Metaverse resource scheduling method, apparatus and device, and medium and product

Abstract

The present application discloses a metaverse resource scheduling method, apparatus and device, and a medium and a product. The method comprises: using a pre-trained graph neural network model to perform resource demand forecasting, and formulating a first resource allocation strategy on the basis of a generated resource demand forecasting value; and using the graph neural network model to analyze resource usage data, dynamically adjusting the first resource allocation strategy on the basis of an analysis result, and performing update to obtain a second resource allocation strategy. The present application relates to making use of self-learning and self-adaptive capabilities of a graph neural network to perform analysis and adjustment on the basis of dynamic resource usage data detected in real time and to dynamically adjust a resource allocation strategy, thereby ensuring that stable performance of a system can still be maintained even under sudden increase or change in resource demand.

Description

Metaverse resource scheduling methods, devices, equipment, media and products

[0001] Cross-reference to related applications

[0002] This application is based on and claims priority to Chinese Patent Application No. 202411571650.1, filed on November 5, 2024, the entire contents of which are incorporated herein by reference. Technical Field

[0003] This application relates to the field of electronic digital data processing technology, and in particular to methods, apparatus, equipment, media and products for scheduling metaverse resources. Background Technology

[0004] With the development of IT cloud-native technologies, the complexity of application systems is increasing exponentially, requiring continuous investment of various operations, maintenance, and development personnel to ensure business and platform stability. However, the decentralized deployment of personnel necessitates new collaborative models to efficiently complete tasks. The metaverse, as a virtual digital space, is increasingly becoming an important part of people's lives and work.

[0005] In the metaverse environment, there exist numerous nodes and edges. These nodes include users, virtual items, and servers, while edges include user interactions, item transactions, and server connections. Traditional resource scheduling methods are typically based on static rules or predefined optimization strategies, such as round-robin, priority, or simple load balancing algorithms. These strategies cannot adapt to the dynamically changing relationships between nodes and edges in the metaverse. In the metaverse, the relationships between users, virtual items, and service nodes may change at any time, making it difficult for traditional methods to dynamically adjust resource allocation strategies to cope with this complexity and variability. Summary of the Invention

[0006] The main purpose of this application is to provide a method, apparatus, equipment, medium and product for scheduling metaverse resources.

[0007] To achieve the above objectives, this application provides a metaverse resource scheduling method, which includes the following operations:

[0008] Resource demand is predicted using a pre-trained graph neural network model, and a first resource allocation strategy is formulated based on the generated resource demand predictions.

[0009] The graph neural network model is used to analyze resource usage data, and the first resource allocation strategy is dynamically adjusted based on the analysis results to obtain the second resource allocation strategy.

[0010] In one embodiment, after analyzing resource usage data using the graph neural network model, dynamically adjusting the first resource allocation strategy based on the analysis results, and updating to obtain the second resource allocation strategy, the process further includes:

[0011] Based on the pre-calculated priority index, the second resource allocation strategy is optimized through the graph neural network model to generate a resource allocation scheme.

[0012] In one embodiment, after the step of optimizing the second resource allocation strategy using the graph neural network model based on pre-calculated priority indicators to generate a resource allocation scheme, the method further includes:

[0013] Resource scheduling operations are performed according to the resource allocation scheme, and resource configuration is dynamically adjusted.

[0014] In one embodiment, the step of constructing graph structure data in the metaverse environment, and the step of optimizing the second resource allocation strategy based on pre-calculated priority indicators using the graph neural network model to generate a resource allocation scheme, includes:

[0015] Calculate the importance parameters and resource utilization of nodes and edges in the metaverse environment;

[0016] Calculate the priority indicators for nodes and edges based on user demand parameters, importance parameters, and resource utilization rates.

[0017] The graph neural network model optimizes the second resource allocation strategy based on the priority indices of nodes and edges to generate a resource allocation scheme.

[0018] In one embodiment, the operation of using a pre-trained graph neural network model to predict resource demand and formulating a first resource allocation strategy based on the generated resource demand prediction values ​​includes:

[0019] Resource demand predictions for nodes and edges in pre-constructed graph structure data are performed using a pre-trained graph neural network model, and the predicted resource demand values ​​for nodes and edges are obtained respectively.

[0020] Based on the predicted resource requirements of nodes and edges, calculate the resource allocation for nodes and edges respectively.

[0021] In one embodiment, before the operation of using a pre-trained graph neural network model to predict the resource demand of nodes and edges in a pre-constructed graph structure data, and obtaining the predicted resource demand values ​​of nodes and edges respectively, the method further includes:

[0022] A graph dataset is generated by extracting feature vectors of nodes and edges from pre-constructed graph structure data.

[0023] The pre-built graph neural network model is trained using the graph dataset.

[0024] In one embodiment, the operation of training a pre-built graph neural network model using the graph dataset includes:

[0025] The graph dataset is input into a pre-built graph neural network model;

[0026] The backpropagation algorithm is used to calculate the gradient of the loss function with respect to the parameters of the graph neural network model;

[0027] The parameters of the graph neural network model are updated based on the gradient using the gradient descent algorithm.

[0028] The loss function is minimized through iterative training until it converges to a preset threshold, resulting in an optimized graph neural network model.

[0029] In one embodiment, the operation of analyzing resource usage data using the graph neural network model, dynamically adjusting the first resource allocation strategy based on the analysis results, and updating to obtain the second resource allocation strategy includes:

[0030] The resource usage data of nodes and edges is monitored by a pre-deployed resource monitoring module, and the resource usage data is dynamically analyzed to calculate the resource utilization rate.

[0031] The graph neural network model is used to analyze the input resource usage data and calculate the resource usage trends of nodes and edges.

[0032] The bottleneck nodes and bottleneck edges are identified using the graph neural network model, and the resource stress of the bottleneck nodes and bottleneck edges is calculated based on the resource utilization rate.

[0033] Based on the resource usage trend and resource scarcity, the first resource allocation strategy is dynamically adjusted using the graph neural network model to update and obtain the second resource allocation strategy.

[0034] Furthermore, to achieve the above objectives, this application also provides a metaverse resource scheduling device, the metaverse resource scheduling device comprising:

[0035] The first strategy module is used to predict resource demand using a pre-trained graph neural network model and formulate a first resource allocation strategy based on the generated resource demand prediction values.

[0036] The second strategy module is used to analyze resource usage data using the graph neural network model, dynamically adjust the first resource allocation strategy based on the analysis results, and update the second resource allocation strategy.

[0037] In addition, to achieve the above objectives, this application also provides a terminal device, which includes a memory, a processor, and a metaverse resource scheduler stored in the memory and executable on the processor. When the metaverse resource scheduler is executed by the processor, it implements the operation of the metaverse resource scheduling method as described above.

[0038] In addition, to achieve the above objectives, this application also provides a computer-readable storage medium storing a metaverse resource scheduler, which, when executed by a processor, implements the operation of the metaverse resource scheduling method as described above.

[0039] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the operation of the metaverse resource scheduling method described above. Attached Figure Description

[0040] Figure 1 is a flowchart illustrating a first exemplary embodiment of the metaverse resource scheduling method of this application;

[0041] Figure 2 is a flowchart illustrating a second exemplary embodiment of the metaverse resource scheduling method of this application.

[0042] Figure 3 is a flowchart of the third exemplary embodiment of the metaverse resource scheduling method of this application;

[0043] Figure 4 is a flowchart illustrating the fourth exemplary embodiment of the metaverse resource scheduling method of this application;

[0044] Figure 5 is a schematic diagram of the module structure of the metaverse resource scheduling device according to an embodiment of this application;

[0045] Figure 6 is a schematic diagram of the hardware operating environment involved in the metaverse resource scheduling method in this application embodiment.

[0046] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0047] It should be understood that the embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0048] This application takes into account that traditional resource scheduling methods typically use rule-based or simple optimization algorithms to manage the allocation of computing, storage, and network resources. These methods are often suitable for traditional computing environments, such as data centers or traditional network services, but they have significant limitations in applicability when facing complex virtual environments like the metaverse. Furthermore, traditional resource scheduling methods are often based on static rules or predefined optimization strategies. These optimization strategies may be based on round-robin, priority, or simple load balancing algorithms, but they cannot adapt to the complex dynamic relationships within the metaverse.

[0049] Based on this, the first embodiment of this application proposes a solution that utilizes a pre-trained graph neural network model to predict resource demand and formulates a first resource allocation strategy based on the generated resource demand predictions. The graph neural network model is then used to analyze resource usage data, and the first resource allocation strategy is dynamically adjusted based on the analysis results to update and obtain a second resource allocation strategy. This method leverages the self-learning and adaptive capabilities of graph neural networks, analyzing and adjusting based on real-time monitored dynamic resource usage data. This allows for timely responses to changes in resource demand, dynamically adjusting the resource allocation strategy, and ensuring system stability even during sudden increases or changes in resource demand. This real-time dynamic scheduling capability significantly improves the system's response speed and flexibility.

[0050] The following is a detailed description of the first exemplary embodiment of the primitive universe resource scheduling method:

[0051] Referring to Figure 1, which is a flowchart illustrating a first exemplary embodiment of the metaverse resource scheduling method of this application, the metaverse resource scheduling method includes operations S10 to S20:

[0052] Operation S10 uses a pre-trained graph neural network model to predict resource demand and formulates a first resource allocation strategy based on the generated resource demand prediction values.

[0053] In this embodiment, historical resource usage data is first collected, including but not limited to CPU utilization, memory usage, network bandwidth, storage I / O, and the relationship between these resource usages and business load and time (such as weekdays and weekends, day and night). Then, a dataset is constructed using the historical resource usage data, including cleaning and standardizing the data, and handling missing values, outliers, etc.

[0054] Furthermore, a graph neural network model is constructed and trained. The dataset obtained based on the resource demand prediction is input into the trained graph neural network model. The graph neural network model predicts the resource demand in the future and outputs the prediction result, that is, the resource demand prediction value is obtained.

[0055] Based on the projected resource demand, resources can be rationally allocated to meet business needs while maintaining efficient resource utilization. This allocation can be tailored to consider factors such as business requirements, resource constraints, resource priority, scalability, and cost-effectiveness. A preliminary resource allocation plan, or first resource allocation strategy, is then developed.

[0056] In operation S20, the graph neural network model is used to analyze the resource usage data, and the first resource allocation strategy is dynamically adjusted based on the analysis results to update and obtain the second resource allocation strategy.

[0057] In this embodiment, resource usage data is first monitored and collected, including but not limited to the real-time consumption of computing resources, storage resources, and network resources. A graph neural network model is then used to perform in-depth analysis of the resource usage data, identifying abnormal patterns, trend changes, and potential performance bottlenecks. Based on the analysis results, shortcomings in the first resource allocation strategy are identified, such as uneven resource allocation, excessive reservation, or insufficient resources. Based on this, the first resource allocation strategy is dynamically adjusted, for example, by adjusting resource allocation ratios, priorities, and schedules, to optimize resource utilization efficiency. The adjusted resource allocation plan is then integrated into a second resource allocation strategy.

[0058] Furthermore, resource scheduling is executed according to the second resource allocation strategy, and resource usage and business performance are continuously monitored. Feedback data is collected to further optimize the graph neural network model and resource allocation strategy. Based on the feedback data, the graph neural network model is updated and optimized to improve prediction accuracy. Based on the continuously optimized graph neural network model, the resource allocation strategy is iteratively updated to ensure optimal resource utilization.

[0059] This application also considers that traditional resource scheduling methods are inefficient when processing large-scale graph-structured data. The nodes and edges in the metaverse environment constitute a vast and complex graph structure, and traditional methods may require significant computation time and resources to process these complex relationships and data volumes. For example, traditional load balancing algorithms may simply allocate resources based on the current load, without being able to utilize the potential information in the graph structure for more precise resource optimization.

[0060] Furthermore, traditional methods suffer from bottlenecks in achieving multi-dimensional resource scheduling. Resource scheduling in the metaverse environment involves not only the allocation of computing resources but also the coordinated management of storage and network resources. Traditional methods often focus on scheduling a single resource type, failing to effectively balance and optimize across multiple resource dimensions. When a particular resource type becomes a bottleneck, the overall system performance and user experience are negatively impacted.

[0061] Based on this, the second embodiment of this application proposes a solution: generating a graph dataset by extracting feature vectors of nodes and edges from pre-constructed graph structure data; training a pre-constructed graph neural network model using the graph dataset; predicting resource requirements of nodes and edges in the pre-constructed graph structure data using the pre-trained graph neural network model, and obtaining predicted resource requirements of nodes and edges respectively; calculating the resource allocation amount of nodes and edges respectively based on the predicted resource requirements of nodes and edges.

[0062] The second embodiment of this application effectively learns and captures the complex relationships between nodes and edges in the metaverse through a pre-built and trained graph neural network model. Compared with traditional resource scheduling methods, graph neural networks have stronger processing capabilities and can identify important features and patterns in massive and complex graph structure data, thereby achieving more accurate resource demand prediction and allocation.

[0063] Secondly, by comprehensively considering the scheduling of computing, storage, and network resources, and through multi-dimensional analysis using a graph neural network model, coordinated management among various resources can be achieved. This avoids the bottleneck problem caused by single-resource scheduling in traditional methods, thereby improving the overall resource utilization efficiency of the system.

[0064] The following is a detailed description of the second exemplary embodiment of the primordial universe resource scheduling method:

[0065] Please refer to Figure 2, which is a flowchart illustrating a second exemplary embodiment of the metaverse resource scheduling method of this application. In this embodiment, operation S10: using a pre-trained graph neural network model to predict resource demand and formulating a first resource allocation strategy based on the generated resource demand prediction values ​​includes operations S11 to S12:

[0066] Operation S11 uses a pre-trained graph neural network model to predict the resource requirements of nodes and edges in the pre-constructed graph structure data, and obtains the predicted resource requirements of nodes and edges respectively.

[0067] Operation S12 calculates the resource allocation for nodes and edges based on the predicted resource requirements of nodes and edges.

[0068] In one feasible implementation, operations S01 to S02 may be included before operation S11:

[0069] Operation S01: Generate a graph dataset by extracting the feature vectors of nodes and edges from the pre-constructed graph structure data;

[0070] In this embodiment, graph structure data is pre-constructed. The operation for constructing graph structure data is as follows:

[0071] First, construct the node set V and edge set E in the metaverse environment.

[0072] All key elements in the metaverse are identified and categorized as nodes and edges. For example, the node set V may include user nodes, virtual item nodes, and server nodes. Correspondingly, the edge set E covers user interaction edges, item transaction edges, and server connection edges.

[0073] Furthermore, define the node feature vector X. v and edge eigenvectors X e , where X v ={x1,x2,…,x n}, x i Let X represent the feature of the i-th node. e ={e1,e2,…,e m}, e j This represents the feature vector of the j-th edge. A feature vector is a numerical array used to describe specific attributes of a node or edge. For a node, the feature vector might include its activity level, user level, the value of virtual items, server load capacity, etc. For example, a user node's feature vector might include its login frequency, number of social relationships, and spending history. Similarly, edge feature vectors describe specific attributes of the edge, such as the frequency of user interaction, the amount of item transactions, and the data transfer rate between servers.

[0074] Furthermore, construct node type labels and edge type labels: node type labels can include user type labels, virtual item type labels, and server type labels, respectively denoted by T. u T v and T s This indicates that edge type labels can include user interaction type labels, item transaction type labels, and server connection type labels, respectively represented by R. u R t R s Type labels are identifiers used to categorize nodes and edges, helping the model better understand the structure and meaning of the data.

[0075] Finally, the node set V and the edge set E are constructed into graph structure data G, and each node and edge is attached with a corresponding feature vector and type label. The formula for G is defined as follows:

[0076] Based on the definition of graph structure data G, a graph structure is constructed in the metaverse environment. The graph structure data is represented as follows:

[0077] G(V,E,X v ,X e ,T,R);

[0078] T = {T(v)} i )|vi ∈V};R={R(e ij )|e ij ∈E};

[0079] Wherein, T(v) i R(e) represents the type label of the i-th node. ij ) represents the type label of the ij-th edge.

[0080] For example, in graph-structured data G, a user node consists of its unique identifier, feature vectors (such as login frequency, number of social relationships, etc.), and user type label (such as "advanced user"). Similarly, a user interaction edge may also contain its unique identifier, feature vectors (such as interaction frequency, relationship strength, etc.), and user interaction type label (such as "friend relationship").

[0081] Based on the above operations, graph structure data can be obtained.

[0082] Based on the graph structure data obtained from the above operations, feature vectors of nodes and edges are extracted to generate a graph dataset for training. The implementation method is as follows:

[0083] First, the constructed graph structure data G(V, E, X) is processed. v X e Data cleaning is performed on the node feature vectors X, T, and R to remove redundant and invalid data. v and edge eigenvectors X e Standardization is performed to transform the feature values ​​to a uniform scale range. Then, feature vectors for nodes and edges are extracted; the node feature vector X... v Represented as Edge feature vector X e Represented as Finally, a graph dataset for training is generated, which includes a preprocessed set of nodes V, an edge set E, and node feature vectors X. v and edge eigenvectors X e The graph dataset is represented as: (V, E, X) v X e ).

[0084] For example, suppose there is a user node A and a server node B in a metaverse environment, connected by a user interaction edge C. When extracting feature vectors, the feature vector for user node A is [user type, high current load, frequent interaction], the feature vector for server node B is [server type, low current load, moderate interaction frequency], and the feature vector for user interaction edge C is [high bandwidth, low latency, user interaction edge type]. These feature vectors will be used to generate a graph dataset containing nodes A, B, and edge C.

[0085] Operation S02: Train the pre-built graph neural network model using the graph dataset.

[0086] In one feasible implementation, operation S02 may include operations S021 to S024:

[0087] In step S021, the graph dataset is input into a pre-built graph neural network model;

[0088] It should be noted that, before implementing operation S021, a graph neural network model can be constructed in the following way:

[0089] In some embodiments, the graph neural network model includes an input layer, multiple modified graph convolutional layers, and an output layer.

[0090] The input layer receives node feature vectors and edge feature vectors, and performs preliminary processing to adapt them for the computation of the graph convolutional layer.

[0091] Improved graph convolutional layers are used to extract complex relationships between nodes and edges. The following formula is the calculation formula for obtaining the node representation for each improved graph convolution:

[0092] Among them, H (1) Let N(i) represent the node representation at level l, and let W represent the set of neighboring nodes of node i. (1) This represents the weight matrix of the l-th layer. Let C represent the self-loop weight matrix of the l-th layer. ij σ is a normalization constant, σ represents the activation function, and P(i) represents the set of metapaths associated with node i. It is the meta-path weight matrix, g(e ik ) is the characteristic function related to the metapath, α ij This represents the scheduling priority weight between node i and node j.

[0093] Scheduling priority weight α ij α is calculated using a priority function specific to metaverse resource scheduling. ij The calculation formula is as follows:

[0094] Where ψ is a priority function specific to resource scheduling in the metaverse, and ψ is defined as:

[0095] ψ(h i ,h j ,r ij ) = a T ·ReLU(W a [h i ||hj ||r ij ]);

[0096] Among them, h i and h j Let i and j represent the feature representation vectors of node i and node j, respectively.

[0097] Where a is a learnable priority vector, W a It is a learnable weight matrix, || denotes the vector concatenation operation, ReLU represents the activation function, and r ij The feature vector related to resource scheduling in the metaverse is calculated based on the following formula, r. ij The calculation formula is as follows:

[0098] r ij =[load(i),load(j),bandwidth(e ij ),latency(e ij ),interaction(i,j)];

[0099] Where load(i) and load(j) represent the current loads of nodes i and j, respectively, and bandwidth(e ij ) represents edge e ij bandwidth, latency(e ij ) represents edge e ij The delay, interaction(i,j) represents the interaction frequency between node i and node j;

[0100] Furthermore, based on the above implementation method, a dynamic adaptive learning mechanism is introduced, and an adaptive learning algorithm is designed to dynamically adjust the structure and parameters of the graph neural network to adapt to changing resource requirements. The adaptive learning mechanism can be expressed by the following formula:

[0101] Where η is the learning rate. λ is the loss function, Y is the actual resource requirement, and λ is the regularization parameter. By dynamically adjusting the learning rate and the regularization parameter, adaptive adjustment of the network structure and parameters can be achieved.

[0102] Furthermore, the output layer generates resource demand prediction results, utilizing the final representation H of nodes and edges. (1) Resource demand forecasting is performed, including the demand values ​​for computing resources, storage resources, and network resources. The output layer generates the forecast results through a fully connected layer and a Softmax function.

[0103] In summary, after constructing the above graph neural network model, the following operation is performed: input the graph dataset into the pre-constructed graph neural network model.

[0104] In operation S022, the backpropagation algorithm is used to calculate the gradient of the loss function with respect to the parameters of the graph neural network model;

[0105] In some embodiments, the gradients of the model parameters are calculated layer by layer using the backpropagation algorithm. Where L represents the loss function, W (l) This represents the weight matrix of the l-th layer.

[0106] The loss function L is defined as the resource demand prediction error, that is, the difference between the resource demand value predicted by the model and the actual resource demand value. The definition formula is as follows:

[0107] in, Let y represent the resource demand value predicted by the model for the i-th sample. i This represents the actual resource requirement value for the i-th sample, and N represents the total number of samples.

[0108] The basic principle of the backpropagation algorithm described above is to calculate the gradient of the loss function with respect to the network parameters in order to update the parameters and thus minimize the loss function. In some embodiments, for each neuron in the output layer, the partial derivative of the loss function with respect to the output value of that neuron is calculated. Using the chain rule, the gradient of the loss function with respect to the activation values ​​of the output layer is decomposed into the product of several factors. Using these factors, the gradient of the loss function with respect to the weights and biases of the output layer is calculated. For each hidden layer, the gradient of the loss function with respect to the activation values ​​of that layer is calculated layer by layer in the reverse order of forward propagation, and the gradient of the loss function with respect to the weights and biases of that layer is calculated using the chain rule. During backpropagation, the error is gradually propagated back from the output layer to the input layer, and the corresponding gradients of the weights and biases are updated at each step.

[0109] In operation S023, the parameters of the graph neural network model are updated based on the gradient using the gradient descent algorithm.

[0110] Updating the parameters of a graph neural network model using the gradient descent algorithm can be expressed by the following formula:

[0111] This formula indicates that, by taking the current parameter value W... (1) Subtracting the product of the gradient and the learning rate η yields the new parameter value W. (l+1) .

[0112] In operation S024, the loss function is minimized through iterative training until the loss function converges to a preset threshold, thereby obtaining the optimized graph neural network model.

[0113] In some embodiments, based on the gradient obtained from the above operations, the parameters of the model are gradually adjusted according to a predetermined learning rate to minimize the loss function.

[0114] Finally, the loss function L is minimized through iterative training until it converges to a preset threshold. This means that the model's prediction error has been reduced to an acceptable range, and the model has been sufficiently trained and optimized.

[0115] In summary, based on the above operations S01 to S02, a graph neural network model is constructed and trained. Based on this, operation S11 is performed: the pre-trained graph neural network model is used to predict the resource demand of nodes and edges in the pre-constructed graph structure data, and the predicted resource demand values ​​of nodes and edges are obtained respectively.

[0116] In this embodiment, the preprocessed graph dataset is input into a trained graph neural network model. The graph neural network model is then used to predict the resource requirements of nodes and edges in the metaverse environment, thereby obtaining the computational resource requirements of the nodes. Storage resource requirements and network resource demand value For each node v i ∈V, based on its node feature vector X i Based on the characteristics of neighboring nodes, calculate their predicted computational resource requirements. The calculation formula is as follows:

[0117] Among them, W c To calculate the resource demand weight matrix, W represents the weights used in calculating resource demand forecasts. x W is the node feature weight matrix, representing the weight of the influence of node features on computing resource requirements. n Let b be the neighbor node feature weight matrix, representing the weight of the influence of neighbor node features on computational resource requirements. c To compute the resource demand bias vector, let α represent the bias used to compute resource demand forecasts. ij σ is the scheduling priority weight, representing the scheduling priority between node i and node j; σ is the activation function, representing a nonlinear activation function; ReLU is the rectified linear unit, representing a nonlinear function used for activation; and N(i) is the set of neighboring nodes of node i, representing the set of nodes directly connected to node i.

[0118] Further, calculate v for each node. iPredicted storage resource requirements for ∈V and network resource demand forecasts The calculation formula is as follows:

[0119] Among them, W s Here is a storage resource demand weight matrix, representing the weights used for storage resource demand forecasting, b s Let g be the storage resource demand bias vector, representing the bias in the storage resource demand forecast. s,i W is a storage resource feature vector, representing specific characteristics of storage resource requirements. n Here, b represents the weights used for network resource demand forecasting. n Let g be the network resource demand bias vector, representing the bias in network resource demand forecasting. n,i This is a network resource feature vector, representing the specific characteristics of network resource demand.

[0120] Furthermore, for each edge e ij ∈E, based on its edge eigenvector e ij Based on the characteristics of the connected nodes, calculate their predicted computational resource requirements. Storage resource demand forecast and network resource demand forecasts The calculation formula is as follows:

[0121] W e This represents the edge feature vector e. ij The weight matrix W. When calculating resource demand forecasts, W... e Used to map edge feature vectors to a new feature space. W xi and W xj Describes the feature vector x used for nodes i and j. i and x j The weight matrix W. When calculating resource demand forecasts, W... xi and W xj It is used to map node feature vectors to a new feature space and combine them with edge features for comprehensive calculation.

[0122] Among them, W ce To calculate the resource demand weight matrix, let b represent the weights used in calculating resource demand forecasts. ce To compute the resource demand bias vector, g represents the bias used to compute resource demand forecasts. ce,ij W is a computational resource feature vector, representing specific characteristics of computational resource requirements. se Here is a storage resource demand weight matrix, representing the weights used for storage resource demand forecasting, b seLet g be the storage resource demand bias vector, representing the bias in the storage resource demand forecast. se,ij W is a storage resource feature vector, representing specific characteristics of storage resource requirements. ne Here, b represents the weights used for network resource demand forecasting. ne Let g be the network resource demand bias vector, representing the bias in network resource demand forecasting. ng,ij This is a network resource feature vector, representing the specific characteristics of network resource demand.

[0123] In summary, we obtain the predicted resource requirements for nodes and edges. Based on this, operation S12 is performed.

[0124] Operation S12 calculates the resource allocation for nodes and edges based on the predicted resource requirements of nodes and edges.

[0125] In some embodiments, based on the predicted resource requirements of nodes and edges. Calculate the total resource requirements for each node and edge.

[0126] The formula for calculating the total resource requirements of a node is as follows:

[0127] The formula for calculating the total resource requirement of an edge is as follows:

[0128] Furthermore, the resource allocation vector R i and R represents the resource allocation of nodes and edges, respectively. i and The definition formula is as follows:

[0129] R i =[R ci ,R s,i ,R n，i ];

[0130] Among them, R c,i For the computing resources allocated to node i, R s,i For the storage resources allocated to node i, R n,i For network resources allocated to node i, To assign edge e ij Computing resources To assign edge e ij Storage resources, To assign edge e ij Network resources.

[0131] In some embodiments, the initial allocation of resources to nodes is performed using the following formula:

[0132] Among them, R i This represents the total amount of resources allocated to node i.

[0133] In some embodiments, the initial allocation of resources to edges is performed using the following formula:

[0134] in, This indicates that the edge e is assigned to the edge. ij Total resources.

[0135] Please refer to Figure 3, which is a flowchart illustrating a third exemplary embodiment of the metaverse resource scheduling method of this application. In this embodiment, operation S20: analyzing resource usage data using the graph neural network model, dynamically adjusting the first resource allocation strategy based on the analysis results, and updating the second resource allocation strategy includes operations S21 to S24:

[0136] In operation S21, the resource usage data of nodes and edges is monitored by a pre-deployed resource monitoring module, and the resource usage data is dynamically analyzed to calculate the resource utilization rate.

[0137] In operation S22, the graph neural network model is used to analyze the input resource usage data and calculate the resource usage change trends of nodes and edges.

[0138] In operation S23, the bottleneck node and bottleneck edge are identified using the graph neural network model, and the resource stress of the bottleneck node and bottleneck edge is calculated based on the resource utilization rate.

[0139] In operation S24, based on the resource usage trend and the resource scarcity, the first resource allocation strategy is dynamically adjusted using the graph neural network model to update and obtain the second resource allocation strategy.

[0140] It should be noted that a resource monitoring module can be deployed in the metaverse environment before implementing operation S21 to monitor the usage of computing resources, storage resources and network resources in real time, and thus collect resource usage data.

[0141] In some embodiments, a resource monitoring module is deployed to monitor the computing resource usage of each node and edge in real time, collect data such as CPU utilization, memory usage, and computing task load of each node and edge, and generate a computing resource usage dataset D. c ;

[0142] The resource monitoring module also monitors the storage resource usage of each node and edge in real time, collecting data such as storage space utilization, data read / write speed, and storage task load, and generating a storage resource usage dataset D. s ;

[0143] The resource monitoring module also monitors the network resource usage of each node and edge in real time, collecting data such as bandwidth utilization, network latency, and network traffic of each node and edge, and generating a network resource usage dataset D. n .

[0144] Furthermore, the resource usage data of nodes and edges is monitored by a pre-deployed resource monitoring module, and the resource usage data is dynamically analyzed to calculate the resource utilization rate.

[0145] In some embodiments, the resource monitoring module dynamically analyzes the collected usage data of computing resources, storage resources, and network resources to calculate the real-time resource utilization rate U. c,i U s,i U n,i , in:

[0146] U c,i : Calculate resource utilization rate, representing the CPU utilization rate of node i at a certain moment. The formula is:

[0147] Among them, CPU utilization is the CPU resources currently used by node i, and total CPU capacity is the maximum CPU resources that node i can provide.

[0148] U s,i Storage resource utilization rate represents the amount of storage space used by node i at a certain moment. The formula is:

[0149] Among them, storage space usage is the storage resources currently used by node i, and total storage space is the maximum storage resources that node i can provide.

[0150] U n,i Network resource utilization rate represents the bandwidth utilization rate of node i at a certain moment. The formula is:

[0151] Among them, bandwidth utilization rate is the network bandwidth resource currently used by node i, and total bandwidth capacity is the maximum network bandwidth resource that node i can provide.

[0152] Calculate resource utilization, representing edge e ij CPU utilization at a given moment. The formula is:

[0153] Among them, CPU utilization is e ij The current CPU resources used, the total CPU capacity is e ij The maximum CPU resources that can be provided.

[0154] Storage resource utilization rate, representing edge e ij The amount of storage space used at a given moment. The formula is:

[0155] Among them, the storage space usage is edge e ij The current storage resources used have a total storage space of e. ij The maximum storage resources that can be provided.

[0156] Network resource utilization rate, representing edge e ij The bandwidth utilization rate at a given moment. The formula is:

[0157] Among them, bandwidth utilization is edge e ij The current network bandwidth resources used have a total bandwidth capacity of e. ij The maximum network bandwidth resources that can be provided.

[0158] Furthermore, the graph neural network model is used to analyze the input resource usage data to calculate the resource usage trends of nodes and edges.

[0159] In some embodiments, the dynamic resource usage data D monitored in real time c D s D n The data is input into a graph neural network model, which analyzes the input data and calculates the resource usage trend ΔU for each node and edge. c,i ΔU s，i ΔU n,i , The formula for calculating the trend of change is as follows:

[0160] ΔU c,i =U c，i (t)-U c，i (t-1);

[0161] ΔU s,i =U s,i (t)-U s,i (t-1);

[0162] ΔU n,i =U n，i (t)-U n，i (t-1);

[0163] Among them, U c,i (t) represents the computing resource utilization rate of node i at time t, U s，i U represents the storage resource utilization rate of node i at time t. n，i This represents the network resource utilization rate of node i at time t. Representing edge e ij The utilization rate of computing resources, storage resources, and network resources at time point t.

[0164] Furthermore, the graph neural network model is used to identify bottleneck nodes and bottleneck edges in resource usage, and the resource stress of the bottleneck nodes and bottleneck edges is calculated based on the resource utilization rate.

[0165] In some embodiments, a graph neural network model is used to identify potential bottlenecks in resource usage and to calculate the resource stress T of the bottleneck node and bottleneck edge. i and The formula for calculating tension is as follows:

[0166] Furthermore, based on the resource usage trend and the resource scarcity, the first resource allocation strategy is dynamically adjusted using the graph neural network model to update and obtain the second resource allocation strategy.

[0167] In some embodiments, based on the resource stress (including computational resource stress, storage resource stress, and network resource stress) of each node and edge calculated in previous operations, and the corresponding resource usage trends, the graph neural network model can iteratively update the resource allocation vectors of nodes and edges using optimization techniques in deep learning algorithms, such as gradient descent or its variants. These resource allocation vectors represent the allocation of each node and edge across different resource dimensions. The update process aims to maximize overall resource utilization efficiency while ensuring that the resource needs of critical tasks or bottleneck nodes are met. Finally, the graph neural network model outputs the updated resource allocation vectors, updating the resource allocation vectors R of nodes and edges. i and

[0168] This application further considers that, in traditional resource scheduling methods, due to the failure to fully consider the differences in node importance, changes in resource utilization, and the real-time nature of user needs, the resource needs of critical tasks and important users cannot be met in a timely manner, resulting in resource waste and a decline in overall system performance.

[0169] Based on this, the third embodiment of this application proposes a solution: using a pre-trained graph neural network model to predict resource demand, and formulating a first resource allocation strategy based on the generated resource demand prediction value; using the graph neural network model to analyze resource usage data, dynamically adjusting the first resource allocation strategy according to the analysis results, and updating to obtain a second resource allocation strategy; and optimizing the second resource allocation strategy through the graph neural network model according to pre-calculated priority indicators to generate a resource allocation scheme.

[0170] This application's embodiments comprehensively consider factors such as the importance of nodes and edges, resource utilization, and user needs to generate an intelligent resource scheduling scheme. This scheme identifies high-priority nodes and edges, prioritizing their resource needs and ensuring the performance of important tasks and the user experience. Simultaneously, intelligent optimization reduces resource waste and improves resource utilization.

[0171] The following is a detailed description of the third exemplary embodiment of the primordial universe resource scheduling method:

[0172] Referring to Figure 4, which is a flowchart illustrating a fourth exemplary embodiment of the metaverse resource scheduling method of this application, the metaverse resource scheduling method includes operations S10 to S40:

[0173] Operation S10 uses a pre-trained graph neural network model to predict resource demand and formulates a first resource allocation strategy based on the generated resource demand prediction values.

[0174] In operation S20, the graph neural network model is used to analyze the resource usage data, and the first resource allocation strategy is dynamically adjusted based on the analysis results to update and obtain the second resource allocation strategy.

[0175] In operation S30, the second resource allocation strategy is optimized using the graph neural network model based on the pre-calculated priority index to generate a resource allocation scheme.

[0176] Operation S40 executes resource scheduling operations according to the resource allocation scheme and dynamically adjusts resource configuration.

[0177] It should be noted that, in this embodiment, the implementation methods of operations S10 to S20 can be referred to the above description, and will not be repeated hereafter.

[0178] In one feasible implementation, operation S30 may include operations S31 to S33:

[0179] Operation S31 calculates the importance parameters and resource utilization of nodes and edges in the metaverse environment;

[0180] In some embodiments, the importance I of each node is calculated.i And the importance of each edge The importance of a node is determined by its role and connectivity, while the importance of an edge is determined by the importance of the nodes it connects to and the frequency of their interactions.

[0181] Among them, W j The importance weight of neighboring nodes is represented by ||N(i)||, which represents the number of neighboring nodes of node i.

[0182] in, Representing edge e ij The frequency of interaction;

[0183] Calculate the resource utilization rate U for each node and edge. i and Resource utilization includes computing resource utilization, storage resource utilization, and network resource utilization:

[0184] Operation S32: Calculate the priority indicators of nodes and edges based on user demand parameters, importance parameters, and resource utilization rates.

[0185] In some embodiments, the priority P of each node and edge is calculated according to user requirements. i and Priority is determined based on the urgency and importance of user needs, calculated using the following formula:

[0186] P i =d i ·I i ·U i ;

[0187] Where, d i and These represent the user demand weights for nodes and edges, respectively.

[0188] It should be noted that user demand values ​​can be obtained through methods such as market research, user interviews, questionnaires, and data analysis. These methods aim to gain a deep understanding of users' actual needs, preferences, and the degree to which they value different functions and features. In some embodiments, detailed questionnaires can be designed to assess the urgency and importance of different aspects of a product or service to users, such as functional requirements, performance requirements, price sensitivity, and user experience. By collecting and analyzing this feedback data, the value of user demand can be quantified, and then used to calculate the priority sum of each node and edge.

[0189] During the calculation process, each user request is assigned a weight, reflecting its importance to the user. Urgency is assessed by the time frame required to fulfill the user's expressed request. Combining these weights and urgency factors, the priority of each node and edge can be calculated using a formula.

[0190] In operation S33, the second resource allocation strategy is optimized and a resource allocation scheme is generated by using the graph neural network model based on the priority indicators of nodes and edges.

[0191] In some embodiments, the priorities of nodes and edges are input into a graph neural network model for resource scheduling optimization to generate a final resource allocation scheme. The final resource allocation scheme includes the reallocation of computing resources, storage resources, and network resources to ensure optimal resource utilization and maximum satisfaction of user needs.

[0192] Based on this, operation S40 is executed: resource scheduling operation is performed according to the resource allocation scheme, and resource configuration is dynamically adjusted.

[0193] In some embodiments, the system performs resource scheduling operations according to the final resource allocation scheme, dynamically adjusts the resource allocation of each node and edge, and allocates computing resources, storage resources and network resources to ensure that the resource needs of each node are met while ensuring the stability and efficiency of system performance.

[0194] Furthermore, this application also proposes a metaverse resource scheduling device. Referring to Figure 5, which is a schematic diagram of the module structure of the metaverse resource scheduling device of this application, the metaverse resource scheduling device includes:

[0195] The first strategy module 10 uses a pre-trained graph neural network model to predict resource demand and formulates a first resource allocation strategy based on the generated resource demand prediction values.

[0196] The second strategy module 20 uses the graph neural network model to analyze resource usage data, dynamically adjusts the first resource allocation strategy based on the analysis results, and updates the second resource allocation strategy.

[0197] The metaverse resource scheduling device provided in this application employs the metaverse resource scheduling method described in the above embodiments, aiming to solve the technical problem of untimely dynamic resource scheduling in traditional resource scheduling methods, and to achieve efficient resource management and system performance optimization in the metaverse environment. Compared with the prior art, the beneficial effects of the metaverse resource scheduling device provided in this application are the same as those of the metaverse resource scheduling method provided in the above embodiments, and other technical features in the metaverse resource scheduling device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0198] This application provides a metaverse resource scheduling device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the metaverse resource scheduling method in the above embodiment 1.

[0199] Referring to Figure 6 below, a schematic diagram of the structure of the metaverse resource scheduling device used to implement the embodiments of this application is shown. The metaverse resource scheduling device in the embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. The metaverse resource scheduling device shown in Figure 6 is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0200] As shown in Figure 6, the metaverse resource scheduling device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the metaverse resource scheduling device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the metaverse resource scheduling device to communicate wirelessly or wiredly with other devices to exchange data. While the figure shows a metaverse resource scheduling device with various systems, it should be understood that implementation or possession of all the systems shown is not required. More or fewer systems may be implemented alternatively.

[0201] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0202] The metaverse resource scheduling device provided in this application employs the metaverse resource scheduling method described in the above embodiments, aiming to solve the technical problem of untimely dynamic resource scheduling in traditional resource scheduling methods, and to achieve efficient resource management and system performance optimization in the metaverse environment. Compared with the prior art, the beneficial effects of the metaverse resource scheduling device provided in this application are the same as those of the metaverse resource scheduling method provided in the above embodiments, and other technical features in this metaverse resource scheduling device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0203] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0204] The above description is only a partial embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0205] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the metaverse resource scheduling method in the above embodiments.

[0206] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0207] The aforementioned computer-readable storage medium may be included in the metaverse resource scheduling device; or it may exist independently and not be assembled into the metaverse resource scheduling device.

[0208] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0209] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0210] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0211] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described metaverse resource scheduling method. It aims to solve the technical problem of untimely dynamic resource scheduling in traditional resource scheduling methods, and to achieve efficient resource management and system performance optimization in the metaverse environment. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the metaverse resource scheduling method provided in the above embodiments, and will not be repeated here.

[0212] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the operation of the metaverse resource scheduling method described above.

[0213] The computer program product provided in this application aims to solve the technical problem of untimely dynamic resource scheduling in traditional resource scheduling methods, and to achieve efficient resource management and system performance optimization in the metaverse environment. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the metaverse resource scheduling method provided in the above embodiments, and will not be repeated here.

[0214] Compared to existing technologies, the metaverse resource scheduling method, apparatus, device, medium, and computer product proposed in this application extracts the business characteristic information of the target service, standardizes the business characteristic information to obtain standard characteristic data, hashes the standard characteristic data to obtain unique characteristic data, numerically processes and concatenates the unique characteristic data to obtain a first business characteristic value, accumulates the first business characteristic value of the target service to obtain a target business characteristic value, and finally compares the target business characteristic value with the characteristic value set to obtain the metaverse resource scheduling result. This is more efficient, flexible, and reliable than the traditional method of generating unique keys or consecutive serial numbers for each service to identify duplicate services. Based on this application's solution, by transforming complex scenarios into a series of simple transformations, the comparison process becomes highly intuitive and efficient. The system only needs to simply compare whether these two values ​​are equal to quickly determine whether two services are completely identical.

[0215] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0216] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0217] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, controlled terminal, or network device, etc.) to execute the methods of each embodiment of this application.

[0218] The above are only some embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural changes made based on the content of this application’s specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A metaverse resource scheduling method, wherein, The metaverse resource scheduling method includes: Resource demand is predicted using a pre-trained graph neural network model, and a first resource allocation strategy is formulated based on the generated resource demand predictions. The resource usage data is analyzed using the graph neural network model, and the first resource allocation strategy is dynamically adjusted based on the analysis results to obtain the second resource allocation strategy.

2. The metaverse resource scheduling method as described in claim 1, wherein, The step of analyzing resource usage data using the graph neural network model, dynamically adjusting the first resource allocation strategy based on the analysis results, and updating the second resource allocation strategy further includes: Based on the pre-calculated priority index, the second resource allocation strategy is optimized through the graph neural network model to generate a resource allocation scheme.

3. The metaverse resource scheduling method as described in claim 2, wherein, The step of optimizing the second resource allocation strategy based on the pre-calculated priority index using the graph neural network model to generate a resource allocation scheme further includes: Resource scheduling operations are performed according to the resource allocation scheme, and resource configuration is dynamically adjusted.

4. The metaverse resource scheduling method as described in claim 2, wherein, The step of optimizing the second resource allocation strategy using the graph neural network model based on pre-calculated priority indicators to generate a resource allocation scheme includes: Calculate the importance parameters and resource utilization of nodes and edges in the metaverse environment; Based on the user demand parameters, the importance parameters, and the resource utilization rate, calculate the priority indicators for the nodes and the edges respectively; The resource allocation scheme is generated by optimizing the second resource allocation strategy based on the priority index of the nodes and edges using the graph neural network model.

5. The metaverse resource scheduling method as described in claim 4, wherein, The importance parameter of a node is determined based on the node's role and connectivity.

6. The metaverse resource scheduling method as described in claim 4, wherein, The importance parameter of the edge is determined based on the importance and interaction frequency of the nodes connected by the edge.

7. The metaverse resource scheduling method as described in claim 4, wherein, The resource utilization rate is obtained based on the computational resource utilization rate, storage resource utilization rate, and network resource utilization rate.

8. The metaverse resource scheduling method as described in claim 1, wherein, The operation of using a pre-trained graph neural network model to predict resource demand and formulating a first resource allocation strategy based on the generated resource demand prediction values ​​includes: The pre-trained graph neural network model is used to predict the resource demand of nodes and edges in the pre-constructed graph structure data, and the predicted resource demand values ​​of the nodes and edges are obtained respectively. Based on the predicted resource demand values ​​of the nodes and edges, calculate the resource allocation amounts for the nodes and edges respectively.

9. The metaverse resource scheduling method as described in claim 8, wherein, Before the step of using the pre-trained graph neural network model to predict resource demand for nodes and edges in the pre-constructed graph structure data, and obtaining the predicted resource demand values ​​for the nodes and edges respectively, the method further includes: A graph dataset is generated by extracting the feature vectors of the nodes and edges from the pre-constructed graph structure data; The pre-built graph neural network model is trained using the graph dataset.

10. The metaverse resource scheduling method as described in claim 9, wherein, The operations for constructing the graph structure data include: Construct the set of nodes and the set of edges in the metaverse environment; Define the feature vector of the node and the feature vector of the edge; Construct the type labels for the nodes and the type labels for the edges; The node set and the edge set are used to construct the graph structure data. Each node in the node set is appended with a corresponding feature vector and a type label. Similarly, each edge in the edge set is appended with a corresponding feature vector and a type label.

11. The metaverse resource scheduling method as described in claim 9, wherein, The operation of generating a graph dataset by extracting feature vectors of nodes and edges from the pre-constructed graph structure data includes: Data cleaning is performed on the constructed graph structure data; The feature vectors of the nodes and the feature vectors of the edges in the graph structure data are subjected to eigenvalue scaling. The graph dataset is generated based on the feature vectors of the nodes and the feature vectors of the edges extracted from the graph structure data.

12. The metaverse resource scheduling method as described in claim 9, wherein, The operation of training the pre-built graph neural network model using the graph dataset includes: The graph dataset is input into the pre-built graph neural network model; The backpropagation algorithm is used to calculate the gradient of the loss function with respect to the parameters of the graph neural network model; The parameters of the graph neural network model are updated based on the gradient using the gradient descent algorithm. The loss function is minimized through iterative training until it converges to a preset threshold, thus obtaining the optimized graph neural network model.

13. The metaverse resource scheduling method as described in claim 1, wherein, The operation of analyzing resource usage data using the graph neural network model, dynamically adjusting the first resource allocation strategy based on the analysis results, and updating the second resource allocation strategy includes: The resource usage data of nodes and edges is monitored by a pre-deployed resource monitoring module, and the resource usage data is dynamically analyzed to calculate the resource utilization rate. The graph neural network model is used to analyze the input resource usage data and calculate the resource usage trends of the nodes and edges. The bottleneck nodes and bottleneck edges are identified using the graph neural network model, and the resource stress of the bottleneck nodes and bottleneck edges is calculated based on the resource utilization rate. Based on the resource usage trend and resource scarcity, the first resource allocation strategy is dynamically adjusted using the graph neural network model to update and obtain the second resource allocation strategy.

14. A metaverse resource scheduling device, wherein, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the metaverse resource scheduling method as described in any one of claims 1 to 13.

15. A storage medium, wherein, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the operation of the metaverse resource scheduling method as described in any one of claims 1 to 13.

16. A computer program product, wherein, The computer program product includes a computer program that, when executed by a processor, implements the operation of the metaverse resource scheduling method as described in any one of claims 1 to 13.

Images