processed by a group of nodes, in, is the node group responsible for the first subspace, is the node group responsible for the second subspace, is the node group responsible for the I-th subspace, is the node group responsible for the Rth subspace.

3. According to claim 1, a data space consensus method based on improved DPoS is characterized in that: In step 1, each subspace contains a number of node sets, and the I-th subspace is defined to include N I Group node set, Nth I The group node set is in, For N I The first node of the group node set, For N I The second node of the group node set, For N I The Qth node in the group node set.

4. According to claim 1, a data space consensus method based on improved DPoS is characterized in that: In step 2, define the Nth I The i-th node of the group node set The workload is History records The contribution is For nodes The initial trust value is Among them, w L 、w H 、w C They are the weights of workload, history record, and contribution respectively.

5. According to claim 1, a data space consensus method based on improved DPoS is characterized in that: In step 3, the initial candidate nodes are selected based on the initial trust value and the random disturbance. The rules for selecting the initial candidate nodes are as follows: Among them, SORT() is a sorting function, sorting from high to low. N I The initial trust value, GAUSS (0, σ) is the Gaussian perturbation, σ is the standard deviation of the Gaussian perturbation, A k is the amplitude, f k is the frequency, φ k As the phase, through the SORT() sorting function, Z I The set of all nodes in the subspace is sorted from high to low.

6. According to claim 5, a data space consensus method based on improved DPoS is characterized in that: In step 3, in order to avoid accidental shocks, remember the rule is V1, the second iteration is V2, and after the LOOP iteration is V LOOP ,calculate in, is phase addition, which means adding the corresponding items before sorting. The node set corresponding to the maximum number of times the maximum value appears in the LOOP iterations is The node set corresponding to the maximum trust value after summing in LOOP iterations is Two nodes are elected to become the initial candidate nodes.

7. According to claim 1, a data space consensus method based on improved DPoS is characterized in that: In step 4, a multi-level consensus mechanism is established, including campaign rules, voting behavior, reward mechanism, block probability and supervision mechanism.

8. According to claim 7, a data space consensus method based on improved DPoS is characterized in that: The election rules are as follows: No. Z I subspaces, including N I Group nodes, the number of node groups that can participate in voting is N I -2; The voting behavior is as follows: Each node can vote and choose or As the initial candidate node, it can only cast one vote in each node set, and the one vote is selected according to the maximum value method, that is: Among them, ticket A represents voting Ticket B means voting The reward mechanism is as follows: For voting nodes, there are: Where η is the benchmark coefficient of reward allocation, λ is the reward allocation adjustment parameter, and r i is the reward for winning, rewardpool is the reward pool, is voting node n i voting weight; The block probability is as follows: in, for The probability of being a candidate node to generate blocks, for The probability of being a candidate node to generate blocks, For not choosing As a reward for candidate nodes, For not choosing As a reward for candidate nodes, It is in Z I In the word space, all nodes are cast to The total number, as The adjustment parameters of the block probability, As Adjustment parameters of block probability; The supervision mechanism is as follows: Candidate nodes are classified by probability and Elected super node N * , build a supervision mechanism to and To conduct supervision: in, for The supervision score of θ is the weight of the initial trust value. for The initial trust value of , ε is the block quality weight coefficient, for The amount of blocks produced, for The effective block output is μ, μ is the malicious behavior weight coefficient, For the detected The number of malicious behaviors, v is the weight coefficient of the number of blocks produced, for The number of blocks produced, for The supervision score, for The initial trust value of for The amount of blocks produced, for The effective block output, For the detected Number of malicious behaviors, for The number of blocks produced; Allow candidate nodes The supervision score is temporarily lower than The supervision score of The supervision score is temporarily lower than supervision rating.

9. According to claim 8, a data space consensus method based on improved DPoS is characterized in that: The "temporary" is limited in two directions: supervision score limitation and time period limitation.

10. A data space consensus method based on improved DPoS according to claims 1-9, characterized in that: The methods provided in claims 1-9 are applied to blockchain systems to improve the efficiency and security of large-scale data processing.

Citation Information

Cited By

  • Data space consensus method based on improved DPoS

    CN120528753A