Data processing method for trusted spaceborne edge computing nodes based on DAG blockchain

Abstract

This application relates to a data processing method for trusted spaceborne edge computing nodes based on DAG blockchain. The method includes: dividing the spaceborne edge computing service area into multiple fixed geographical service regions; constructing multi-attribute evaluation indicators for each computing node based on node reputation, remaining service time, and load status; filtering among the computing nodes based on the multi-attribute evaluation indicators to obtain candidate nodes that meet the filtering criteria; updating the local DAG blockchain ledger according to the basic results; and updating the reputation inheritance proof generated by the computing node to the local DAG blockchain ledger corresponding to the new service region when the computing node crosses the boundary of the service region. By using a regionalized + parallel DAG ledger approach, the processing concurrency is improved, thereby increasing the node data processing efficiency.

Description

Technical Field

[0001] This application relates to the field of blockchain technology, and in particular to a data processing method, apparatus, computer equipment, storage medium, and computer program product for a trusted spaceborne edge computing node based on DAG blockchain. Background Technology

[0002] In recent years, the construction of space infrastructure centered on giant low-Earth orbit satellite constellations has been accelerated. Compared with the processing mode that relies on ground cloud computing centers, moving some computing, reasoning and decision-making capabilities to low-Earth orbit satellites for on-board edge computing can effectively achieve rapid on-site perception and analysis.

[0003] In traditional solutions, computing tasks are deployed on satellite edge nodes to achieve distributed computing and on-chain result confirmation. However, there are still problems such as low data processing efficiency of computing nodes and reputation breakdown when nodes cross regions. Summary of the Invention

[0004] Based on this, it is necessary to provide a data processing method, apparatus, computer equipment, computer-readable storage medium, and computer program product for a trusted spaceborne edge computing node of a DAG blockchain that can improve data processing efficiency and ensure the smooth inheritance of node reputation, in order to address the above-mentioned technical problems.

[0005] Firstly, this application provides a data processing method for a trusted onboard edge computing node in a DAG blockchain. The method includes: The spaceborne edge computing service space is divided into multiple fixed geographical service areas, wherein each service area includes multiple computing nodes and multiple task publishing nodes; the task publishing nodes are used to publish computing tasks to the computing nodes, and each computing node is used to maintain a local DAG blockchain ledger corresponding to the service area, and the DAG blockchain ledger is used to record the processing result information of the computing tasks; For each computing node, the node reputation, remaining service time, and load status of the computing node are obtained, and a multi-attribute evaluation index for each computing node is constructed based on the node reputation, remaining service time, and load status. Based on the multi-attribute evaluation index, candidate nodes that meet the screening conditions matching the redundancy requirements are obtained from each of the computing nodes; wherein, the candidate nodes are used to process the basic tasks published by the task publishing node and obtain the basic results of the candidate nodes for the basic tasks. The local DAG blockchain ledger is updated based on the aforementioned basic results; When the computing node crosses the boundary of a service area, the reputation inheritance proof generated by the computing node is updated in the local DAG blockchain ledger corresponding to the new service area.

[0006] Secondly, this application also provides a data processing device for a trusted spaceborne edge computing node of a DAG blockchain. The device includes: The partitioning module is used to divide the spaceborne edge computing service into multiple fixed geographical service areas, wherein each service area includes multiple computing nodes and multiple task publishing nodes; the task publishing nodes are used to publish computing tasks to the computing nodes, and each computing node is used to maintain a local DAG blockchain ledger corresponding to the service area, and the DAG blockchain ledger is used to record the processing result information of the computing tasks; The information acquisition module is used to acquire the node reputation, remaining service time and load status of each computing node, and to construct a multi-attribute evaluation index for each computing node based on the node reputation, remaining service time and load status. The filtering module is used to filter among the computing nodes based on the multi-attribute evaluation index to obtain candidate nodes that meet the filtering conditions matching the redundancy requirements; wherein, the candidate nodes are used to process the basic tasks published by the task publishing node to obtain the basic results of the candidate nodes for the basic tasks. The ledger update module is used to update the local DAG blockchain ledger based on the aforementioned basic results; The inheritance module is used to update the reputation inheritance proof generated by the computing node to the local DAG blockchain ledger corresponding to the new service area when the computing node crosses the boundary of the service area.

[0007] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the above method steps.

[0008] Fourthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the above method steps.

[0009] Fifthly, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, implements the above method steps.

[0010] The aforementioned DAG blockchain's trusted spaceborne edge computing node data processing method, apparatus, computer equipment, storage medium, and computer program products The spaceborne edge computing service space is divided into multiple fixed geographical service areas. Each service area includes multiple computing nodes and multiple task publishing nodes. The task publishing nodes are used to publish computing tasks to the computing nodes. Each computing node is used to maintain the local DAG blockchain ledger corresponding to the service area. The DAG blockchain ledger is used to record the processing results of the computing tasks. Dividing the service space into multiple fixed areas and combining the service areas with the DAG ledger reduces the network-wide synchronization overhead. By using the regionalized + parallel DAG ledger approach, the processing concurrency is improved, thereby improving the data processing efficiency of the nodes.

[0011] For each computing node, the node reputation, remaining service time, and load status are obtained. Based on the node reputation, remaining service time, and load status, multi-attribute evaluation indicators are constructed for each computing node. This provides an objective and multi-dimensional decision-making basis for subsequent accurate node selection, thereby improving the overall efficiency of node data processing.

[0012] Based on multi-attribute evaluation metrics, candidate nodes that meet the redundancy requirements are selected from each computing node. These candidate nodes are used to process the basic tasks published by the task publishing nodes, and the basic results of the candidate nodes for the basic tasks are obtained. The quality of nodes participating in the core computing is ensured through precise selection, and the reliability and fault tolerance of the results are enhanced by using multiple nodes to process the basic tasks in parallel.

[0013] The local DAG blockchain ledger is updated based on the basic results, which realizes the immutable recording and distributed synchronization of task processing results, ensuring the traceability of the computing process and data consistency within the region.

[0014] When a computing node crosses the boundary of a service region, the reputation inheritance proof generated by the computing node is updated in the local DAG blockchain ledger corresponding to the new service region. This method, on the one hand, achieves seamless connection between node identity and trust level through cross-region migration of reputation proofs, ensuring the continuity and reliability of services provided by onboard nodes when dynamically switching regions. On the other hand, by using a regionalized + parallel DAG ledger approach, it improves processing concurrency, thereby increasing node data processing efficiency. Attached Figure Description

[0015] Figure 1 This is an application environment diagram of a data processing method for a trusted spaceborne edge computing node of a DAG blockchain in one embodiment. Figure 2This is a flowchart illustrating a data processing method for a trusted onboard edge computing node of a DAG blockchain in one embodiment. Figure 3 This is a schematic diagram of a local DAG blockchain service scenario in one embodiment; Figure 4 This is a flowchart illustrating a method for determining node reputation in one embodiment; Figure 5 This is a flowchart illustrating a method for determining candidate nodes in one embodiment; Figure 6 This is a flowchart illustrating a computing node verification method in one embodiment; Figure 7 This is a flowchart illustrating the cross-regional node reputation inheritance process in one embodiment; Figure 8 This is a schematic diagram of simulation results for different ablation mechanisms in one embodiment; Figure 9 This is a schematic diagram of simulation results under different proportions of malicious nodes in one embodiment; Figure 10 This is a schematic diagram of simulation results under different numbers of geographic service areas in one embodiment; Figure 11 This is a schematic diagram of simulation results under different task arrival rates in one embodiment; Figure 12 This is a structural block diagram of a data processing device for a trusted onboard edge computing node of a DAG blockchain in one embodiment. Figure 13 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0017] The data processing method for trusted satellite-borne edge computing nodes in DAG blockchain provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, the task issuing node 101 can be a ground-based node used to send computing tasks to the satellite, and the computing node 102 can be a node in the satellite used to perform computing. The task issuing node 101 and the computing node 102 can transmit signals to each other through a satellite-to-ground communication link.

[0018] The task publishing node 101 can divide the spaceborne edge computing service into multiple fixed geographical service areas. Each service area includes multiple computing nodes 102 and multiple task publishing nodes 101. The task publishing node 101 is used to publish computing tasks to the computing nodes 102. Each computing node 102 is used to maintain the local DAG blockchain ledger corresponding to the service area. The DAG blockchain ledger is used to record the processing result information of the computing tasks. For each computing node 102, the task publishing node 101 can obtain the node reputation, remaining service time and load status of the computing node 102, and construct multi-attribute evaluation indicators for each computing node 102 based on the node reputation, remaining service time and load status. The task publishing node 101 can filter among the computing nodes 102 based on multi-attribute evaluation indicators to obtain candidate nodes that match the redundancy requirements; among them, the candidate nodes are used to process the basic tasks published by the task publishing node 101 and obtain the basic results of the candidate nodes for the basic tasks.

[0019] Computation node 102 can update the local DAG blockchain ledger based on the base results.

[0020] When compute node 102 crosses the boundary of a service area, compute node 102 can update the reputation inheritance proof generated by compute node 102 to the local DAG blockchain ledger corresponding to the new service area.

[0021] In one embodiment, such as Figure 2 As shown, a data processing method for a trusted onboard edge computing node in a DAG blockchain is provided, which can be applied to... Figure 1 Taking the nodes in the example, the explanation includes: S201 divides the space for spaceborne edge computing services into multiple fixed geographical service areas.

[0022] Each service area includes multiple computing nodes and multiple task publishing nodes. The task publishing nodes are used to publish computing tasks to the computing nodes. Each computing node is used to maintain the local DAG blockchain ledger corresponding to the service area. The DAG blockchain ledger is used to record the processing result information of the computing tasks.

[0023] Among them, the spaceborne edge computing service space is the overall logical and physical range covered by computing resources carried by satellites (or other space-based platforms).

[0024] A service area can be a fixed subspace obtained by geographically dividing the service space. A service area can contain multiple computing nodes and multiple task publishing nodes.

[0025] A computing node is an entity located within the service area that possesses computing capabilities and is responsible for maintaining the blockchain ledger. Specifically, a computing node can be a processor with computing capabilities on a satellite.

[0026] A task publishing node is an entity located within the service area that needs computing services and publishes computing tasks to the computing node.

[0027] A computing task can be a specific work instruction or data processing request issued by a task issuing node and executed by a computing node.

[0028] A local DAG blockchain ledger is a distributed ledger organized using a DAG (Directed Acyclic Graph) structure instead of a traditional single-chain structure for the current service area. DAG blockchain ledgers allow multiple transactions to be written concurrently, solving the performance problem of traditional single-chain blockchains and providing a basis for improving node data processing efficiency in the future.

[0029] Processing result information includes the final output data generated by the computing node after completing the computing task, which needs to be recorded on the DAG ledger, along with its verification information.

[0030] Specifically, the space for spaceborne edge computing services is divided into multiple fixed geographical service areas. Each service area It includes multiple computing nodes. And multiple task publisher nodes (TPs), the task publisher nodes (TPs) are used to distribute tasks to compute nodes. Deploy computing tasks to each computing node Used to maintain the local DAG blockchain ledger corresponding to the service area. ,in Represents the set of transactions in the ledger. This represents the hash reference relationship between a transaction and its parent transaction. The DAG blockchain ledger is used to record the processing results of computational tasks.

[0031] Specifically, consider by A giant low-Earth orbit (LEO) satellite constellation consisting of LEO satellite nodes, the set of satellites is denoted as Each satellite is equipped with an onboard edge computing unit and Each ISL laser communication terminal possesses the foundation for deploying lightweight blockchain nodes, such as cryptographic computation and node authentication. System time is discretized into... Each of equal length The time slots. Computing nodes can divide the spaceborne edge computing service space into... A fixed geographic service area The computing node can obtain its current latitude and longitude location and determine its current geographic service area based on the latitude and longitude. Time slot, within the service area Computing nodes in the air provide computing services for it. Service area It includes multiple task publishing nodes (TPs) and is organized into a local decentralized P2P coverage network through inter-satellite links, jointly maintaining a local DAG ledger. ,in Represents the set of transactions in the ledger. This indicates the hash reference relationship between a transaction and its parent transaction.

[0032] It should be noted that within the service area Within the system, multiple task nodes require access to the LEO satellite constellation internet to obtain highly reliable computing services. TPs can be individuals, enterprises, or even satellites. TPs do not possess high-intensity computing power; they access the internet through... P2P network initiation has redundancy requirements The computational tasks published by TP are used to obtain task computation services. ,in For input data, The number of CPU cycles required per bit of the task. This is for redundant consensus.

[0033] Understandable, in ground service areas Within the local ledger, all tasks and computation results (processing result information) of all computing nodes are published to the local ledger via transactions. Up. Whenever a compute node completes a redundant task... A transaction will be published. A transaction is represented by a quadruple: .

[0034] in , Indicates transaction Unique hash value This indicates the task ID to which the exchange belongs. This indicates that the computing node that generated the transaction has a globally unique ID within the constellation. express Transaction type; They represent The unique hash value that points to the predecessor transaction in the task chain and node chain. Then it means The transaction payload is divided into two types: ordinary computation transactions and area handover maintenance transactions. In ordinary computation transactions, the payload is the encrypted result of redundant task calculations; while in area handover maintenance transactions, the payload is a reputation inheritance certificate for the satellite node that is about to undergo area handover.

[0035] like Figure 3 As shown, for a DAG blockchain ledger, the DAG blockchain ledger contains task chains and node chains and their interconnections. The task chain is a link between different transactions of the same task, and the task chain is used to reach consensus on the task results (each task has the same corresponding calculation result). The node chain is a link between different task transactions completed in the same computing node.

[0036] Normally, later-published transactions link to previously published transactions, forming a task chain with other transactions. However, due to the redundant computation of multiple satellite nodes performing tasks in parallel in P2P networks, transactions are published on the chain concurrently, resulting in three special cases. Firstly, if satellites... In satellite The transaction was released later, but both were related to the task. If redundant calculations are completed too close together, it will cause satellite... In publishing the transaction No satellite has been received yet Published transactions Therefore, the transaction and There will be no link between them. Secondly, due to network latency and jitter, if the satellite... In publishing the transaction At that time, the data from the satellite had not yet been synchronized. Published transactions , No link Thirdly, satellites. Publish the transaction after completing the calculation. At that time, when a published transaction is selected Verification is required before making a link. Check if the calculation result in the previous step is consistent with the current calculation result. If they are inconsistent, then... No link Above; if any published transactions fail to pass verification, Link to Genesis Transaction .

[0037] In the first two scenarios, if two transactions yield the same result, they are indirectly linked, and all transactions with identical results eventually form a DAG subgraph task chain. In the third scenario, transactions with different results cannot be linked, thus forming a fork. The number of interconnected transactions with the same result within a task chain is the weight of that chain. For the same task, the chain containing the most transactions is called the main chain, and the others are called parasitic chains. (See reference...) Figure 3 As shown, the main chain of Task 1 consists of the processing results of four task nodes (excluding the Genesis Task transaction). The parasitic chain of Task 1 consists of the processing result of one task node.

[0038] Furthermore, when the task Main chain weight When, explain the task The results are from a computation that has achieved consensus among a majority of nodes in the current trusted mission collaborative computing satellite ensemble. It should be noted that each computing node, upon completing its mission... After the calculation, the result of the computation task is obtained. The computing node publishes (writes) the result to the DAG blockchain ledger in the form of a transaction, and then reaches a consensus based on the shape of the task chain in the ledger.

[0039] Dividing the service space into multiple fixed regions and combining the service regions with the DAG ledger reduces the network-wide synchronization overhead. By using a region-based approach combined with a parallel DAG ledger, the processing concurrency is improved, thereby increasing the efficiency of node data processing.

[0040] S202. For each computing node, obtain the node reputation, remaining service time, and load status of the computing node, and construct multi-attribute evaluation indicators for each computing node based on the node reputation, remaining service time, and load status.

[0041] Among them, node reputation can be a comprehensive score that quantitatively assesses the trustworthiness of a node based on its past historical behavior.

[0042] The remaining service time of a node can be the remaining time that a computing node is expected to maintain normal operation (or remain in the current service area).

[0043] Node load status can be an indicator of the resource utilization or busyness of a computing node at a given moment.

[0044] Multi-attribute evaluation metrics can be a system of metrics or a comprehensive scoring model that combines multiple single-dimensional metrics (reputation, remaining time, load status) to comprehensively evaluate the overall capabilities of a node.

[0045] Specifically, for each computing node, the task publishing node obtains the node reputation, remaining service time, and load status of the computing node, and constructs a multi-attribute evaluation index for each computing node based on the node reputation, remaining service time, and load status.

[0046] Specifically, the task issuing node obtains the node reputation of the computing nodes, which may include: the task issuing node obtaining the task index and node index corresponding to each computing node; before task scheduling, the task issuing node locates the latest transaction of each computing node based on the node index; where the latest transaction represents the task processing result closest to the current time; when the computing node is in a basic task scenario, it backtracks along the directed edges of the node chain based on the node chain of the latest transaction and the task index to obtain the number of main chains and parasitic chains of the node chain to which the computing node belongs; based on the number of main chains and parasitic chains, the node reputation of the computing node is determined, where the number of main chains is positively correlated with the node reputation, and the number of parasitic chains is negatively correlated with the node reputation; when the computing node is in a reputation inheritance proof task scenario, it obtains the historical reputation base of the computing node by backtracking to the most recent reputation inheritance declaration transaction; based on the historical reputation base and the verification result of the historical reputation base, the node reputation of the computing node is determined.

[0047] Specifically, the task issuing node obtains the node reputation of the computing node, which includes: the task issuing node selecting a nearby satellite. Access P2P network to synchronize DAG ledger and according to calculate Node reputation of each satellite .

[0048] Based on node reputation, remaining service time, and load status, we construct multi-attribute evaluation indicators for each computing node to provide objective and multi-dimensional decision-making basis for subsequent accurate node selection, thereby improving the overall efficiency of node data processing.

[0049] S203, based on multi-attribute evaluation indicators, filters among the computing nodes to obtain candidate nodes that meet the screening conditions that match the redundancy requirements.

[0050] Among them, candidate nodes are used to process the basic tasks published by the task publishing nodes and obtain the basic results of the candidate nodes for the basic tasks.

[0051] Candidate nodes can be computing nodes that meet the filtering criteria, such as nodes with high reputation, long remaining service time, and low load. The filtering criteria are not limited here and can be flexibly selected based on the actual situation.

[0052] Redundancy requirement, also known as redundancy consensus, can be equal to the number of candidate nodes selected. For example, if the redundancy requirement is 1, the corresponding selection criterion for redundancy requirement will select 1 candidate node.

[0053] A basic task can be a raw computational work unit issued by a task-issuing node that needs to be processed independently and in parallel by candidate nodes. Specifically, a basic task can be a task that a candidate node can receive and process normally within its current service area. Furthermore, if a computing node is about to leave its current service area, it can act as a task issuer, selecting candidate nodes and issuing tasks involving cross-regional reputation authentication to them. That is, the computing node uses the computation of cross-regional reputation authentication tasks by candidate nodes to prove its trustworthiness to the next service area, thereby achieving cross-regional reputation inheritance.

[0054] The basic results are the raw output data obtained after each candidate node has independently completed the basic tasks.

[0055] Specifically, each computing node can be screened according to its reputation, remaining service time, and load status. The intersection of the three screening results is then used to obtain candidate nodes that meet the screening criteria that match the redundancy requirements.

[0056] Specifically, entropy weighting can be applied to each of the multi-attribute evaluation indicators to obtain the objective weights of each computing node for different attribute evaluation indicators. Based on the decision matrix constructed from these objective weights, the globally optimal positive ideal solution and the worst negative ideal solution are determined. For each computing node, the distances to the globally optimal positive ideal solution and the worst negative ideal solution are calculated according to the node's position in the decision matrix, determining the node's proximity. A higher proximity indicates that the computing node is closer to the optimal state under the comprehensive consideration of honesty, reliability, stable residency, and computational load. The computing nodes are then sorted in descending order according to their proximity to obtain candidate nodes that meet the screening criteria.

[0057] Precise selection ensures the quality of nodes participating in core computing, and the use of multiple nodes to process basic tasks in parallel enhances the reliability and fault tolerance of results.

[0058] S204, update the local DAG blockchain ledger based on the various basic results.

[0059] The underlying results may or may not be consistent. By writing these underlying results into the DAG blockchain ledger, consensus is reached on each result. If consensus is reached, a transaction is generated based on the confirmed underlying results, and this transaction is linked to or points to existing transactions in the current local DAG ledger, forming a new directed acyclic graph structure, thus completing the ledger update. It should be noted that each computing node, upon completing its task... After the calculation, the calculated task results are published (written) into the DAG blockchain ledger in the form of transactions, and then consensus is reached based on the shape of the task chain in the ledger.

[0060] The local DAG blockchain ledger is updated based on the basic results, which realizes the immutable recording and distributed synchronization of task processing results, ensuring the traceability of the computing process and data consistency within the region.

[0061] S205, when a computing node crosses the boundary of a service area, the reputation inheritance proof generated by the computing node is updated to the local DAG blockchain ledger corresponding to the new service area.

[0062] Among them, the reputation inheritance proof can be a cryptographic credential generated by a computing node that crosses the service area boundary to prove its reputation and trusted history accumulated in the original service area.

[0063] A new service area can be a different fixed geographic service area that a computing node enters after crossing a boundary.

[0064] Specifically, the proof of credit inheritance can be a committee-aggregated proof. When a computing node crosses the boundary of a service area, the computing node first issues a temporary computing task. Multiple committee nodes are then selected from among the computing nodes to perform the temporary computing task. Once the committee nodes reach a consensus, the computing node (the node to be migrated) generates a committee aggregation proof using BLS aggregation signature. The node to be migrated then writes the committee aggregation proof into the local DAG blockchain ledger corresponding to the new service area. In this way, the computing node's reputation is inherited across regions.

[0065] By enabling cross-regional migration of reputation proof, a seamless connection between node identity and trust level is achieved, ensuring the continuity and reliability of services provided by onboard nodes when dynamically switching regions.

[0066] In the data processing method of the trusted spaceborne edge computing node of the aforementioned DAG blockchain, The spaceborne edge computing service space is divided into multiple fixed geographical service areas. Each service area includes multiple computing nodes and multiple task publishing nodes. The task publishing nodes are used to publish computing tasks to the computing nodes. Each computing node is used to maintain the local DAG blockchain ledger corresponding to the service area. The DAG blockchain ledger is used to record the processing results of the computing tasks. Dividing the service space into multiple fixed areas and combining the service areas with the DAG ledger reduces the network-wide synchronization overhead. By using the regionalized + parallel DAG ledger approach, the processing concurrency is improved, thereby improving the data processing efficiency of the nodes.

[0067] For each computing node, the node reputation, remaining service time, and load status are obtained. Based on the node reputation, remaining service time, and load status, multi-attribute evaluation indicators are constructed for each computing node. This provides an objective and multi-dimensional decision-making basis for subsequent accurate node selection, thereby improving the overall efficiency of node data processing.

[0068] Based on multi-attribute evaluation metrics, candidate nodes that meet the redundancy requirements are selected from each computing node. These candidate nodes are used to process the basic tasks published by the task publishing nodes, and the basic results of the candidate nodes for the basic tasks are obtained. The quality of nodes participating in the core computing is ensured through precise selection, and the reliability and fault tolerance of the results are enhanced by using multiple nodes to process the basic tasks in parallel.

[0069] The local DAG blockchain ledger is updated based on the basic results, which realizes the immutable recording and distributed synchronization of task processing results, ensuring the traceability of the computing process and data consistency within the region.

[0070] When a computing node crosses the boundary of a service region, the reputation inheritance proof generated by the computing node is updated in the local DAG blockchain ledger corresponding to the new service region. This method, on the one hand, achieves seamless connection between node identity and trust level through cross-region migration of reputation proofs, ensuring the continuity and reliability of services provided by onboard nodes when dynamically switching regions. On the other hand, by using a regionalized + parallel DAG ledger approach, it improves processing concurrency, thereby increasing node data processing efficiency.

[0071] In one embodiment, such as Figure 4 The flowchart shown above illustrates the method for determining node reputation. Specifically, obtaining the node reputation of a computing node in step S202 includes: S401, obtain the task index and node index corresponding to each computing node.

[0072] The task index, in particular, can be an index within the DAG blockchain ledger used to locate transactions in the task chain. It can record the set of end transactions of the task chain in the current DAG blockchain ledger. The task chain contains the task processing results of multiple computing nodes.

[0073] A node index is an index used to locate transactions within a node chain in a DAG blockchain ledger. Specifically, a node index... It can record the set of end transactions of the node chain in the current DAG blockchain ledger, and the node chain contains multiple computing nodes.

[0074] In the memory of the compute node, the time complexity of retrieving the task index and node index is O(n log n). .

[0075] S402, the task publishing node locates the latest transaction of each computing node based on the node index before task scheduling.

[0076] The latest transaction represents the task processing result most recently at the current moment.

[0077] Specifically, before scheduling tasks, the task publishing node needs to base its work on the DAG ledger. Assess the reputation of each computing node. Task publishing nodes are identified via node indexes. Positioning computing node (satellite) Latest transactions .

[0078] S403, when the computing node is in a basic task scenario, backtrack along the directed edges of the node chain according to the node chain of the latest transaction and the task index to obtain the number of main chains and parasitic chains of the node chain to which the computing node belongs; based on the number of main chains and parasitic chains, determine the node reputation of the computing node.

[0079] Among them, the number of main chains is positively correlated with node reputation, while the number of parasitic chains is negatively correlated with node reputation.

[0080] In this context, the basic task scenario can be defined as the computing node being located within a region but not within a soft boundary. Conversely, if the computing node is located within a soft boundary, it can be considered to be in a reputation proof task scenario.

[0081] The number of main chains can represent the number of main chains to which the transaction results contained in the node chain corresponding to a computing node belong. For example, if the node chain corresponding to a computing node contains 4 transaction results, and one of these transaction results belongs to the main chain, then the number of main chains is 1.

[0082] The number of parasitic chains can represent the number of parasitic chains to which the transaction results contained in the node chain corresponding to the computing node belong.

[0083] For example, refer to Figure 3 As shown, for the topmost computing node, there are 4 redundant computing task transaction results in the node chain to which the computing node belongs. Among them, transaction result 1 belongs to the main chain, and transaction results 2, 3 and 4 belong to the parasitic chain.

[0084] Specifically, along the computing nodes (satellites) ) node chain The directed edges are backtracked in reverse. During the backtracking process, for each ordinary task transaction encountered, it is checked whether it is on the main chain of the task chain. If the transaction is on the main chain, the reputation of that node is increased. If it is on the parasitic chain, then deduct. This determines the node's reputation. It should be noted that a regular task transaction check indicates a task transaction in the node chain that belongs to the reputation inheritance proof category. This is not included in the reputation assessment; rather, it reflects the potential obligation of high-reputation nodes to maintain system operation. Reputation is only assessed for transactions related to ordinary consensus computation tasks. Furthermore, during node chain traversal, if the traversal proceeds backwards to the most recent reputation inheritance declaration transaction... This will interrupt the retrieval process and prevent further reverse backtracking, i.e., for transactions... Previous deep historical redundant calculation records have been isolated and will no longer be included in this round of reputation assessment. Also, during the task chain traversal, if a satellite is retrieved... Genesis Transaction once issued a cross-regional credit inheritance certificate task. If the satellite is in a soft boundary state, then reputation retrieval is interrupted and the satellite is removed from the available scheduling pool.

[0085] S404, In the case of a computing node being in a reputation inheritance proof task scenario, obtain the historical reputation base of the computing node by tracing back to the most recent reputation inheritance declaration transaction; and determine the node reputation of the computing node based on the historical reputation base and the verification result of the historical reputation base.

[0086] In the context of reputation inheritance proof task, the computing node is about to approach the region boundary. At this time, the computing node can be considered to be at the soft boundary, and the computing node is in the reputation proof task scenario.

[0087] Historical reputation baseline represents the sum of scores of a computing node in other regions preceding the current region. The higher the historical reputation baseline, the higher the probability that the computing node has reached consensus on historical tasks, and the more trustworthy the node is.

[0088] The most recent reputation inheritance declaration transaction represents the processing result of the reputation inheritance declaration task of this compute node in the previous service area. The purpose of this processing result is to isolate the deep historical redundant calculation records on the node chain corresponding to the compute node, thereby reducing the amount of data processing while ensuring the accurate calculation of the node's reputation.

[0089] The historical reputation baseline verification result represents the consensus confirmation result for the most recent reputation inheritance declaration transaction of the computing node. If the verification result is greater than the preset redundancy consensus level, the verification result is 1; otherwise, the verification result is 0.

[0090] Among them, node reputation The calculation formula is as follows: ; ; Specifically, this includes: for computing nodes (satellites) When searching for a credit inheritance statement transaction At that time, the historical reputation base of the computing node is obtained by tracing back to the most recent reputation inheritance statement transaction. Verify the historical credit base; if the verification passes ( ),but The value is 1 if the verification fails. ),but =0. Along the computing node (satellite) ) node chain The directed edges are backtracked in reverse. During the backtracking process, for each ordinary task transaction encountered, it is checked whether it is on the main chain of the task chain. If the transaction is on the main chain, the reputation of that node is increased. If it is on the parasitic chain, then deduct. .

[0091] In this embodiment, firstly, reputation is quantified through the structural features of the DAG graph, improving the accuracy of the assessment. Secondly, instead of scanning the entire history of a node, the latest transaction is used as a starting point to backtrack along the directed edges of the node chain, significantly reducing computational overhead and making it suitable for the limited computing resources of onboard nodes.

[0092] In one embodiment, refer to Figure 5 As shown, S203 above filters each computing node based on multi-attribute evaluation metrics to obtain candidate nodes that meet the redundancy requirements, specifically including: S501 performs soft boundary screening on each computing node, selecting intermediate nodes that meet the soft boundary screening conditions.

[0093] The soft boundary filtering condition can be a filtering condition configured for compute nodes. Specifically, the soft boundary filtering condition can be to remove compute nodes that are about to leave the current service area and retain other nodes (i.e., intermediate nodes).

[0094] Intermediate nodes represent computing nodes that are located in the current service area and will remain in the current service area for a period of time in the future.

[0095] Specifically, soft boundary screening conditions can be characterized by a dynamic safety threshold. The expression is as follows: ; in, It can represent a computing node Execution time (computation node) within the current region (Time required to complete the calculation and obtain the result after all tasks in the task queue are executed). This can represent a queue of redundant computational tasks to be executed. It can represent the safety margin coefficient. Take a value greater than or equal to 1.2. It can represent the average time spent.

[0096] Among them, average time It can be done in each time slot Maintain a physical time length The sliding time window is used to calculate the average duration of tasks that successfully reach consensus within the sliding time window. This average duration is the average time spent.

[0097] Remaining service time The remaining service duration of the service area can be calculated using the track information of the computing nodes. and time slots Calculations show that When the remaining service time Less than the dynamic security threshold At this time, the satellite enters the soft boundary buffer zone, refuses to accept new tasks, only processes historical tasks that have entered the queue, and initiates a service area handover procedure of connecting first and then disconnecting.

[0098] For example, the service area is first determined based on a soft boundary mechanism. Active node subset within This involves filtering at each computing node. After removing nodes that are in the soft boundary buffer, the result is a summary of the task. The set of intermediate nodes that satisfy the soft boundary screening criteria The expression is as follows: ; Assumption set Includes One usable satellite .

[0099] S502, entropy weight processing is performed on the multi-attribute evaluation indicators corresponding to each intermediate node to obtain the objective weight of each intermediate node for different attribute evaluation indicators.

[0100] Among these, the multiple attributes include cost-based attributes and benefit-based attributes, with benefit-based attributes including node reputation. and remaining service time Cost-related attributes include the overall load of nodes. Objective weights are the weights of multi-attribute evaluation indicators after entropy weighting.

[0101] Specifically, a multi-attribute initial decision matrix is ​​constructed for the multi-attribute evaluation indicators of V satellites. ,in Indicates the index of the intermediate node. Indicates the attribute index. This represents the node's reputation level. This represents the remaining service time of a node in the current service area, and is a benefit-oriented indicator. This represents the overall load status of a node and is a cost-based indicator. Since the three indicators mentioned above differ in their physical dimensions and numerical tendencies, an extremum method is used to optimize the matrix. Normalization is performed to map all indicators to the [0,1] interval, resulting in a standardized decision matrix. For benefit-type indicators ( ), using forward normalization: ; For cost-type indicators ( Inverse normalization is used to unify the optimization direction: ; The entropy weight method is used to dynamically allocate weights based on the dispersion of each indicator data. First, the weights are calculated... The node at the th The proportion of features under each indicator : ; Then calculate the first Information entropy of the indicator (like Then define ): ; Further conclusions Objective information weight of each indicator And satisfy : ; S503, based on the decision matrix constructed from each objective weight, determines the globally optimal positive ideal solution and the worst negative ideal solution.

[0102] Specifically, a weighted normalized decision matrix is ​​constructed by combining objective weights. The individual element is Based on this, the globally optimal positive ideal solution under the current environment is determined. With the worst-case negative ideal solution : ; ; S504. For each intermediate node, the distance between it and the global optimal positive ideal solution and the worst negative ideal solution is calculated based on the position of the intermediate node in the decision matrix, and the proximity of the intermediate node is determined.

[0103] Among them, the greater the proximity, the closer the intermediate node is to the optimal state under the comprehensive consideration of honesty, reliability, stable residence, and computational load.

[0104] The position of an intermediate node in the decision matrix can represent the position of the intermediate node in the normalized decision matrix. The position in the middle.

[0105] The position of the intermediate node in the decision matrix, and Calculate the Euclidean distance for each intermediate node. Distance to the ideal solution Distance to the negative ideal solution : ; Among them, the proximity of the intermediate nodes The expression is as follows: ; S505: Sort the intermediate nodes in descending order according to their proximity to obtain candidate nodes that meet the screening criteria that match the redundancy requirements.

[0106] Among them, for candidate set All intermediate nodes in Sort the nodes in descending order to obtain the top-ranked candidate nodes. For example, the node with the highest score can be selected. Composed of satellites Collective deployment of redundant computing And the calculation results are confirmed by consensus.

[0107] In this embodiment, firstly, nodes about to leave are eliminated through soft boundary screening conditions, avoiding task termination, reducing unnecessary computational overhead, and improving overall computational efficiency. Secondly, the entropy weight method is used to determine objective weights, avoiding subjective weighting bias. The entropy weight method dynamically calculates weights based on the dispersion of actual data for each node, eliminating human intervention and subjective bias, and improving the accuracy of screening. Thirdly, the fit output is a continuous value between 0 and 1, supporting descending order sorting, and can be directly used for prioritizing candidate nodes (e.g., before selection). (One as a candidate node).

[0108] In one embodiment, refer to Figure 6 As shown, in the case where a computing node crosses the boundary of a service area, S205 updates the reputation inheritance proof generated by the computing node to the local DAG blockchain ledger corresponding to the new service area. Specifically, this includes: S601: When a computing node crosses the boundary of a service area, the computing node is used as a temporary task publishing node to publish a reputation inheritance proof task in the local DAG blockchain ledger of the original service area.

[0109] Specifically, when computing nodes Within the original service area When a soft boundary is encountered, it will be treated as a special temporary TP direction. Issue a credit inheritance certificate task .

[0110] S602, based on the proximity of each computing node, selects a set of baseline nodes from among the computing nodes.

[0111] The baseline node set includes computational nodes whose proximity meets the proximity screening criteria. The baseline node set may include computational nodes whose proximity ranks in the top 30%.

[0112] Specifically, based on the proximity of each computing node, a set of benchmark nodes, namely the Reputation Inheritance Proof Committee, is selected from among the computing nodes. .

[0113] For example, refer to Figure 7 S601 specifically includes: Step 1 When the satellite of Less than the system security threshold At that time, the satellite enters the soft boundary buffer zone; Step 2, the satellite Selected Genesis Transaction, a platform for issuing credibility verification tasks. Step 3, satellite Perform reputation verification tasks and publish transactions. .

[0114] First, by using soft boundary triggering, early warnings are provided. When the remaining service time is less than the system's security threshold, the migration process is triggered, allowing sufficient exchange time before the satellite leaves the current area to avoid mission failure or loss of reputation due to sudden departure. Second, a genesis transaction is published, and committee elections ensure decentralization. The genesis transaction serves as an anchor point, forming a complete cryptographic evidence chain.

[0115] S603, each computing node in the base node set independently calculates the credibility of the temporary task issuing node and uses its private key to perform digital signature.

[0116] Specifically, for each computing node in the set of reference nodes, i.e., the satellites within the committee... It needs to be based on the current situation. State-independent computation credibility ,use private key Digitally sign the results and to China releases credit inheritance certificate task transaction The transaction payload .

[0117] First, each committee member independently calculates the credibility score and signs it; an error or breach by a single member does not affect the overall result. Second, each transaction includes a credibility score and a signature, allowing for subsequent verification of the signature's validity via public key, ensuring the authenticity of the data source.

[0118] S604: After the reputation inheritance proof task reaches consensus in the local DAG blockchain ledger, the node to be migrated collects the main chain transactions and digital signatures, and uses BLS aggregate signatures to generate committee aggregate proofs.

[0119] Specifically, when the credit inheritance proof task exist After China and the US reached a consensus, the satellite... collect Transactions on the main chain of the mission will be received by the satellite set participating in this consensus. and the corresponding transaction payload. Satellite The signature will be aggregated based on BLS. Digital signature Adding the results on the elliptic curve yields the committee's proof of aggregation. BLS aggregation compresses multiple independent signatures into a fixed-length aggregated signature, reducing storage space and making it suitable for scenarios with limited spaceborne link bandwidth.

[0120] For example, refer to Figure 7S603 and S604 include: Step 4, satellite Collect main chain transactions to obtain a consensus set. With digital signatures Step 5, satellite Aggregate digital signatures Fly in with proof of credit .

[0121] First, verifying aggregate signatures requires only one bilinear pairing operation (or a constant number of times), and the verification time is independent of the domain, making it much faster than verifying signatures one by one. Second, based on the elliptic curve discrete logarithm problem, attackers cannot forge legitimate signatures without knowing the private key, ensuring the cryptographic security of the reputation proof. Third, aggregate signatures are lightweight credentials independent of the original regional ledger, and can be carried by satellites into new regions without needing to query the original regional ledger online.

[0122] S605, the node to be migrated writes the committee aggregation proof into the local DAG blockchain ledger corresponding to the new service area.

[0123] Specifically, when the node to be migrated (satellite) Flying into the new service area At that time, to the new regional DAG ledger China issued a special statement of credit inheritance for the transaction. The transaction payload It includes the reputation status of the node (satellite) to be migrated before the switch region, the trusted proof of that reputation status, and the set of satellite nodes that have completed the trusted proof. (If the DAG ledger...) There are no satellites in China The node chain indicates the nodes to be migrated. First flight into the service area ,trade will be as The initial head node of the node chain; if Satellites already exist in China The node chain indicates Previously joined the region And participate in consensus calculations and transactions. This will link to the end of the original node chain. Transaction It needs to be isolated from the mission chain and used only as a satellite. In the region The origin of the node chain during this service cycle. For In the region All subsequent transactions that participate in consensus calculation and are subsequently published must be linked to the transaction in sequence. after.

[0124] Specifically, the nodes to be migrated will aggregate the committee's proof. After being written into the local DAG blockchain ledger corresponding to the new service area, for the access service area TP, in relation to satellite When conducting a credit assessment, it can be done through transactions. Get Aggregators based on constellation satellite public key registry use the same method. China Satellite Node Public Key and will and Perform bilinear mapping verification. If the verification passes, TP will declare the reputation inheritance transaction. The historical reputation of the satellite was incorporated into this reputation assessment, thereby achieving satellite The credit status is continuous.

[0125] For example, refer to Figure 7 S605 includes: Step 6, satellite Towards Send a special reputation inheritance statement transaction Step 7, TP passes verification. , will satellite Historical reputation is included in the reputation assessment. Among these, the satellite at this time... The node to be migrated. Satellite Incorporating historical reputation into the reputation assessment can mean that the committee aggregates the proof and writes it into the local DAG blockchain ledger corresponding to the new service area.

[0126] Firstly, Satellite enters new area Send proactively afterwards Firstly, it eliminates the need for proactive probing in new areas, reducing discovery latency and enabling rapid network access. Secondly, it allows for TP verification in new areas. (Including verification of BLS aggregate signatures) (The correctness of the verification) Once verified, historical reputation is directly incorporated into the current assessment, avoiding the cold start problem of accumulating trust from scratch.

[0127] In this embodiment, the cross-regional migration of reputation proof enables seamless connection between node identity and trust level, ensuring the continuity and reliability of services provided by onboard nodes when dynamically switching regions.

[0128] In one embodiment, a method for maintaining the reputation and scheduling tasks of trusted spaceborne edge computing nodes based on DAG blockchain is provided, which can be applied to the Starlink Phase I low-Earth orbit satellite constellation scenario (which could be...). Figure 1A simulation experiment was established using the model shown, simulating the entire process of task release, task distribution, worker node computation, transaction release, Gossip broadcast, DAG ledger growth, and task consensus confirmation. The simulation parameters are shown in Table 1. The simulation experiment was analyzed from four aspects: 1) Verifying the contribution of each core mechanism to system performance improvement through ablation comparison of multiple mechanisms; 2) Analyzing the security robustness of the above data processing methods under different attack intensities by changing the proportion of malicious nodes under a fixed constellation size; 3) Discussing the impact of regional division granularity on system performance by changing the number of geographic service areas; 4) Analyzing the scalability and stability of the system under different business load pressures by changing the task arrival rate.

[0129] Table 1. Simulation Experiment Parameter Settings ; The experiment uses task consensus success rate, average reputation value of each satellite, average load variance of each region, average number of times a malicious node is selected, and average reputation value of a malicious node as the main evaluation indicators. Among them, task consensus success rate is used to measure whether redundant computing tasks can complete a reliable consensus within the satellite service window; average reputation value of each satellite reflects the positive incentive effect of the system on honest service behavior; average load variance of each region is used to measure the load balance of the scheduling strategy in a local region; and average number of times a malicious node is selected and average reputation value reflect whether the system can continuously identify and prevent malicious nodes from participating in the computation.

[0130] Furthermore, set in Service area within time slot The set of all computational tasks executed in the process is .satellite Able to perform tasks The prerequisite for making an effective contribution is that the consensus confirmation time of its calculation results in the local DAG ledger is [missing information]. Earlier than the time the node leaves the service area At the same time, for a task to reach consensus, the number of redundant computation results in effective aggregation on the main chain must be greater than [a certain value]. Therefore, service area Mid-task consensus success rate Defined as: ; Furthermore, a normalization coefficient is introduced. , Satellite in time slot The comprehensive load characteristics are quantified as ,area The average load across all nodes is: ; Therefore, the region In the time slot Load variance Defined as: ; For example, Figure 8 The paper presents comparative results on five metrics for enabling regionalization, soft boundary, cross-regional reputation maintenance, and entropy-weighted TOPSIS star selection and scheduling mechanisms. The baseline represents the simulation results for maintaining the global DAG ledger. Figure 8 (a) and Figure 8 (b) Throughout the simulation, the success rate of mission consensus and the average reputation value of satellites remained at extremely low levels, indicating that in the scenario of a giant low-Earth orbit satellite constellation, maintaining the global DAG ledger would result in serious state synchronization overhead. Mission scheduling, state updates and consensus verification all need to be propagated across a large network, which makes it difficult to meet the timeliness requirements of redundant computing tasks. This proves the necessity of geographic service regionalization and local DAG ledger maintenance mechanisms.

[0131] In comparison, FF-Greedy, TF-Greedy, TT-Greedy, and the aforementioned data processing method (Proposed) all employ a regionalized local management approach, resulting in significantly higher task consensus success rates and average satellite reputation values ​​compared to the baseline. FF-Greedy, lacking soft boundary and reputation maintenance mechanisms and using a greedy highest-reputation satellite selection strategy, achieves a certain consensus success rate in the later stages of simulation, but its overall convergence speed and final performance are lower than other schemes that utilize enhancement mechanisms. TF-Greedy, by introducing a soft boundary mechanism, further improves its task consensus success rate, indicating that soft boundaries effectively eliminate critical satellites and reduce task consensus interruptions. TT-Greedy, by enabling both soft boundaries and a cross-regional reputation inheritance mechanism, achieves success rates and average reputation values ​​closer to those of the aforementioned data processing method (Proposed), demonstrating that continuous maintenance of reputation status enhances the node trust selection capability in cross-regional service scenarios.

[0132] However, relying solely on the highest reputation for greedy star selection remains clearly insufficient. Figure 8(c) The average load variance of FF-Greedy and TF-Greedy in each region is relatively high and fluctuates significantly during simulation. Although TT-Greedy improves the continuity of node trustworthiness through reputation inheritance, its star selection strategy is still the highest reputation greedy selection, thus it still has shortcomings in load balancing. In contrast, the data processing method (Proposed) mentioned above incorporates reputation value, remaining service time, and node load into the entropy-weighted TOPSIS multi-attribute decision-making process. While prioritizing node trustworthiness and service stability, it can allocate some tasks to candidate nodes with reliable reputation and low load, thus keeping the load variance at a low level.

[0133] Figure 8 (d) and Figure 8 (e) This further demonstrates the ability of different schemes to suppress malicious nodes. In FF-Greedy and TF-Greedy, malicious nodes are selected on average a higher number of times, and their average reputation value increases, indicating that in the absence of effective reputation maintenance and cross-regional reputation inheritance, malicious nodes may gain opportunities to participate in the computation through region switching. After introducing reputation inheritance in TT-Greedy, the number of times malicious nodes are selected decreases significantly. The proposed data processing method further performs multi-attribute comprehensive ranking, making the star selection process no longer solely reliant on a single reputation indicator, further reducing the probability of malicious nodes entering the worker node set. Simultaneously, the average reputation value of malicious nodes continues to decrease, indicating that the system can exert long-term constraints on malicious behavior.

[0134] To further evaluate the robustness of SERC-CRRASA under different attack intensities, experiments were conducted with a fixed satellite constellation size and the same workload on honest nodes. The proportion of malicious nodes was set to 10%, 20%, 30%, 40%, and 50%, respectively. The corresponding results are as follows: Figure 9 As shown.

[0135] Depend on Figure 9 (a) When the proportion of malicious nodes increases from 10% to 50%, the overall success rate of the system's task consensus remains at a high level. However, as the proportion of malicious nodes increases, the convergence speed of the success rate curve decreases in the early stages. This is because as the number of malicious nodes increases, the proportion of low-trust nodes in the candidate node set rises. The algorithm needs to continuously identify and reduce the selection priority of malicious nodes through reputation feedback. Therefore, the system will experience a longer reputation differentiation process in the early stages of simulation. However, from the overall trend, even with a high proportion of malicious nodes, the system can still gradually complete the screening of trustworthy nodes and maintain a high task consensus success rate.

[0136] Figure 9(b) shows that, under different proportions of bad nodes, the average reputation value of each satellite in the system steadily increases with the increase of simulation time slots, and the curves as a whole show a near-linear growth trend. This result is consistent with the experimental setup. To ensure that the comparative experiment only examines the impact of the single variable of the proportion of bad nodes, the task arrival rate is reduced accordingly when the proportion of bad nodes increases, so that the effective task load undertaken by honest nodes remains basically consistent under different experimental scenarios. As the proportion of bad nodes increases from 10% to 50%, the average reputation value of the system does not decrease or stagnate, indicating that under scenarios with a high proportion of bad nodes, SERC-CRRASA can still effectively identify and schedule honest nodes to participate in redundant calculations, so that the normal service behavior of honest nodes continues to receive positive reputation feedback.

[0137] Depend on Figure 9 (c) The average number of times a malicious node is selected increases with the proportion of malicious nodes. This is because as the proportion of malicious nodes increases, the probability of malicious nodes appearing in the candidate set also increases. However, in the later stages of the simulation, the growth rate of the average number of times a malicious node is selected slows down, indicating that SERC-CRRASA can effectively identify malicious nodes and avoid their selection in later stages. Meanwhile... Figure 9 (d) also shows that the average reputation value of malicious nodes generally shows a downward trend under different proportions of malicious nodes, indicating that even if a malicious node is selected for some tasks, its malicious behavior will still be recorded through the local DAG ledger and trigger reputation penalties, thereby reducing the probability of it being selected in the future.

[0138] The number of geographic service areas directly affects the maintenance scale of a local DAG, the number of candidate satellites within the area, and the frequency of cross-regional reputation migration. To analyze the impact of the granularity of area partitioning on the performance of the above data processing methods, the experiment divided the coverage area of ​​the LEO satellite constellation into equal-area geographic service areas of M=9, 16, 25, 36, 49, 64, and 81, respectively. The results are as follows: Figure 10 As shown.

[0139] Depend on Figure 10(a) Under different numbers of geographic service areas, the overall success rate of task consensus can reach a relatively high level, indicating that the above data processing method has a certain adaptability to regional division. However, the convergence speed and final stability level still vary under different M values. When the number of regions is large, the success rate of task consensus is relatively low. The main reason is that as the number of regions increases, the coverage of each service area shrinks, and the number of active satellites available in a single region decreases. This results in fewer trusted nodes that TP can choose when performing redundant computing tasks, and the space for the scheduling algorithm to trade off reputation, remaining service time, and load is compressed. When there are insufficient high-reputation, long-staying, and low-load nodes in a certain region, the system has to select suboptimal nodes to participate in the task, thereby affecting the stability of task execution and the effect of suppressing malicious nodes.

[0140] Figure 10 (b) indicates that as the number of geographic service areas increases, the average number of times a malicious node is selected generally shows an upward trend, further illustrating that overly fine regional division weakens the diversity of the candidate set, making it easier for malicious nodes to gain opportunities to participate in tasks within local areas. Although Figure 10 (c) shows that the average reputation score of malicious nodes generally declines, indicating that the reputation penalty mechanism remains effective. However, as the number of times a malicious node is selected increases, the system still needs to incur higher costs for identification and punishment. Figure 10 The average reputation value results in (d) show that the overall reputation accumulation of the system varies relatively little with different numbers of regions, indicating that changes in the number of regions do not directly disrupt the basic scheduling stability of the data processing method described above. However, while overly fine partitioning can reduce the maintenance scale of a single local DAG, it can lead to an insufficient number of candidate nodes within a region, thereby reducing the optimization space for multi-attribute star selection. Overly coarse partitioning, on the other hand, can provide a larger candidate set, but it increases the pressure on local ledger maintenance and state synchronization. Therefore, in actual deployment, a moderate number of geographical regions should be selected to achieve a balance between local ledger maintenance overhead, candidate node diversity, and secure scheduling capabilities.

[0141] To verify the applicability of the above data processing method under different business load pressures, the experiment set the task arrival rate to 2, 5, 10, 15, and 20, respectively. The corresponding results are as follows: Figure 11 As shown.

[0142] Depend on Figure 11 (a) As the task arrival rate increases from 2 to 20, the overall success rate of system task consensus remains at a high level, indicating that the above data processing method can continue to complete the scheduling of trusted redundant computing tasks and maintain a relatively stable redundant computing consensus capability when the task density increases.

[0143] Figure 11(b) shows that as the mission arrival rate increases, the average load variance in each region also increases, indicating that a high mission arrival rate will cause satellites in the region to undertake more concurrent computing tasks, increasing the system load pressure. However, the load variance did not grow out of control and still fluctuated within a certain range, indicating that the above data processing method can still maintain basic load balancing capability under high mission pressure.

[0144] Figure 11 (c) This further shows that the average number of times a malicious node is selected increases with the task arrival rate. This is because as the number of tasks increases, the total number of star selections performed by the system increases, and the number of times a malicious node is selected also increases. Combined with... Figure 11 (d) Joint analysis of the average reputation value of malicious nodes shows that as the task arrival rate increases, the overall decline in the average reputation value of malicious nodes increases. This indicates that in scenarios with high task arrival rates, although malicious nodes may have more opportunities to participate, their malicious behavior will also be recorded and punished more frequently, resulting in a faster decline in the average reputation value.

[0145] This embodiment proposes a method for maintaining the reputation of trusted satellite-borne edge computing nodes and scheduling tasks based on DAG blockchain. Addressing the challenges of continuously acquiring the trusted state of nodes in a giant LEO constellation, the high overhead of confirming redundant computation results, and the potential for load imbalance caused by single reputation-driven scheduling, this method first constructs a local DAG ledger framework oriented towards the geographic service area. This records and confirms satellite node computation behavior, task result submission processes, and area switching information within a local scope, thereby reducing the communication pressure and maintenance overhead caused by global ledger synchronization. Second, a soft boundary mechanism and a cross-regional reputation inheritance mechanism based on BLS aggregate signatures are designed, enabling satellites to avoid the risk of task consensus interruption before leaving their current service area and to transfer historical reputation states to the new service area in a verifiable manner. Subsequently, a node reputation awareness mechanism based on local DAG ledger retrieval is proposed, and an entropy-weighted TOPSIS multi-attribute joint satellite selection algorithm is designed, combining node reputation, remaining service time, and comprehensive load factors. Experimental results show that the method of the present invention can improve the success rate of task consensus and the average reputation value of the system, reduce the regional load variance, and continuously suppress the selection probability and reputation gain of malicious nodes under different conditions of malicious node ratio, regional division granularity and task arrival rate.

[0146] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0147] Based on the same inventive concept, this application also provides a data processing apparatus for a trusted spaceborne edge computing node of a DAG blockchain, used to implement the data processing method for the trusted spaceborne edge computing node of the DAG blockchain described above. The solution provided by this apparatus is similar to the implementation scheme described in the above method. Therefore, the specific limitations of one or more embodiments of the data processing apparatus for a trusted spaceborne edge computing node of a DAG blockchain provided below can be found in the limitations of the data processing method for the trusted spaceborne edge computing node of the DAG blockchain described above, and will not be repeated here.

[0148] In one embodiment, such as Figure 12 As shown, a data processing device for a trusted onboard edge computing node of a DAG blockchain is provided, comprising: a partitioning module 1201, an information acquisition module 1202, a filtering module 1203, a ledger update module 1204, and an inheritance module 1205, wherein: The partitioning module 1201 is used to divide the spaceborne edge computing service into multiple fixed geographical service areas, wherein each service area includes multiple computing nodes and multiple task publishing nodes; the task publishing nodes are used to publish computing tasks to the computing nodes, and each computing node is used to maintain a local DAG blockchain ledger corresponding to the service area, and the DAG blockchain ledger is used to record the processing result information of the computing tasks. The information acquisition module 1202 is used to acquire the node reputation, remaining service time and load status of each computing node, and construct a multi-attribute evaluation index for each computing node based on the node reputation, remaining service time and load status. The filtering module 1203 is used to filter the computing nodes based on the multi-attribute evaluation index to obtain at least two candidate nodes that meet the filtering conditions; wherein, each candidate node is used to process the basic task published by the task publishing node to obtain the basic result of each candidate node for the basic task. Ledger update module 1204 is used to update the local DAG blockchain ledger based on the aforementioned basic results; The inheritance module 1205 is used to update the reputation inheritance proof generated by the computing node to the local DAG blockchain ledger corresponding to the new service area when the computing node crosses the boundary of the service area.

[0149] The modules in the data processing device of the trusted satellite-borne edge computing node of the aforementioned DAG blockchain can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0150] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 13 As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores local DAG blockchain ledger data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a data processing method for a trusted onboard edge computing node of a DAG blockchain.

[0151] Those skilled in the art will understand that Figure 13 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0152] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the above-described method steps.

[0153] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the above-described method steps.

[0154] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the above-described method steps.

[0155] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0156] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0157] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A data processing method for a trusted spaceborne edge computing node based on DAG blockchain, characterized in that, The method includes: The spaceborne edge computing service space is divided into multiple fixed geographical service areas, wherein each service area includes multiple computing nodes and multiple task publishing nodes; the task publishing nodes are used to publish computing tasks to the computing nodes, and each computing node is used to maintain a local DAG blockchain ledger corresponding to the service area, and the DAG blockchain ledger is used to record the processing result information of the computing tasks; For each computing node, the node reputation, remaining service time, and load status of the computing node are obtained, and a multi-attribute evaluation index for each computing node is constructed based on the node reputation, remaining service time, and load status. Based on the multi-attribute evaluation index, candidate nodes that meet the screening conditions matching the redundancy requirements are obtained from each of the computing nodes; wherein, the candidate nodes are used to process the basic tasks published by the task publishing node and obtain the basic results of the candidate nodes for the basic tasks. The local DAG blockchain ledger is updated based on the aforementioned basic results; When the computing node crosses the boundary of a service area, the reputation inheritance proof generated by the computing node is updated in the local DAG blockchain ledger corresponding to the new service area.

2. The method according to claim 1, characterized in that, The process of obtaining the node reputation of the computing node includes: Obtain the task index and node index corresponding to each computing node; The task publishing node locates the latest transaction of each computing node based on the node index before task scheduling; wherein, the latest transaction represents the task processing result most recently at the current time. When the computing node is in a basic task scenario, backtracking is performed along the directed edges of the node chain based on the node chain of the latest transaction and the task index to obtain the number of main chains and parasitic chains of the node chain to which the computing node belongs; based on the number of main chains and the number of parasitic chains, the node reputation of the computing node is determined, wherein the number of main chains is positively correlated with the node reputation, and the number of parasitic chains is negatively correlated with the node reputation; When the computing node is in a reputation inheritance proof task scenario, the historical reputation base of the computing node is obtained by tracing back to the most recent reputation inheritance statement transaction of the computing node; based on the historical reputation base and the verification result of the historical reputation base, the node reputation of the computing node is determined.

3. The method according to claim 1, characterized in that, The process of filtering candidate nodes based on the multi-attribute evaluation index to obtain candidate nodes that meet the filtering conditions matching the redundancy requirements includes: Soft boundary filtering is performed on each of the computing nodes to filter out intermediate nodes that meet the soft boundary filtering conditions. Entropy weight processing is performed on the multi-attribute evaluation index corresponding to each intermediate node to obtain the objective weight of each intermediate node for different attribute evaluation indexes. Based on the decision matrix constructed using the objective weights mentioned above, the globally optimal positive ideal solution and the worst negative ideal solution are determined. For each intermediate node, the distance between it and the global optimal positive ideal solution and the worst negative ideal solution is calculated based on the position of the intermediate node in the decision matrix, and the proximity of the intermediate node is determined. The greater the proximity, the closer the intermediate node is to the optimal state under the comprehensive consideration of honesty, reliability, stable residence and computational load. The intermediate nodes are sorted in descending order according to their proximity to obtain candidate nodes that meet the screening criteria that match the redundancy requirements.

4. The method according to claim 3, characterized in that, When the computing node crosses the boundary of a service area, updating the reputation inheritance proof generated by the computing node to the local DAG blockchain ledger corresponding to the new service area includes: When the computing node crosses the boundary of the service area, the computing node is used as a temporary task publishing node to publish the reputation inheritance proof task in the local DAG blockchain ledger of the original service area. Based on the proximity of each computing node, a set of benchmark nodes is selected from each computing node; wherein, the set of benchmark nodes includes computing nodes whose proximity meets the proximity selection criteria. Each computing node in the set of baseline nodes independently calculates the reputation of the temporary task issuing node and performs digital signature using its private key; After the reputation inheritance proof task reaches consensus in the local DAG blockchain ledger, the node to be migrated collects the main chain transactions and the digital signatures, and uses BLS aggregate signatures to generate committee aggregate proofs. The node to be migrated writes the committee aggregation proof into the local DAG blockchain ledger corresponding to the new service area.

Images