Collaborative operation method and system for edge computing nodes

Abstract

The invention relates to the technical field of edge computing, and discloses a collaborative operation method and system for edge computing nodes, and the method comprises the steps: matching a first node group, computing a comprehensive cognitive index, screening a second node group, updating the second node group, and computing a comprehensive verification index. The system corresponds to the method. According to the method, before task allocation, task features are deeply analyzed, nodes are matched with functions and computing power, blind allocation is avoided, the nodes are comprehensively evaluated through self-cognition and mutual cognition, the second node group better meeting the task computing power requirement is screened out, the resource utilization rate is increased, the node group is dynamically updated in task execution, and the task allocation efficiency is improved. Low-efficiency nodes are eliminated in time, appropriate nodes are supplemented, it is guaranteed that calculation tasks are smoothly and efficiently carried out, a maintenance instruction is generated based on the comprehensive verification index, passive maintenance is changed into active maintenance, resource waste and potential safety hazards are reduced, and the overall efficiency and reliability of edge calculation node collaborative operation are improved.

Description

Technical Field

[0001] The present application relates to the field of edge computing technology, and specifically to a collaborative operation method and system for edge computing nodes. Background Art

[0002] With the rapid development of technologies such as the Internet of Things and 5G, edge computing has emerged and has been widely used in many fields. Edge computing stores computing and data close to the data source or user end to reduce data transmission delays and improve response speed. However, in terms of the collaborative operation of edge computing nodes, existing technologies still have many problems, as follows:

[0003] Traditional edge computing node collaboration often lacks a comprehensive and accurate recognition and evaluation mechanism for nodes. When allocating tasks, it is difficult to reasonably match nodes based on their actual capabilities and status, which may cause some nodes to be overloaded while others are idle, reducing overall computing efficiency.

[0004] During task execution, there is a lack of effective dynamic adjustment strategies, which prevents the timely elimination of inefficient nodes and the addition of suitable nodes, which may cause computing tasks to face the risk of stalling, delay or even failure.

[0005] The maintenance of edge computing nodes mostly relies on regular manual inspections or fixed maintenance cycles. There is a lack of an intelligent maintenance instruction generation mechanism based on the actual operation status of the nodes and the performance of collaborative operations. This makes it difficult to achieve precise maintenance, which can easily lead to waste of resources and potential safety hazards, and cannot fully leverage the advantages of collaborative operations of edge computing nodes.

[0006] In summary, a new technical solution for the collaborative operation of edge computing nodes is urgently needed to solve the above problems. Summary of the invention

[0007] The purpose of this application is to provide a collaborative operation method and system for edge computing nodes to solve the technical problems raised in the above background technology.

[0008] To achieve the above objectives, this application discloses the following technical solutions:

[0009] In a first aspect, the present application discloses a collaborative operation method of edge computing nodes, the method comprising:

[0010] S1: parsing the computing task and matching to obtain a corresponding first node group; wherein the first node group is a set of edge computing nodes that functionally correspond to the computing task;

[0011] S2: Comprehensively recognize each edge computing node in the first node group to obtain a corresponding comprehensive recognition index; wherein the comprehensive recognition index is calculated by a self-recognition score and a mutual recognition score, the self-recognition score is the result of the edge computing node performing self-recognition, and the mutual recognition score is the result of the edge computing node performing mutual recognition;

[0012] S3: Utilizing the comprehensive cognitive index of the first node group and combining it with a preset node allocation rule, a second node group is obtained by screening; wherein the node allocation rule is used to screen a set of edge computing nodes that meet the computing power of the computing task in the first node group, and the second node group is a set of edge computing nodes that correspond to the computing task in terms of computing power, and the number of nodes in the second node group is less than or equal to the number of nodes in the first node group;

[0013] S4: Utilize the second node group to perform the computing task, continuously utilize the preset node self-elimination rule and node screening rule to update the second node group, and utilize the updated second node group to perform the computing task until the end after the update; wherein, the node self-elimination rule is used to screen edge computing nodes in the second node group that do not meet the computing efficiency when performing the computing task, and the node screening rule is used to screen edge computing nodes in the first node group that meet the computing power of the computing task;

[0014] S5: Comprehensively verify each edge computing node in the second node group to obtain a corresponding comprehensive verification index and generate a corresponding node maintenance instruction; wherein, the comprehensive verification index is calculated by a self-verification score and a mutual verification score, the self-verification score is the result of self-verification of the edge computing node, the mutual verification score is the result of mutual verification of the edge computing node, and the node maintenance instruction is used to maintain the edge computing node.

[0015] Preferably, the analytical calculation task is specifically:

[0016] Obtaining the computing task;

[0017] The computing task is parsed to obtain a corresponding computing amount, a computing type, and a computing priority; wherein the computing amount, the computing type, and the computing priority are used to match the first node group.

[0018] Preferably, the calculation process of the comprehensive cognitive index is:

[0019] Obtaining the first node group;

[0020] Performing comprehensive cognition on the first node group, wherein the comprehensive cognition includes self-cognition and mutual cognition;

[0021] The self-cognition is the cognition performed by the edge computing node based on its own real-time operation status, and the self-cognition score is calculated;

[0022] The mutual recognition is the recognition performed by the edge computing node based on the real-time operation status of the edge computing nodes of the same type as itself, and the mutual recognition score is calculated;

[0023] The comprehensive cognitive index is calculated using the self-cognition score and the mutual cognition score.

[0024] Preferably, the self-cognition is specifically: based on the real-time operation of the edge computing node and combined with the historical computing efficiency, the self-cognition score is calculated;

[0025] The mutual recognition is specifically: based on the real-time operation status of the edge computing nodes of the same type, the mutual recognition score is calculated in combination with the historical computing efficiency of the edge computing nodes of the same type.

[0026] Preferably, the screening process of the second node group includes:

[0027] Analyze the computing task to obtain a corresponding computing power allocation plan;

[0028] The node allocation rule is called, and in combination with the computing power allocation plan, the edge computing nodes that meet the computing power of the computing task are screened in the first node group to obtain the second node group.

[0029] Preferably, the updating process of the second node group includes:

[0030] Using the node self-elimination rule to filter edge computing nodes in the second node group that do not meet the computing efficiency when executing the computing task, and eliminating the calculated minimum value;

[0031] The node self-elimination rule is specifically as follows:

[0032]

[0033] Among them, EN i is the edge computing node numbered i that is screened out and does not meet the computing efficiency in the second node group when executing the computing task, min is the minimum value operator, The integral of the efficiency P calculated by the edge computing node numbered i in the preset elimination period T, P i_τ is the computational efficiency threshold of the preset edge computing node numbered i, α i is the efficiency compensation value of the edge computing node numbered i, and the efficiency compensation value is used to characterize the compensation for the computing efficiency of the edge computing node during the elimination cycle.

[0034] Preferably, the updating process of the second node group further comprises:

[0035] When the edge computing nodes in the second node group that do not meet the computing efficiency when executing the computing task are screened out, the sub-computing tasks of the edge computing node are obtained, and the node screening rule is called to screen the edge computing nodes in the first node group that meet the computing power of the computing task, and the computing tasks of the eliminated edge computing nodes are assigned to the screened edge computing nodes.

[0036] Preferably, the calculation process of the comprehensive verification index is:

[0037] Obtain each edge computing node in the second node group that executes the computing task;

[0038] Performing comprehensive verification on each edge computing node in the second node group that executes the computing task, wherein the comprehensive verification includes self-verification and mutual verification;

[0039] The self-verification is a verification performed by the edge computing node based on its own operation records, and the self-verification score is calculated;

[0040] The mutual verification is a verification performed by the edge computing node based on the operation records of the edge computing nodes of the same type as itself, and a mutual verification score is calculated;

[0041] The comprehensive verification index is calculated using the self-verification score and the mutual verification score.

[0042] Preferably, the self-verification is specifically: based on the operation record of the edge computing node, the self-verification score is calculated in combination with the self-cognition evaluation;

[0043] The mutual verification is specifically: based on the operation records of the edge computing nodes of the same type, and combined with the historical computing efficiency of the edge computing nodes of the same type, the mutual verification score is calculated.

[0044] In a second aspect, the present application discloses a collaborative operation system of edge computing nodes, the system is applicable to the collaborative operation method of edge computing nodes as described above, and the system includes:

[0045] A first node group matching module, wherein the first node group matching module is configured to: parse the computing task and match to obtain a corresponding first node group; wherein the first node group is a collection of edge computing nodes that functionally correspond to the computing task;

[0046] A comprehensive cognitive index calculation module, wherein the comprehensive cognitive index calculation module is configured to: comprehensively recognize each edge computing node in the first node group to obtain a corresponding comprehensive cognitive index; wherein the comprehensive cognitive index is calculated by a self-cognition score and a mutual cognition score, wherein the self-cognition score is a result of the edge computing node performing self-cognition, and the mutual cognition score is a result of the edge computing node performing mutual cognition;

[0047] A second node group screening module, wherein the second node group screening module is configured to: use the comprehensive cognitive index of the first node group and a preset node allocation rule to screen and obtain a second node group; wherein the node allocation rule is used to screen a set of edge computing nodes that meet the computing power of the computing task in the first node group, the second node group is a set of edge computing nodes that correspond to the computing task in terms of computing power, and the number of nodes in the second node group is less than or equal to the number of nodes in the first node group;

[0048] A second node group update module, wherein the second node group update module is configured to: use the second node group to perform the computing task, continuously use the preset node self-elimination rule and node screening rule to update the second node group, and use the updated second node group to perform the computing task until the end after the update; wherein the node self-elimination rule is used to screen edge computing nodes in the second node group that do not meet the computing efficiency when performing the computing task, and the node screening rule is used to screen edge computing nodes in the first node group that meet the computing power of the computing task;

[0049] A comprehensive verification index calculation module, wherein the comprehensive verification index calculation module is configured to: comprehensively verify each edge computing node in the second node group to obtain a corresponding comprehensive verification index and generate a corresponding node maintenance instruction; wherein the comprehensive verification index is calculated by a self-verification score and a mutual verification score, the self-verification score is the result of self-verification by the edge computing node, the mutual verification score is the result of mutual verification by the edge computing node, and the node maintenance instruction is used to maintain the edge computing node.

[0050] Beneficial effects: The collaborative operation method and system of the edge computing nodes of the present application utilizes a series of processes such as parsing computing tasks and matching the first node group, comprehensively cognizing and calculating the comprehensive cognitive index of each node, selecting the second node group based on the comprehensive cognitive index combined with the node allocation rules, executing tasks and continuously updating the second node group, and comprehensively verifying and generating maintenance instructions to achieve efficient and accurate collaborative operation of edge computing nodes. Before task assignment, the task characteristics are deeply analyzed, and nodes are matched in combination with functions and computing power to avoid blind assignment. Through self-cognition and mutual cognition, the nodes are comprehensively evaluated to select the second node group that better meets the task computing power requirements and improve resource utilization. During task execution, the node group is dynamically updated, inefficient nodes are eliminated in a timely manner, and suitable nodes are supplemented to ensure smooth and efficient computing tasks. Finally, maintenance instructions are generated based on the comprehensive verification index, changing passive to active maintenance, reducing resource waste and safety hazards, and improving the overall efficiency and reliability of collaborative operations of edge computing nodes. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0052] Figure 1 A flowchart of a collaborative operation method of edge computing nodes provided in an embodiment of the present application;

[0053] Figure 2 A structural block diagram of a collaborative operation system of edge computing nodes provided in an embodiment of the present application. DETAILED DESCRIPTION

[0054] The technical solutions in the embodiments of the present application are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application.

[0055] In this article, the term "comprising" is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of more restrictions, the elements defined by the sentence "comprising..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0056] The first aspect of this embodiment discloses Figure 1 A collaborative operation method of an edge computing node is shown, the method comprising:

[0057] S1: parsing the computing task and matching to obtain the corresponding first node group; wherein the first node group is a collection of edge computing nodes that functionally correspond to the computing task;

[0058] S2: Comprehensively recognize each edge computing node in the first node group to obtain a corresponding comprehensive recognition index; wherein the comprehensive recognition index is calculated by a self-recognition score and a mutual recognition score, the self-recognition score is the result of the edge computing node's self-recognition, and the mutual recognition score is the result of the edge computing node's mutual recognition;

[0059] S3: Using the comprehensive cognitive index of the first node group and combining it with a preset node allocation rule, a second node group is obtained by screening; wherein the node allocation rule is used to screen a set of edge computing nodes that meet the computing power of the computing task in the first node group, and the second node group is a set of edge computing nodes that correspond to the computing task in terms of computing power, and the number of nodes in the second node group is less than or equal to the number of nodes in the first node group;

[0060] S4: Utilize the second node group to perform the computing task, continuously utilize the preset node self-elimination rule and node screening rule to update the second node group, and utilize the updated second node group to perform the computing task until the end after the update; wherein the node self-elimination rule is used to screen the edge computing nodes in the second node group that do not meet the computing efficiency when performing the computing task, and the node screening rule is used to screen the edge computing nodes in the first node group that meet the computing power of the computing task;

[0061] S5: Comprehensively verify each edge computing node in the second node group to obtain a corresponding comprehensive verification index and generate a corresponding node maintenance instruction; wherein the comprehensive verification index is calculated by a self-verification score and a mutual verification score, the self-verification score is the result of self-verification by the edge computing node, the mutual verification score is the result of mutual verification by the edge computing node, and the node maintenance instruction is used to maintain the edge computing node.

[0062] It should be noted that the node maintenance instruction of this embodiment is an instruction for maintaining an edge computing node that is well known to existing technical personnel in this field, and its triggering mechanism can be but is not limited to comparing the obtained comprehensive verification index with the preset comprehensive verification index to trigger the node maintenance instruction.

[0063] Based on the above, this embodiment utilizes a series of processes such as parsing computing tasks and matching the first node group, comprehensively cognizing and calculating the comprehensive cognitive index of each node, screening the second node group based on the comprehensive cognitive index combined with the node allocation rules, executing tasks and continuously updating the second node group, and comprehensively verifying and generating maintenance instructions to achieve efficient and accurate edge computing node collaboration. Before task assignment, in-depth analysis of task characteristics, combining functions with computing power matching nodes, to avoid blind allocation. Through self-cognition and mutual cognition, the nodes are comprehensively evaluated to select the second node group that better meets the task computing power requirements, thereby improving resource utilization. During task execution, the node group is dynamically updated, inefficient nodes are eliminated in a timely manner, and suitable nodes are supplemented to ensure smooth and efficient computing tasks. Finally, maintenance instructions are generated based on the comprehensive verification index, changing passive to active maintenance, reducing resource waste and safety hazards, and improving the overall performance and reliability of edge computing node collaboration.

[0064] Specifically, the analytical calculation tasks are as follows:

[0065] Get computing tasks;

[0066] The computing task is parsed to obtain the corresponding computing amount, computing type and computing priority; wherein the computing amount, computing type and computing priority are used to match the first node group.

[0067] It should be noted that the analytical calculation task of this embodiment is based on the experience known to technical personnel in this field, and the division of calculation amount, calculation type and calculation priority corresponds to the matching of the first node group, thereby achieving node screening in a larger range.

[0068] Through the above, this embodiment provides an accurate basis for subsequent node matching and task allocation, thereby accurately locating the first node group corresponding to the task function among many edge computing nodes, making the initial matching of tasks and nodes more targeted, and avoiding node selection errors or resource waste due to unclear task information. In a simple example, when processing image recognition tasks, nodes with image processing capabilities can be quickly screened out based on the amount of calculation and type, laying the foundation for subsequent efficient collaborative operations, and effectively improving the accuracy and efficiency of edge computing node collaborative operations at the initial stage of the task.

[0069] Specifically, the calculation process of the comprehensive cognitive index is:

[0070] Get the first node group;

[0071] Conduct comprehensive cognition on the first node group, which includes self-cognition and mutual cognition;

[0072] Self-cognition is the cognition of the edge computing node based on its own real-time operation status, and the self-cognition score is calculated;

[0073] Mutual recognition is the recognition of edge computing nodes based on the real-time operation status of edge computing nodes of the same type as themselves, and the mutual recognition score is calculated;

[0074] The comprehensive cognitive index was calculated using the self-cognition score and the mutual cognition score.

[0075] It should be noted that the comprehensive cognition of this embodiment is an optimization of the cognition of the edge computing node, and the optimization of self-cognition is achieved through the data of mutual cognition, thereby providing more accurate data for the screening of the second node group. Furthermore, the comprehensive cognition index calculation method of this embodiment can be, but is not limited to, correspondingly assigning self-cognition scores and mutual cognition scores for calculation.

[0076] Through the above, this embodiment realizes a comprehensive and objective evaluation of edge computing nodes. Self-cognition is based on its own real-time operation status and historical computing efficiency, and mutual cognition is based on the real-time and historical status of nodes of the same type. The comprehensive cognitive index obtained by combining the two can reflect the node status in multiple dimensions, thereby overcoming the limitations of traditional single-dimensional evaluation and being able to more accurately judge the adaptability of nodes in collaborative operations. In a simple example, when allocating big data processing tasks, nodes with stable and efficient computing can be screened out based on the comprehensive cognitive index, reducing computing delays and errors caused by node performance fluctuations or poor adaptation, thereby improving the reliability and scientificity of node selection in collaborative operations of edge computing nodes.

[0077] Specifically, self-cognition is as follows: based on the real-time operation of the edge computing node and combined with the historical computing efficiency, the self-cognition score is calculated;

[0078] Mutual recognition is specifically: the mutual recognition score is calculated based on the real-time operation status of edge computing nodes of the same type and the historical computing efficiency of edge computing nodes of the same type.

[0079] As a preferred implementation of this embodiment, this embodiment uses a self-cognition score formula to calculate the self-cognition score, wherein the self-cognition score formula is specifically:

[0080]

[0081] Among them, ScS i is the self-awareness score of the edge computing node numbered i, ScS p_i is the self-cognition evaluation of the edge computing node numbered i based on the real-time operation status. It should be noted that the calculation of the self-cognition evaluation can be any existing self-cognition evaluation method of the edge computing node. Exemplarily, the self-cognition evaluation is performed based on the computing efficiency (completion, accuracy, etc.) of the edge computing node. is the average value of the historical computing efficiency of the edge computing node numbered i, thereby realizing the correction of self-cognition of the real-time operation status combined with historical data.

[0082] As a preferred implementation of this embodiment, this embodiment uses a mutual recognition score formula to calculate the mutual recognition score, wherein the mutual recognition score formula is specifically:

[0083]

[0084] Among them, McS i is the mutual recognition score, It is the average value of the historical computing efficiency of edge computing nodes of the same type as the edge computing node numbered i, so as to achieve mutual understanding of the real-time operation status combined with the historical data of collaborative operations.

[0085] Specifically, the screening process of the second node group includes:

[0086] Analyze the computing tasks to obtain the corresponding computing power allocation plan;

[0087] The node allocation rule is called, and in combination with the computing power allocation plan, the edge computing nodes that meet the computing power of the computing task are screened in the first node group to obtain the second node group.

[0088] It should be noted that the node allocation rule of this embodiment is obtained based on the computing power allocation experience known to those skilled in the art.

[0089] Based on the above, this embodiment uses the computing task to analyze and obtain the computing power allocation plan and combines the node allocation rules to screen the second node group, thereby achieving accurate adaptation of task computing power and node resources. According to the computing power requirements analyzed by the computing task, the node allocation rules are called to screen in the first node group to ensure that the selected second node group can meet the task computing power requirements and optimize the number of nodes. In a simple example, when processing complex mathematical model calculation tasks, the required number and type of nodes can be accurately determined according to the computing power allocation plan to avoid excess or insufficient node resources, so that the collaborative operation of edge computing nodes is more reasonable and efficient in computing power allocation, ensuring the smooth progress of tasks.

[0090] Specifically, the update process of the second node group includes:

[0091] The node self-elimination rule is used to filter out edge computing nodes in the second node group that do not meet the computing efficiency when executing the computing task, and the minimum value obtained by calculation is eliminated;

[0092] The specific node self-elimination rules are as follows:

[0093]

[0094] Among them, EN i is the edge computing node with the number i that is screened out and does not meet the computing efficiency in the second node group when executing the computing task. min is the minimum value operator. The integral of the efficiency P calculated by the edge computing node numbered i in the preset elimination period T, P i_τ is the computational efficiency threshold of the preset edge computing node numbered i, α i is the efficiency compensation value of the edge computing node numbered i, and the efficiency compensation value is used to characterize the compensation for the computing efficiency of the edge computing node during the elimination cycle. In a specific example, this embodiment uses the historical records of the compensation for the computing efficiency of the edge computing node during the elimination cycle to fit the efficiency compensation value.

[0095] Through the above, this embodiment uses the node self-elimination rule to realize the dynamic optimization of edge computing nodes during task execution. By setting a screening formula based on computing efficiency integral, threshold and compensation value, inefficient nodes in the second node group can be identified and eliminated in a timely manner. When processing continuous data flow computing tasks, computing jams caused by node performance degradation or failure can be effectively eliminated to ensure the smoothness and stability of task execution. At the same time, the mechanism of eliminating inefficient nodes and supplementing suitable nodes enables the collaborative operation of edge computing nodes to adjust resource allocation in real time according to task progress, thereby improving overall computing efficiency and adaptability.

[0096] Specifically, the updating process of the second node group further includes:

[0097] When the edge computing nodes in the second node group that do not meet the computing efficiency when executing the computing task are screened out, the sub-computing tasks of the edge computing node are obtained, and the node screening rules are called to screen the edge computing nodes with computing power that meet the computing task in the first node group, and the computing tasks of the eliminated edge computing nodes are assigned to the screened edge computing nodes.

[0098] It is understandable that although the edge computing nodes each perform the corresponding sub-computing tasks, for the computing task, the computing task is completed only after each sub-computing task is completed. Therefore, this embodiment designs a node self-elimination rule and configures the corresponding node screening rule to perform node self-update, thereby optimizing the collaborative operation of the edge computing nodes. It should be noted that the node screening rule of this embodiment is a screening rule for computing power based on the experience known to those skilled in the art.

[0099] Through the above, this embodiment utilizes the mechanism of transferring sub-computing tasks when eliminating inefficient nodes to achieve seamless task execution and efficient resource utilization. When it is found that there are nodes in the second node group that do not meet the computing efficiency, their sub-tasks are quickly obtained and suitable nodes are selected in the first node group to undertake. For example, in a multi-stage data processing task, if a node has a problem in the middle stage, its unfinished tasks can be assigned in time to avoid task interruption and idle resources, ensure the continuity and efficiency of the collaborative operation of edge computing nodes, and reduce the impact of node failure or inefficiency on the entire task process.

[0100] Specifically, the calculation process of the comprehensive verification index is:

[0101] Obtain each edge computing node that performs computing tasks in the second node group;

[0102] Performing comprehensive verification on each edge computing node that performs computing tasks in the second node group, the comprehensive verification including self-verification and mutual verification;

[0103] Self-verification is the verification performed by the edge computing node based on its own operation records, and the self-verification score is calculated;

[0104] Mutual authentication is the authentication performed by the edge computing node based on the operation records of the edge computing nodes of the same type as itself, and the mutual authentication score is calculated;

[0105] The comprehensive verification index is calculated using the self-verification score and the mutual verification score.

[0106] It should be noted that the comprehensive verification of this embodiment is an optimization of the verification of edge computing nodes. The purpose of the verification is to provide a data basis for the maintenance of edge computing nodes, and to optimize self-verification through mutual verification data, thereby providing more accurate data for the screening of edge computing node maintenance. Furthermore, the comprehensive verification index calculation method of this embodiment can be, but is not limited to, performing corresponding assignments on the self-verification score and the mutual verification score for calculation.

[0107] Specifically, self-verification is as follows: based on the operation records of edge computing nodes, the self-verification score is calculated in combination with self-cognitive evaluation;

[0108] Mutual verification is specifically as follows: mutual verification scores are calculated based on the operating records of edge computing nodes of the same type and combined with the historical computing efficiency of edge computing nodes of the same type.

[0109] As a preferred implementation of this embodiment, this embodiment uses a self-verification score formula to calculate the self-verification score, wherein the self-verification score formula is specifically:

[0110]

[0111] Among them, SvS v_i is the self-verification score of the edge computing node numbered i, ScS real_i It is a self-verification evaluation of the execution sub-computing task of the edge computing node numbered i, and the method of self-verification evaluation corresponds to the aforementioned self-cognition evaluation, thereby realizing the self-verification correction of the operation status record combined with the self-cognition evaluation.

[0112] As a preferred implementation of this embodiment, this embodiment uses a mutual authentication score formula to calculate the mutual authentication score, wherein the mutual authentication score formula is specifically:

[0113]

[0114] Among them, MvS i To achieve mutual verification of the scores, the real-time operation status can be verified by combining the historical data of the collaborative operation, thereby achieving mutual verification of the scores by combining the historical computing efficiency of the same type of edge computing nodes.

[0115] Based on the above, this embodiment uses self-verification and mutual verification combined with historical computing efficiency to calculate a comprehensive verification index, thereby achieving precise guidance on the generation of edge computing node maintenance instructions. The self-verification score reflects the quality of its own operation, the mutual verification score reflects the comparison of the same category, and the comprehensive verification index combines the two to accurately locate node problems. In a specific example, in the edge computing of the industrial Internet of Things, personalized maintenance instructions can be generated for different node problems based on this index, such as focusing on checking the network module for nodes with unstable data transmission, and optimizing algorithms for nodes with large calculation errors, thereby improving the accuracy and effectiveness of node maintenance in the collaborative operation of edge computing nodes and extending the service life and reliability of nodes.

[0116] The second aspect of this embodiment discloses Figure 2 A collaborative operation system of edge computing nodes is shown, and the system is applicable to the collaborative operation method of edge computing nodes as described above, and the system includes:

[0117] A first node group matching module, wherein the first node group matching module is configured to: parse the computing task and match to obtain a corresponding first node group; wherein the first node group is a collection of edge computing nodes that functionally correspond to the computing task;

[0118] A comprehensive cognitive index calculation module, wherein the comprehensive cognitive index calculation module is configured as follows: each edge computing node in the first node group of comprehensive cognition obtains a corresponding comprehensive cognitive index; wherein the comprehensive cognitive index is calculated by a self-cognition score and a mutual cognition score, wherein the self-cognition score is the result of the self-cognition of the edge computing node, and the mutual cognition score is the result of the mutual cognition of the edge computing node;

[0119] A second node group screening module, the second node group screening module is configured to: use the comprehensive cognitive index of the first node group, combined with a preset node allocation rule to screen and obtain a second node group; wherein the node allocation rule is used to screen a set of edge computing nodes that meet the computing power of the computing task in the first node group, the second node group is a set of edge computing nodes corresponding to the computing task in terms of computing power, and the number of nodes in the second node group is less than or equal to the number of nodes in the first node group;

[0120] A second node group update module, wherein the second node group update module is configured to: use the second node group to perform computing tasks, continuously use preset node self-elimination rules and node screening rules to update the second node group, and use the updated second node group to perform computing tasks until completion after the update; wherein the node self-elimination rule is used to screen edge computing nodes in the second node group that do not meet the computing efficiency when performing computing tasks, and the node screening rule is used to screen edge computing nodes in the first node group that meet the computing power of the computing tasks;

[0121] A comprehensive verification index calculation module, the comprehensive verification index calculation module is configured as follows: comprehensively verifying each edge computing node in the second node group to obtain a corresponding comprehensive verification index and generating a corresponding node maintenance instruction; wherein the comprehensive verification index is calculated by a self-verification score and a mutual verification score, the self-verification score is the result of self-verification by the edge computing node, the mutual verification score is the result of mutual verification by the edge computing node, and the node maintenance instruction is used to maintain the edge computing node.

[0122] It should be noted that the collaborative operation system of the edge computing nodes of this embodiment corresponds to the collaborative operation method of the edge computing nodes mentioned above. Therefore, the contents not specifically described in the collaborative operation system of the edge computing nodes of this embodiment may include but are not limited to functional definitions, working principles, and technical effects, etc., and may all refer to the records in the collaborative operation method of the edge computing nodes mentioned above, and this text will not go into details here.

[0123] In summary, the collaborative operation method and system of the edge computing nodes of this embodiment utilizes a series of processes such as parsing computing tasks and matching the first node group, comprehensively cognizing and calculating the comprehensive cognitive index of each node, screening the second node group based on the comprehensive cognitive index combined with the node allocation rule, executing tasks and continuously updating the second node group, and comprehensively verifying and generating maintenance instructions to achieve efficient and accurate collaborative operation of edge computing nodes. Before task assignment, the task characteristics are deeply analyzed, and nodes are matched in combination with functions and computing power to avoid blind assignment. Through self-cognition and mutual cognition, the nodes are comprehensively evaluated to select the second node group that better meets the task computing power requirements, thereby improving resource utilization. The node group is dynamically updated during task execution, inefficient nodes are eliminated in a timely manner, and suitable nodes are supplemented to ensure smooth and efficient computing tasks. Finally, maintenance instructions are generated based on the comprehensive verification index, changing passive to active maintenance, reducing resource waste and safety hazards, and improving the overall performance and reliability of collaborative operations of edge computing nodes.

[0124] In the embodiments provided in the present application, it should be understood that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code or any appropriate combination thereof. For hardware implementation, the processor can be implemented in one or more of the following units: application specific integrated circuit (ASIC), digital signal processor (DSP), digital signal processing device (DSPD), programmable logic device (PLD), field programmable gate array (FPGA), processor, controller, microcontroller, microprocessor, other electronic units designed to implement the functions described herein or their combination. For software implementation, part or all of the flow of the embodiment can be completed by instructing the relevant hardware through a computer program. When implemented, the above program can be stored in a computer-readable storage medium or transmitted as one or more instructions or codes on a computer-readable storage medium. Computer-readable storage media include computer storage media and communication media, wherein the communication medium includes any medium that is convenient for transmitting a computer program from one place to another. The storage medium can be any available medium that a computer can access. The computer-readable storage medium can include but is not limited to RAM, ROM, EEPROM, CD-ROM or other optical disk storage, disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of an instruction or data structure and can be accessed by a computer.

[0125] Finally, it should be noted that the above is only a preferred embodiment of the present application and is not intended to limit the present application. Although the present application 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 application shall be included in the protection scope of the present application.

Claims

1. A collaborative operation method of edge computing nodes, characterized in that: The method includes: S1: parsing the computing task and matching to obtain a corresponding first node group; wherein the first node group is a set of edge computing nodes that functionally correspond to the computing task; S2: Comprehensively recognize each edge computing node in the first node group to obtain a corresponding comprehensive recognition index; wherein the comprehensive recognition index is calculated by a self-recognition score and a mutual recognition score, the self-recognition score is the result of the edge computing node performing self-recognition, and the mutual recognition score is the result of the edge computing node performing mutual recognition; S3: Utilizing the comprehensive cognitive index of the first node group and combining it with a preset node allocation rule, a second node group is obtained by screening; wherein the node allocation rule is used to screen a set of edge computing nodes that meet the computing power of the computing task in the first node group, and the second node group is a set of edge computing nodes that correspond to the computing task in terms of computing power, and the number of nodes in the second node group is less than or equal to the number of nodes in the first node group; S4: Utilize the second node group to perform the computing task, continuously utilize the preset node self-elimination rule and node screening rule to update the second node group, and utilize the updated second node group to perform the computing task until the end after the update; wherein, the node self-elimination rule is used to screen edge computing nodes in the second node group that do not meet the computing efficiency when performing the computing task, and the node screening rule is used to screen edge computing nodes in the first node group that meet the computing power of the computing task; S5: Comprehensively verify each edge computing node in the second node group to obtain a corresponding comprehensive verification index and generate a corresponding node maintenance instruction; wherein, the comprehensive verification index is calculated by a self-verification score and a mutual verification score, the self-verification score is the result of self-verification of the edge computing node, the mutual verification score is the result of mutual verification of the edge computing node, and the node maintenance instruction is used to maintain the edge computing node.

2. The collaborative operation method of edge computing nodes according to claim 1, characterized in that: The analytical calculation tasks are specifically: Obtaining the computing task; The computing task is parsed to obtain a corresponding computing amount, a computing type, and a computing priority; wherein the computing amount, the computing type, and the computing priority are used to match the first node group.

3. The collaborative operation method of edge computing nodes according to claim 1, characterized in that: The calculation process of the comprehensive cognitive index is: Obtaining the first node group; Performing comprehensive cognition on the first node group, wherein the comprehensive cognition includes self-cognition and mutual cognition; The self-cognition is the cognition performed by the edge computing node based on its own real-time operation status, and the self-cognition score is calculated; The mutual recognition is the recognition performed by the edge computing node based on the real-time operation status of the edge computing nodes of the same type as itself, and the mutual recognition score is calculated; The comprehensive cognitive index is calculated using the self-cognition score and the mutual cognition score.

4. The collaborative operation method of edge computing nodes according to claim 3, characterized in that: The self-cognition is specifically: based on the real-time operation of the edge computing node and combined with the historical computing efficiency, the self-cognition score is calculated; The mutual recognition is specifically: based on the real-time operation status of the edge computing nodes of the same type, the mutual recognition score is calculated in combination with the historical computing efficiency of the edge computing nodes of the same type.

5. The collaborative operation method of edge computing nodes according to claim 1, characterized in that: The screening process of the second node group includes: Analyze the computing task to obtain a corresponding computing power allocation plan; The node allocation rule is called, and in combination with the computing power allocation plan, the edge computing nodes that meet the computing power of the computing task are screened in the first node group to obtain the second node group.

6. The collaborative operation method of edge computing nodes according to claim 1, characterized in that: The updating process of the second node group includes: Using the node self-elimination rule to filter edge computing nodes in the second node group that do not meet the computing efficiency when executing the computing task, and eliminating the calculated minimum value; The node self-elimination rule is specifically as follows: Among them, EN i is the edge computing node numbered i that is screened out and does not meet the computing efficiency in the second node group when executing the computing task, min is the minimum value operator, The integral of the efficiency P calculated by the edge computing node numbered i in the preset elimination period T, P i_τ is the computational efficiency threshold of the preset edge computing node numbered i, α i is the efficiency compensation value of the edge computing node numbered i, and the efficiency compensation value is used to characterize the compensation for the computing efficiency of the edge computing node during the elimination cycle.

7. The collaborative operation method of edge computing nodes according to claim 1, characterized in that: The updating process of the second node group further includes: When the edge computing nodes in the second node group that do not meet the computing efficiency when executing the computing task are screened out, the sub-computing tasks of the edge computing node are obtained, and the node screening rule is called to screen the edge computing nodes in the first node group that meet the computing power of the computing task, and the computing tasks of the eliminated edge computing nodes are assigned to the screened edge computing nodes.

8. The collaborative operation method of edge computing nodes according to claim 1, characterized in that: The calculation process of the comprehensive verification index is: Obtain each edge computing node in the second node group that executes the computing task; Performing comprehensive verification on each edge computing node in the second node group that executes the computing task, wherein the comprehensive verification includes self-verification and mutual verification; The self-verification is a verification performed by the edge computing node based on its own operation records, and the self-verification score is calculated; The mutual verification is a verification performed by the edge computing node based on the operation records of the edge computing nodes of the same type as itself, and a mutual verification score is calculated; The comprehensive verification index is calculated using the self-verification score and the mutual verification score.

9. The collaborative operation method of edge computing nodes according to claim 8, characterized in that: The self-verification is specifically: based on the operation record of the edge computing node, the self-verification score is calculated in combination with the self-cognition evaluation; The mutual verification is specifically: based on the operation records of the edge computing nodes of the same type, and combined with the historical computing efficiency of the edge computing nodes of the same type, the mutual verification score is calculated.

10. A collaborative operation system of edge computing nodes, the system being applicable to the collaborative operation method of edge computing nodes as described in any one of claims 1 to 9, characterized in that: The system includes: A first node group matching module, wherein the first node group matching module is configured to: parse the computing task and match to obtain a corresponding first node group; wherein the first node group is a collection of edge computing nodes that functionally correspond to the computing task; A comprehensive cognitive index calculation module, wherein the comprehensive cognitive index calculation module is configured to: comprehensively recognize each edge computing node in the first node group to obtain a corresponding comprehensive cognitive index; wherein the comprehensive cognitive index is calculated by a self-cognition score and a mutual cognition score, wherein the self-cognition score is a result of the edge computing node performing self-cognition, and the mutual cognition score is a result of the edge computing node performing mutual cognition; A second node group screening module, wherein the second node group screening module is configured to: use the comprehensive cognitive index of the first node group and a preset node allocation rule to screen and obtain a second node group; wherein the node allocation rule is used to screen a set of edge computing nodes that meet the computing power of the computing task in the first node group, the second node group is a set of edge computing nodes that correspond to the computing task in terms of computing power, and the number of nodes in the second node group is less than or equal to the number of nodes in the first node group; A second node group update module, wherein the second node group update module is configured to: use the second node group to perform the computing task, continuously use the preset node self-elimination rule and node screening rule to update the second node group, and use the updated second node group to perform the computing task until the end after the update; wherein the node self-elimination rule is used to screen edge computing nodes in the second node group that do not meet the computing efficiency when performing the computing task, and the node screening rule is used to screen edge computing nodes in the first node group that meet the computing power of the computing task; A comprehensive verification index calculation module, wherein the comprehensive verification index calculation module is configured to: comprehensively verify each edge computing node in the second node group to obtain a corresponding comprehensive verification index and generate a corresponding node maintenance instruction; wherein the comprehensive verification index is calculated by a self-verification score and a mutual verification score, the self-verification score is the result of self-verification by the edge computing node, the mutual verification score is the result of mutual verification by the edge computing node, and the node maintenance instruction is used to maintain the edge computing node.

Images