A computing power scheduling system and method of a computing power network

Abstract

The application discloses a computing power scheduling system and method of a computing power network, relates to the technical field of computing power distribution and scheduling, and comprises the following steps: acquiring all computing power nodes in the computing power network and constructing a computing power node interaction network model; acquiring a computing power task initiating node and computing power task attributes; determining a target attribute group cluster of the computing power task based on the computing power task attributes, and recording all computing power nodes in the target attribute group cluster as demand target nodes of the computing power task; determining an execution trust target node of the computing power task based on the computing power task initiating node; finding an intersection of the demand target nodes of the computing power task and the execution trust target node of the computing power task, and recording the intersection as a comprehensive target node; and determining an execution computing power node of the computing power task from the comprehensive target node in combination with balanced load. The application has the advantages that the interaction network model of the cooperative trust relationship and the attribute group cluster is constructed, multi-dimensional node dynamic screening and optimized matching are realized, and the efficiency and reliability of the computing power scheduling are significantly improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computing power allocation scheduling, in particular to a computing power scheduling system and method of a computing power network. BACKGROUND

[0002] With the continuous expansion of the scale of the computing power network, dynamic scheduling of computing power resources has become a key challenge to improve network efficiency. In the prior art, computing power scheduling often relies on simple load balancing or static node selection strategies, lacking comprehensive consideration of inter-node collaboration trust relationships, dynamic attribute clustering and network topology complexity. Traditional methods are difficult to effectively handle inter-node trust transmission, multi-hop collaboration path optimization and dynamic feedback adjustment after task execution, resulting in low resource utilization and insufficient task allocation reliability. In addition, node selection in the prior art often ignores the influence of node importance on global load balancing, and cannot adapt to the dynamic needs of large-scale heterogeneous computing power networks. Therefore, an intelligent computing power scheduling method that integrates trust evaluation, clustering analysis, load balancing and dynamic adjustment is urgently needed SUMMARY

[0003] To solve the above technical problems, a computing power scheduling system and method of a computing power network are provided, which solve the problem that traditional methods are difficult to effectively handle inter-node trust transmission, multi-hop collaboration path optimization and dynamic feedback adjustment after task execution, resulting in low resource utilization and insufficient task allocation reliability.

[0004] To achieve the above purposes, the technical scheme adopted by the present application is:

[0005] A computing power scheduling method of a computing power network, comprising:

[0006] All computing power nodes in the computing power network are obtained, and a computing power node interaction network model is constructed, the computing power node interaction network model comprising collaboration trust relationships and clustering relationships, an interaction relationship chain between computing power nodes is constructed based on the collaboration trust relationships, and attribute group clusters of the computing power nodes are constructed based on the clustering relationships;

[0007] A computing power task initiating node and a computing power task attribute are obtained;

[0008] The target attribute group cluster of the computing power task is determined based on the computing power task attribute, and all computing power nodes in the target attribute group cluster are recorded as demand target nodes of the computing power task;

[0009] Based on the computing power task initiating node, the execution trust target node of the computing power task is determined in combination with the interaction relationship chain between the computing power nodes;

[0010] The intersection of the demand target nodes of the computing power task and the execution trust target nodes of the computing power task is calculated, and is recorded as a comprehensive target node;

[0011] Determine the execution computing power node of the computing power task from the comprehensive target node in combination with balanced load;

[0012] Assign the computing power task to the execution computing power node for task execution, and adjust the interaction relationship chain among the computing power nodes in combination with the task execution result feedback.

[0013] As a preferred solution, the construction of the interaction relationship chain among the computing power nodes based on the cooperative trust relationship specifically includes:

[0014] Based on the physical distance among the computing power nodes, an initial cooperative trust weight is added between all pairs of computing power nodes;

[0015] If the computing power nodes have cooperated to complete the same computing power task, a cooperative relationship chain is formed among the computing power nodes, and the cooperative trust weight among the computing power nodes is determined based on the completion of the computing power task cooperatively completed by the computing power nodes and the historical cooperative trust weight;

[0016] If the computing power nodes do not form a cooperative relationship chain, but there are at most three cooperative relationship chains that can realize the connectivity of the computing power nodes, a multi-hop relationship chain is formed among the computing power nodes, and the cooperative trust weight of the computing power nodes at both ends of the multi-hop relationship chain is calculated based on the cooperative trust weight among the computing power nodes constituting the multi-hop relationship chain;

[0017] If neither a cooperative relationship chain nor a multi-hop relationship chain is formed among the computing power nodes, the cooperative trust weight among the computing power nodes is the initial cooperative trust weight;

[0018] Summarize all the cooperative relationship chains and multi-hop relationship chains to obtain the interaction relationship chain among the computing power nodes.

[0019] As a further preferred solution: determining the execution computing power node of the computing power task from the comprehensive target node in combination with balanced load specifically includes:

[0020] Determine the computing power load demand of the computing power task and the computing power load capacity of each computing power node in the comprehensive target node;

[0021] Set N=1, and determine whether the sum of the computing power load capacities of N computing power nodes in the comprehensive target node is greater than the computing power load demand of the computing power task, if yes, combine the N computing power nodes into a node topology structure, if no, increase the value of N by 1 and return to continue the determination, if the value of N exceeds the number of computing power nodes included in the comprehensive target node, output an alarm signal to the background terminal;

[0022] The ratio of the computing power load demand of the computing power task to the sum of the computing power load capacities of all computing power nodes in the node topology structure is taken as the task load rate of the node topology structure;

[0023] Screen out the node topology with the minimum task load rate, and all the computing power nodes therein are used as the execution computing power nodes of the computing power task.

[0024] Further, in combination with the above-mentioned computing power network computing power scheduling method, a computing power network computing power scheduling system is further proposed, comprising:

[0025] A node management module is configured to acquire all computing power nodes in the computing power network and construct a computing power node interaction network model comprising cooperative trust relationships and clustering relationships.

[0026] The node management module specifically comprises:

[0027] A weight initialization unit is configured to attach an initial cooperative trust weight to the computing power nodes based on physical distance;

[0028] A cooperative relationship construction unit is configured to generate a cooperative relationship chain and a dynamic weight based on historical cooperative task completion;

[0029] A multi-hop relationship calculation unit is configured to calculate a multi-hop trust weight by multiplying the weights of up to three relationship chains for nodes without direct cooperation, and take the maximum value as the maximum weight;

[0030] A task analysis module is configured to acquire a computing power task initiating node and computing power task attributes, and determine a target attribute group cluster and its demand target node based on the attributes;

[0031] A trust evaluation module is configured to calculate the cooperative trust weight of the computing power task initiating node and other nodes based on the interaction relationship chain, and screen the execution trust target node;

[0032] A node matching module is configured to calculate the intersection of the demand target node and the execution trust target node to generate a comprehensive target node, and select the execution computing power node from the comprehensive target node in combination with a load balancing algorithm;

[0033] The node matching module specifically comprises:

[0034] A load evaluation unit is configured to monitor the computing power load capacity of the comprehensive target node in real time;

[0035] A topology generation unit is configured to search for the minimum node combination satisfying the task demand by an iterative algorithm, and calculate the task load rate of each combination;

[0036] An optimal decision unit is configured to select the node topology with the lowest task load rate as the execution computing power node;

[0037] A dynamic adjustment module is configured to update the cooperative trust relationship and the interaction relationship chain based on the task execution result feedback

[0038] Compared with the prior art, the present application has the following advantages:

[0039] The application realizes multi-dimensional node dynamic screening and optimized matching by constructing an interactive network model of cooperative trust relationship and attribute group clustering, and significantly improves the efficiency and reliability of computing power scheduling. Through trust cooperation and multi-hop path optimization, based on cooperation history and multi-hop trust weight calculation, the credibility of task allocation is enhanced, the participation of inefficient nodes is reduced, load balancing and resource efficient utilization are achieved, the minimum load rate node combination is searched through iteration, combined with the node importance correction coefficient, local overload is avoided and global resource utilization is improved, dynamic feedback and network adaptability: according to the task execution result, the trust relationship chain is adjusted in real time, so that the network has self-optimization ability, complex scene robustness: the attribute group clustering is introduced to quickly lock the target node set, and the search complexity in large-scale network is reduced. The problems of trust transmission, computing power resource cooperation and dynamic load balancing in computing power network are effectively solved, which is suitable for cloud computing, edge computing and other scenes, and has wide application value. BRIEF DESCRIPTION OF DRAWINGS

[0040] Figure 1 The flow chart of the computing power scheduling method of the computing power network for example one is shown in the figure;

[0041] Figure 2 The flow chart of the method for constructing the interactive relationship chain between the computing power nodes according to the cooperative trust relationship for example one is shown in the figure;

[0042] Figure 3 The flow chart of the method for determining the execution computing power node of the computing power task from the comprehensive target node for example one is shown in the figure. DETAILED DESCRIPTION

[0043] The following description is used to disclose the application so that those skilled in the art can implement the application. The preferred embodiments in the following description are only as examples, and other obvious modifications can be thought of by those skilled in the art.

[0044] Referring to Figure 1 The computing power scheduling method of a computing power network, comprising:

[0045] All computing power nodes in the computing power network are acquired, and a computing power node interactive network model is constructed, the computing power node interactive network model comprising cooperative trust relationship and clustering relationship, an interactive relationship chain between the computing power nodes is constructed based on the cooperative trust relationship, and an attribute group clustering of the computing power nodes is constructed based on the clustering relationship, for example, GPU / TPU nodes are divided into a computing-intensive community suitable for AI training tasks, and large-capacity hard disk nodes are divided into a storage-intensive community suitable for data analysis tasks;

[0046] The computing power task initiation node and the computing power task attribute are acquired;

[0047] Determine the target attribute cluster of the computing power task based on the computing power task attribute, and record all computing power nodes in the target attribute cluster as the demand target nodes of the computing power task, for example, matrix operation task → computing-intensive community, data preprocessing task → storage-intensive community;

[0048] Determine the execution trust target node of the computing power task based on the computing power task initiation node and the interaction relationship chain between the computing power nodes;

[0049] Find the intersection of the demand target nodes of the computing power task and the execution trust target nodes of the computing power task, and record it as the comprehensive target node;

[0050] Determine the execution computing power node of the computing power task from the comprehensive target node in combination with the balanced load;

[0051] Assign the computing power task to the execution computing power node for task execution, and adjust the interaction relationship chain between the computing power nodes in combination with the task execution result feedback.

[0052] Referring to Figure 2 In the embodiment, the construction of the interaction relationship chain between the computing power nodes based on the cooperative trust relationship specifically includes:

[0053] Based on the physical distance between the computing power nodes, an initial cooperative trust weight is added between each two computing power nodes;

[0054] Specifically, the initial cooperative trust weight is determined in the following manner:

[0055] Determine the physical distance between the computing power nodes, and set the maximum value of the initial cooperative trust weight, and specifically, the maximum value of the initial cooperative trust weight is set to 0.8, then the initial cooperative trust weight between each two computing power nodes is determined in a linear mapping or nonlinear mapping manner, wherein the calculation formula of the linear mapping is: wherein, is the initial cooperative trust weight between the computing power node i and the computing power node j, is the physical distance between the computing power node i and the computing power node j, is the maximum value of the physical distance between the computing power nodes, is the minimum value of the physical distance between the computing power nodes;

[0056] In the initial stage of system operation, due to the lack of cooperative data between nodes, the initial cooperative trust weight between each two computing power nodes is determined by taking the physical distance which is the most critical factor of communication delay as a consideration factor;

[0057] If the same computing power task is completed by cooperation between computing power nodes, a cooperation relationship chain is formed between the computing power nodes, and the cooperation trust weight between the computing power nodes is determined based on the completion of the computing power task completed by cooperation between the computing power nodes and the historical cooperation trust weight;

[0058] The cooperation trust weight between the computing power nodes is calculated by 0.8*historical cooperation trust weight+0.2*completion score of the computing power task completed this time, dynamically configured, and the completion score of the computing power task completed is calculated by , wherein, is the task demand data volume, M is the actual output data volume, is the task demand data precision, J is the actual output data precision, d is the average communication delay of this cooperation, is the system maximum allowed delay threshold, which is usually set to 500 ms;

[0059] If the cooperation relationship chain is not formed between the computing power nodes, but there are at most three cooperation relationship chains that can realize the connection of the computing power nodes, a multi-hop relationship chain is formed between the computing power nodes, and the cooperation trust weight of the computing power nodes at both ends of the multi-hop relationship chain is calculated based on the cooperation trust weight between the computing power nodes constituting the multi-hop relationship chain;

[0060] If there is only one multi-hop relationship chain between two computing power nodes, the cooperation trust weight between the two computing power nodes is the product of the cooperation trust weights of the computing power nodes constituting the multi-hop relationship chain;

[0061] If there are multiple multi-hop relationship chains between two computing power nodes, the cooperation trust weight between the two computing power nodes is the maximum value in the product of the cooperation trust weights of the computing power nodes of all multi-hop relationship chains;

[0062] For example, if there is only one multi-hop relationship chain A→B, B→D between A and D nodes, the cooperation trust weight between A node and D node is: , is the cooperation trust weight between AB nodes, is the cooperation trust weight between BD nodes;

[0063] If there are two multi-hop relationship chains A→B, B→D and A→C, C→D between A and D nodes, the cooperation trust weight between A node and D node is: and the maximum value of is the cooperation trust weight between AC nodes, is the cooperation trust weight between CD nodes.

[0064] If the cooperation relationship chain and the multi-hop relationship chain are not formed between the computing power nodes, the cooperation trust weight between the computing power nodes is the initial cooperation trust weight.

[0065] Summarize all collaborative relationship chains and multi-hop relationship chains to obtain the interactive relationship chain between computing power nodes.

[0066] By collecting computing task execution results (such as task success rate and response latency) in real time, a dynamic feedback mechanism is established to adaptively correct the collaborative trust relationship chain. Specifically, if a node performs well in task execution (such as high computing efficiency or low error rate), its direct collaborative trust weight with the initiating node will be positively enhanced according to preset rules, and the overall trust weight of its multi-hop path will be simultaneously improved; conversely, if the node has problems with poor task completion quality, its trust weight will be automatically reduced, and the trust decay of the associated multi-hop path will be triggered. This closed-loop feedback mechanism enables the network to quickly identify inefficient or failed nodes and gradually exclude them from the preferred path through weight decay, while strengthening the collaborative priority of high-reliability nodes.

[0067] Based on the interactive relationship chain between computing power nodes constructed above, by screening out computing power nodes whose collaborative trust weight with the computing power task initiating node is greater than the trust threshold, as the execution trust target node of the computing power task, it can be ensured that the execution node of the computing power task is a high-quality collaborative node of the task initiating node, and that the computing power task can be collaboratively executed by high-reliability nodes.

[0068] Further, refer to Figure 3 As shown, in this embodiment: in combination with load balancing, determining the execution computing node of the computing task from the comprehensive target node specifically includes:

[0069] Determine the computing load requirements of the computing task and the computing load capacity of each computing node in the comprehensive target node;

[0070] Set N=1 to determine whether the sum of the computing load capacity of N computing nodes of the comprehensive target node is greater than the computing load requirement of the computing task. If so, combine the N computing nodes into a node topology structure. If not, add 1 to the value of N and return to continue judging. If the value of N exceeds the number of computing nodes included in the comprehensive target node, output an alarm signal to the background terminal;

[0071] The ratio of the computing load requirement of the computing task to the sum of the computing load capacity of all computing nodes in the node topology is used as the task load rate of the node topology;

[0072] Filter out the node topology with the smallest task load rate, and use all computing nodes in it as the execution computing nodes of the computing task.

[0073] By iteratively searching for the node combination with the minimum load rate, local overload can be avoided and global resource utilization can be improved, thereby improving the global rationality of the computing power scheduling of the computing power network. Example

[0074] Furthermore, based on the first embodiment, this embodiment introduces a node importance correction coefficient when determining the execution computing node of the computing task, as follows:

[0075] Based on the collaborative relationship chain of each node, combined with the PageRank algorithm, the node importance of each computing power node in the computing power network is determined;

[0076] The calculation formula for the node importance of the computing power node is: ,in, is the node importance of the i-th computing power node, is the set of nodes that have a collaborative relationship chain with the i-th computing power node, for The node importance of the vth computing power node, For The number of nodes in the collaboration chain with the vth computing power node, M is the total number of nodes in the network, e is the damping coefficient, usually 0.85, and v is The vth computing power node in ;

[0077] After introducing the node importance correction coefficient, the calculation formula for the task load rate of the node topology structure is:

[0078] ;

[0079] in, is the task load rate of the i-th node topology structure with the node importance correction coefficient introduced, X is the computing power load requirement of the computing power task, is the computing load capacity of the kth computing node in the i-th node topology, is the topological structure of the i-th node, K is the union of all node topological structures, is the node importance of the kth computing power node in the i-th node topology, k is the kth computing power node in the i-th node topology, u is the uth computing power node in K, is the node importance of the u-th computing power node in K, α and β are weight coefficients, and the weight coefficient is determined based on the frequency of task initiation in the network. If the frequency of task initiation in the network is high, it is necessary to ensure that the high-importance nodes are idle. In this case, the β value is increased, otherwise, the α value is increased.

[0080] A node importance correction mechanism is used to realize precise allocation of computing power resources and global load optimization. Specifically, the system first uses a progressive search strategy in the comprehensive target node set according to the load demand of the computing power task: starting from a single node (N = 1), it expands to a multi-node combination layer by layer, and selects the minimum node topology structure that meets the task computing power demand. The task load rate of each candidate node combination is calculated by a formula, which not only contains the ratio of task demand to the sum of node computing power, but also introduces a node importance correction coefficient based on the PageRank algorithm. This coefficient quantifies the global influence of the node in the network by analyzing its centrality in the cooperation relationship chain, historical task contribution, and multi-hop path participation frequency, so that the system preferentially selects high-importance node light-load combinations.

[0081] Embodiment three:

[0082] This embodiment combines embodiments one and two to propose a computing power scheduling system for a computing power network, which includes:

[0083] A node management module is used to obtain all computing power nodes in the computing power network and construct a computing power node interaction network model containing cooperation trust relationships and clustering relationships;

[0084] A task analysis module is used to obtain the computing power task initiation node and computing power task attributes, and determine the target attribute group cluster and its demand target node according to the attributes;

[0085] A trust evaluation module is used to calculate the cooperation trust weight of the computing power task initiation node and other nodes based on the interaction relationship chain, and select the execution trust target node;

[0086] A node matching module is used to calculate the intersection of the demand target node and the execution trust target node to generate a comprehensive target node, and select the execution computing power node from it in combination with a load balancing algorithm;

[0087] A dynamic adjustment module is used to update the cooperation trust relationship and the interaction relationship chain according to the task execution result feedback;

[0088] The node management module includes:

[0089] A weight initialization unit is used to attach an initial cooperation trust weight to the computing power node based on the physical distance;

[0090] A cooperation relationship construction unit is used to generate a cooperation relationship chain and dynamic weight according to the historical cooperation task completion;

[0091] The multi-hop relationship calculation unit calculates the multi-hop trust weight by multiplying the weights of at most three relationship chains for nodes without direct cooperation, and takes the maximum value as the maximum weight, wherein the multi-hop relationship calculation unit configures an algorithm as follows: when there is a single multi-hop path, the weight between nodes is calculated by multiplying all the cooperative trust weights in the path; when there are multiple multi-hop paths, the maximum value is reserved by comparing the multiplication results of all paths through a parallel calculation unit.

[0092] The node matching module comprises:

[0093] The load evaluation unit monitors the computing power load capacity of the comprehensive target node in real time.

[0094] The topology generation unit searches for the minimum node combination satisfying the task demand through an iterative algorithm, and calculates the task load rate of each combination.

[0095] The optimal decision unit selects the node topology with the lowest task load rate as the execution computing power node.

[0096] In summary, the advantages of the present application are as follows: by constructing an interactive network model of cooperative trust relationship and attribute group clustering, multi-dimensional dynamic node screening and optimized matching are realized, and the efficiency and reliability of computing power scheduling are significantly improved. Through trust cooperation and multi-hop path optimization, based on cooperative history and multi-hop trust weight calculation, the credibility of task allocation is enhanced, the participation of inefficient nodes is reduced, load balancing and efficient resource utilization are achieved, the minimum load rate node combination is searched through iterative search, combined with the node importance correction coefficient, local overload is avoided and global resource utilization is improved, dynamic feedback and network adaptability: according to the task execution result, the trust relationship chain is adjusted in real time, so that the network has self-optimization ability, complex scene robustness: the attribute group clustering is introduced to quickly lock the target node set, and the search complexity in large-scale network is reduced. The problems of trust transmission, computing power resource cooperation and dynamic load balancing in computing power network are effectively solved, and the present application is suitable for cloud computing, edge computing and other scenes, and has wide application value.

[0097] The basic principles, main features and advantages of the present application are shown and described above. It should be understood by those skilled in the art that the present application is not limited by the above examples, and the above examples and descriptions in the specification are only the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection claimed by the present application is defined by the appended claims and their equivalents.

Claims

1. A computing power scheduling method of a computing power network, characterized in that, The application comprises the following steps: all computing power nodes in a computing power network are acquired, and a computing power node interaction network model is constructed, the computing power node interaction network model comprising a cooperation trust relationship and a clustering relationship, an interaction relationship chain between the computing power nodes is constructed based on the cooperation trust relationship, and an attribute group cluster of the computing power nodes is constructed based on the clustering relationship; a computing power task initiating node and a computing power task attribute are acquired; a target attribute group cluster of the computing power task is determined based on the computing power task attribute, and all computing power nodes in the target attribute group cluster are recorded as demand target nodes of the computing power task; an execution trust target node of the computing power task is determined based on the computing power task initiating node and in combination with the interaction relationship chain between the computing power nodes; an intersection of the demand target nodes of the computing power task and the execution trust target node of the computing power task is calculated, and the intersection is recorded as a comprehensive target node; an execution computing power node of the computing power task is determined from the comprehensive target node in combination with balanced load; the computing power task is distributed to the execution computing power node for task execution, and the interaction relationship chain between the computing power nodes is adjusted in combination with a task execution result feedback; the construction of the interaction relationship chain between the computing power nodes based on the cooperation trust relationship comprises the following steps: initial cooperation trust weights are added between all pairs of the computing power nodes based on physical distances between the computing power nodes; if the computing power nodes have completed a same computing power task, a cooperation relationship chain is formed between the computing power nodes, and cooperation trust weights between the computing power nodes are determined based on completion of the computing power task completed by the computing power nodes and historical cooperation trust weights; if the computing power nodes do not form a cooperation relationship chain but at most three cooperation relationship chains exist to realize connection of the computing power nodes, a multi-hop relationship chain is formed between the computing power nodes, and cooperation trust weights of the computing power nodes at two ends of the multi-hop relationship chain are calculated based on cooperation trust weights between the computing power nodes constituting the multi-hop relationship chain; if the computing power nodes neither form a cooperation relationship chain nor form a multi-hop relationship chain, the cooperation trust weights between the computing power nodes are initial cooperation trust weights; all cooperation relationship chains and multi-hop relationship chains are summarized to obtain the interaction relationship chain between the computing power nodes.

2. The computing power scheduling method of the computing power network according to claim 1, characterized in that, The cooperation trust weights of the computing power nodes at two ends of the multi-hop relationship chain are determined in the following manner: if there is only one multi-hop relationship chain between two computing power nodes, the cooperation trust weights between the two computing power nodes are the product of the cooperation trust weights of the computing power nodes constituting the multi-hop relationship chain; if there are multiple multi-hop relationship chains between two computing power nodes, the cooperation trust weights between the two computing power nodes are the maximum value in the product of the cooperation trust weights of the computing power nodes of all the multi-hop relationship chains.

3. The computing power scheduling method of the computing power network according to claim 1, wherein, The determination of the execution trust target node of the computing power task based on the computing power task initiating node and in combination with the interaction relationship chain between the computing power nodes comprises the following step: computing power nodes having cooperation trust weights greater than a trust threshold with the computing power task initiating node are screened out as the execution trust target nodes of the computing power task.

4. The computing power scheduling method of the computing power network according to claim 1, wherein, The determination of the execution computing power node of the computing power task from the comprehensive target node in combination with balanced load comprises the following steps: the computing power load demand of the computing power task and the computing power load capacity of each computing power node in the comprehensive target node are determined; Setting N=1, judging whether the sum of the computing power load capacity of N computing power nodes is greater than the computing power load demand of the computing power task, if yes, combining the N computing power nodes into a node topology structure, if no, increasing the value of N by 1 and returning to continue judging, if the value of N exceeds the number of computing power nodes included in the comprehensive target node, outputting an alarm signal to a background terminal; Taking the ratio of the computing power load demand of the computing power task to the sum of the computing power load capacity of all computing power nodes in the node topology structure as the task load rate of the node topology structure; Screening out the node topology structure with the minimum task load rate and taking all the computing power nodes therein as the execution computing power nodes of the computing power task.

5. The computing power scheduling method of the computing power network according to claim 4, characterized in that, Introducing a node importance correction coefficient when determining the execution computing power nodes of the computing power task, specifically including: Determining the node importance of each computing power node in the computing power network based on the cooperation relationship chain of each node and combining the PageRank algorithm; Then, the task load rate calculation formula of the node topology structure after introducing the node importance correction coefficient is: ； wherein, is the task load rate of the i-th node topology for introducing the node importance correction coefficient, X is the computing power load demand of the computing power task, is the computing power load capacity of the k-th computing power node in the i-th node topology, is the i-th node topology, K is the union of all node topologies, is the node importance of the k-th computing power node in the i-th node topology, k is the k-th computing power node in the i-th node topology, and u is the u-th computing power node in K, is the node importance of the u-th computing power node in K, and α and β are weight coefficients.

6. A computing power scheduling system of a computing power network, characterized in that, The computing power scheduling method for realizing the computing power network of any one of claims 1-5, comprising: A node management module for acquiring all computing power nodes in the computing power network and constructing a computing power node interaction network model containing cooperation trust relationship and clustering relationship; A task analysis module for acquiring the computing power task initiation node and the computing power task attribute and determining the target attribute group cluster and its demand target node according to the attribute; A trust evaluation module for calculating the cooperation trust weight of the computing power task initiation node and other nodes based on the interaction relationship chain and screening the execution trust target node; A node matching module for calculating the intersection of the demand target node and the execution trust target node to generate a comprehensive target node and selecting the execution computing power node therefrom in combination with a load balancing algorithm; A dynamic adjustment module for updating the cooperation trust relationship and the interaction relationship chain according to the task execution result feedback.

7. The computing power scheduling system of a computing power network according to claim 6, wherein, The node management module comprises: A weight initialization unit for attaching an initial cooperation trust weight to the computing power node based on the physical distance; A cooperation relationship construction unit for generating a cooperation relationship chain and a dynamic weight according to the historical cooperation task completion situation; A multi-hop relationship calculation unit for calculating the multi-hop trust weight through the weight product of up to three relationship chains for nodes without direct cooperation and taking the maximum value as the maximum weight.

8. The computing power scheduling system of a computing power network according to claim 7, wherein, The multi-hop relationship calculation unit is configured to: When there is a single multi-hop path, the weight between nodes is calculated by multiplying all the cooperation trust weights in the path; When there are multiple multi-hop paths, the maximum value is reserved by comparing the multiplication results of all paths through a parallel calculation unit. 9.The computing power scheduling system of a computing power network according to claim 6, wherein, The node matching module comprises: A load evaluation unit for monitoring the computing power load capacity of the comprehensive target node in real time; A topology generation unit for searching the smallest node combination satisfying the task demand through an iterative algorithm and calculating the task load rate of each combination; An optimal decision unit for selecting the node topology with the lowest task load rate as the execution computing power node.

Images