Calculation network multi-element resource abstraction and modeling method

Abstract

The invention discloses a computing network multi-element resource abstraction and modeling method. The method comprises the following steps: S1, demand analysis; s2, resource classification; s3, extracting resource features; s4, constructing an abstract model; s5, resource quantification and measurement; s6, establishing resource association, and analyzing a dependency relationship and a cooperative relationship among different resource elements; s7, verifying and optimizing the model; s8, defining an interface; and S9, performing document processing. According to the method, the required resource elements can be determined in a more targeted manner, excessive or insufficient allocation of the resources is avoided, the resources are classified and subdivided in detail, the resources can be comprehensively and accurately understood and managed, the complex dependence and cooperative relation between the resources can be better processed, the resource utilization efficiency and the system performance are improved, and the resource utilization rate is increased. The effectiveness and adaptability of the model in an actual computing network environment can be ensured, the accuracy and practicability of the model are improved, and efficient interaction and integration between different systems and assemblies can be achieved.

Description

Technical Field

[0001] The present invention relates to the technical field related to computing networks, and in particular to a method for abstracting and modeling multi-factor resources of computing networks. Background Art

[0002] A computing network is a new type of information infrastructure that allocates and flexibly schedules computing and network resources across the cloud, network, and edge based on user or business needs. The construction of a computing network requires expanding network capabilities. This requires not only distributing network resource information but also distributing computing resource information for computing nodes. Furthermore, it must be able to schedule network resources and allocate computing resources based on user or business needs.

[0003] Existing computing network technology is developing rapidly but is still in a stage of continuous evolution. The management and understanding of resources is not mature enough, and there is a lack of systematic, comprehensive and detailed classification methods, which leads to confusion in resource classification. Furthermore, in technical research and practice, the exploration of resource characteristics is not deep enough, and their essence and characteristics are not fully understood. In addition, when constructing models, the combination of theory and practice is not close enough, resulting in unclear abstraction and logic of the model. In addition, there is a lack of unified norms and standards for resource quantification and comparison, making it difficult to conduct accurate evaluation and comparison. There is insufficient understanding of the complexity of the relationship between resources, and a lack of effective analysis and modeling methods. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for abstracting and modeling multi-factor resources in a computing network to solve the problems raised in the above background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for abstracting and modeling multi-factor resources in a computing network, comprising the following steps:

[0006] S1: Demand analysis: clarify the application scenarios and business requirements of the computing network and determine the required resource elements;

[0007] S2: Resource classification, classifying various resources in the computing network and further subdividing each category;

[0008] S3: Resource feature extraction: For each resource type, extract its key features and performance indicators;

[0009] S4: Abstract model construction, based on the extracted resource characteristics, establish an abstract mathematical model or logical model;

[0010] S5: Resource quantification and measurement, defining quantitative metrics and units for each resource characteristic, and conducting accurate evaluation and comparison;

[0011] S6: Establish resource associations, analyze the dependencies and synergies between different resource elements, and reflect the mutual influence between resources by establishing an association model;

[0012] S7: Model verification and optimization: Use actual computing network data and scenarios to verify and test the constructed model, and optimize and adjust the model based on the verification results;

[0013] S8: Interface definition, defining the interface for interacting with the model;

[0014] S9: Documentation: Provide detailed documentation of the abstraction and modeling process, methods, model structure, and parameters.

[0015] Preferably, the demand analysis in step S1 specifically includes the following steps:

[0016] S11: Market research to understand the development trends of the industry in which the computing network is located, the situation of competitors, and potential market opportunities;

[0017] S12: Determine business objectives and communicate with relevant business departments to clarify the short-term and long-term goals of the business that the computing network will support;

[0018] S13: User demand collection: Through user interviews, questionnaires, and user feedback, we gather direct user expectations and requirements for the functions, performance, and usability of the computing network.

[0019] S14: Application scenario definition, which describes in detail the specific scenarios in which the computing network may be applied, including the scenario's environment, conditions, participants, and processes;

[0020] S15: Existing system assessment: Analyze the existing computing network system or related infrastructure to identify its strengths, weaknesses, and limitations;

[0021] S16: Clarify key business processes, identifying key business processes related to the computing network, and the resource requirements and dependencies of these processes;

[0022] S17: Risk and constraint identification: Identify the technical, cost, and time risks and constraints faced in implementing computing network application scenarios and meeting business needs.

[0023] S18: Prioritization: prioritize different needs and application scenarios based on factors such as business importance and urgency.

[0024] Preferably, the resource classification in step S2 is to preliminarily divide the resources in the computing network into major categories, and preliminarily divide the resources into computing resources, network resources, storage resources, energy resources and software resources, security resources and geographical resources. It is also necessary to classify the special resources in the computing network integration business.

[0025] Preferably, the resource feature extraction in step S3 specifically includes the following steps:

[0026] S31: Data collection, collecting detailed technical specifications, performance test reports, usage records and other data related to each resource type;

[0027] S32: Key feature identification, identifying and processing computing resources, network resources, storage resources, energy resources, software resources, security resources, and geographic resources;

[0028] S33: Determine performance indicators. Different indicators are determined for different resources, fully integrating the unique characteristics and performance indicators of resources in computing-network convergence business scenarios.

[0029] S34: Environmental considerations: consider the impact of the physical environment in which the resource is located on its characteristics and performance;

[0030] S35: Load analysis, analyzing the characteristics and performance of resources under different load conditions;

[0031] S36: Dynamic characteristics research, focusing on the dynamic change characteristics of resources;

[0032] S37: Reliability and availability evaluation, identifying characteristics related to reliability;

[0033] S38: Cost considerations, including acquisition costs, operating costs, maintenance costs, and other economic characteristics related to resources.

[0034] Preferably, the abstract model construction in step S4 includes the following steps:

[0035] S41: Select the model type. Based on resource characteristics and business requirements, determine the appropriate model type. The established model is adapted to the complex needs of computing-network convergence services.

[0036] S42: Define variables, convert the extracted key resource characteristics and performance indicators into variables in the model, and clarify their domain and value range;

[0037] S43: Determine constraints. Establish constraints based on physical limitations of resources, business rules, and quality of service requirements.

[0038] S44: Establishing an objective function, constructing the objective function according to the optimization goal of the computing network;

[0039] S45: Model formalization, using mathematical symbols and logical expressions to formally describe variables, constraints and objective functions to form a complete mathematical or logical model;

[0040] S46: Simplification and approximation: If the model is too complex, reasonable simplification and approximation should be performed to improve the solvability and comprehensibility of the model.

[0041] S47: Verify the model structure to check whether the model structure is reasonable and whether it accurately reflects the relationship between resources and business requirements;

[0042] S48: Initial parameter setting, setting initial values ​​for parameters in the model.

[0043] Preferably, the model construction in step S4 is to use linear regression, Sigmoid function or Softmax function to build the model according to the problem and data characteristics, specifically

[0044] Linear regression equation: y = wx + e

[0045] Where y is the dependent variable, x is the independent variable, w is the coefficient, and e is the error term, which is used to describe the linear relationship between the independent variable and the dependent variable, and can be used to estimate the parameters by methods such as the least squares method;

[0046] Sigmoid function: The value range is between (0,1);

[0047] Softmax function:

[0048] Preferably, in the step S5, the resource quantification and measurement for computing resources can be measured in terms of floating-point operations per second, instructions per second, etc. The capacity of storage resources is measured in bytes, the read and write speed is expressed in bytes per second, the network bandwidth is usually measured in bits per second, and the delay is measured in milliseconds. For energy consumption, watts are used as the unit, and the energy consumption distribution under different working conditions is considered.

[0049] Preferably, establishing resource association in step S6 includes the following steps:

[0050] S61: Determine resource elements and identify the various resource types that need to be considered;

[0051] S62: Data collection and preprocessing: collecting data on resource usage, performance indicators, business needs, and other aspects, and cleaning, organizing, and standardizing the data;

[0052] S63: Dependency analysis, through correlation analysis, causal analysis and other methods, to determine the direct or indirect dependencies between resources;

[0053] S64: Synergy analysis, studying how resources cooperate with each other to improve overall performance or efficiency;

[0054] S65: Select an association model. According to the analysis results and data characteristics, select an appropriate mathematical model or algorithm to establish the association model;

[0055] S66: Model construction and parameter estimation, using the selected model, input data for training and fitting.

[0056] Preferably, the model verification and optimization in step S7 includes the following steps:

[0057] S71: Prepare a validation data set. Separate a portion of independent and representative data from the actual network computing data as the validation data set.

[0058] S72: Set evaluation indicators, determine specific indicators for measuring model performance, and select appropriate indicators based on the specific application and goals of the computing network model;

[0059] S73: Model validation: input the validation dataset into the constructed model to obtain the model output and calculate the performance of the model on the validation dataset based on the set evaluation indicators;

[0060] S74: Result analysis: compare the performance of the model on the validation set with the performance of the expected target or baseline model, analyze the error types and patterns of the model, and identify existing problems and deficiencies;

[0061] S75: Determine the optimization direction. Based on the result analysis, determine the direction in which the model needs to be optimized.

[0062] S76: Optimization implementation: modify and adjust the model accordingly according to the determined optimization direction;

[0063] S77: Iterative optimization. If the optimized model still does not achieve the desired effect, repeat the above steps and perform multiple rounds of optimization and verification until the model performance meets the requirements.

[0064] S78: Final evaluation and confirmation: conduct a comprehensive evaluation of the model after multiple optimizations to ensure that its performance in various indicators and actual scenarios reaches an acceptable level;

[0065] S79: Model deployment and monitoring: deploy the optimized model to the actual computing network system and establish a monitoring mechanism to continuously track the performance and effect of the model in actual operation.

[0066] Preferably, the document in step S9 should include the original data and conclusions of the demand analysis, the basis and standards for resource classification, the method and process for resource feature extraction, and detailed records of the mathematical formulas, logical relationships and algorithms used. The model verification and optimization steps and interface definitions must also be recorded, including the test data used, verification indicators and specific operations and instructions for optimization.

[0067] The technical effects and advantages of the present invention are as follows:

[0068] The present invention accurately matches application scenarios and business needs through demand analysis, classifies and subdivides resources, extracts resource characteristics, establishes clear abstract models, quantifies and measures resources, establishes resource associations, verifies optimization models, defines interfaces, and documents and records. It classifies and subdivides resources in detail, which helps to understand and manage resources more comprehensively and accurately, and can deeply explore the key characteristics and performance indicators of resources, providing a basis for more accurate resource management. It can establish a clearer and more effective resource management model, improve the scientific nature of resource planning and scheduling, and establish standards for resource quantification and measurement, which is conducive to more accurate evaluation and comparison of different resources and more reasonable decision-making. It can better handle complex dependencies and collaborative relationships between resources, improve resource utilization efficiency and system performance, ensure the effectiveness and adaptability of the model in the actual computing network environment, improve the accuracy and practicality of the model, and help achieve efficient interaction and integration between different systems and components. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 This is a flow chart of the method for abstracting and modeling multi-factor resources in computing networks according to the present invention.

[0070] Figure 2 This is a demand analysis flow chart of the present invention.

[0071] Figure 3 This is the resource feature extraction flow chart of the present invention.

[0072] Figure 4 Construct a flow chart for the abstract model of the present invention.

[0073] Figure 5 A resource association flow chart is established for the present invention.

[0074] Figure 6 This is the model verification and optimization flow chart of the present invention. DETAILED DESCRIPTION

[0075] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0076] The present invention provides Figure 1-6 The method for abstracting and modeling multi-factor resources in a computing network includes the following steps:

[0077] S1: Demand analysis: clarify the application scenarios and business requirements of the computing network and determine the required resource elements;

[0078] S2: Resource classification, classifying various resources in the computing network and further subdividing each category;

[0079] S3: Resource feature extraction: For each resource type, extract its key features and performance indicators;

[0080] S4: Abstract model construction, based on the extracted resource characteristics, establish an abstract mathematical model or logical model;

[0081] S5: Resource quantification and measurement, defining quantitative metrics and units for each resource characteristic, and conducting accurate evaluation and comparison;

[0082] S6: Establish resource associations, analyze the dependencies and synergies between different resource elements, and reflect the mutual influence between resources by establishing an association model;

[0083] S7: Model verification and optimization: Use actual computing network data and scenarios to verify and test the constructed model, and optimize and adjust the model based on the verification results;

[0084] S8: Interface definition, defining the interface for interacting with the model;

[0085] S9: Documentation: Provide detailed documentation of the abstraction and modeling process, methods, model structure, and parameters.

[0086] The demand analysis in step S1 specifically includes the following steps:

[0087] S11: Market research to understand the development trends of the industry in which the computing network is located, the situation of competitors, and potential market opportunities;

[0088] S12: Determine business objectives and communicate with relevant business departments to clarify the short-term and long-term goals of the business that the computing network will support;

[0089] S13: User demand collection: Through user interviews, questionnaires, and user feedback, we gather direct user expectations and requirements for the functions, performance, and usability of the computing network.

[0090] S14: Application scenario definition, which describes in detail the specific scenarios in which the computing network may be applied, including the scenario's environment, conditions, participants, and processes;

[0091] S15: Existing system assessment: Analyze the existing computing network system or related infrastructure to identify its strengths, weaknesses, and limitations;

[0092] S16: Clarify key business processes, identifying key business processes related to the computing network, and the resource requirements and dependencies of these processes;

[0093] S17: Risk and constraint identification: Identify the technical, cost, and time risks and constraints faced in implementing computing network application scenarios and meeting business needs.

[0094] S18: Prioritization: prioritize different needs and application scenarios based on factors such as business importance and urgency.

[0095] At this initial stage, in-depth and comprehensive research is needed to clearly define the application scenarios and business needs of the computing network. This includes not only a detailed review of existing businesses, but also forward-looking predictions for possible future business expansion and innovation. For example, for high-performance computing scenarios, it may be necessary to support large-scale scientific simulations and complex model training; in terms of big data processing, it must be able to cope with the rapid storage, retrieval and analysis of massive amounts of data; and cloud computing services need to meet the diverse needs of different users for elastic resource configuration. At the same time, it is crucial to accurately identify and determine the various resource elements that need to be taken into consideration. In addition to common computing resources such as central processing units (CPUs), graphics processing units (GPUs), storage resources, including hard disks, memory, and network bandwidth, attention should also be paid to factors such as energy consumption, heat dissipation capacity, resource availability and reliability, and service quality requirements. The determination of these elements will directly affect the accuracy and effectiveness of subsequent modeling.

[0096] The resource classification in step S2 is to preliminarily classify the resources in the computing network into major categories, and preliminarily divide the resources into computing resources, network resources, storage resources, energy resources, software resources, security resources and geographical resources. It is also necessary to classify the special resources in the computing network integration business.

[0097] Computing resource segmentation: can be further divided into CPU resources, GPU resources, FPGA resources, etc.; for CPU resources, it can also be segmented according to parameters such as the number of cores and main frequency.

[0098] Network resource segmentation: divided into bandwidth resources, latency resources, routing resources, etc.; bandwidth resources can be segmented according to different network links (such as backbone network, metropolitan area network, access network) and speed (such as 100Mbps, 1Gbps, etc.).

[0099] Storage resource segmentation: including memory resources, hard disk resources, etc.; hard disk resources can be segmented according to mechanical hard disks, solid-state hard disks, and different capacity sizes.

[0100] Energy resource segmentation: For example, power resources can be segmented according to different power supply methods (mains power, UPS, etc.) and power consumption levels.

[0101] Software resource segmentation: such as operating systems, middleware, application software, etc.; operating systems can be segmented according to different types (Windows, Linux, etc.) and versions.

[0102] Security resource segmentation: such as firewalls, encryption resources, authentication and authorization resources, etc.

[0103] Geographic resource segmentation: segmentation based on the geographical location, regional coverage, etc. of the data center.

[0104] These subdivision steps help to comprehensively and meticulously classify various resources in the computing network, laying the foundation for subsequent feature extraction, modeling, etc. The specific subdivision method should be adjusted and optimized according to the specific application and needs of the computing network.

[0105] The resource feature extraction in step S3 specifically includes the following steps:

[0106] S31: Data collection, collecting detailed technical specifications, performance test reports, usage records and other data related to each resource type;

[0107] S32: Key feature identification, identifying and processing computing resources, network resources, storage resources, energy resources, software resources, security resources, and geographic resources;

[0108] S33: Determine performance indicators. Different indicators are determined for different resources, fully integrating the unique characteristics and performance indicators of resources in computing-network convergence business scenarios.

[0109] S34: Environmental considerations: consider the impact of the physical environment in which the resource is located on its characteristics and performance;

[0110] S35: Load analysis, analyzing the characteristics and performance of resources under different load conditions;

[0111] S36: Dynamic characteristics research, focusing on the dynamic change characteristics of resources;

[0112] S37: Reliability and availability evaluation, identifying characteristics related to reliability;

[0113] S38: Cost considerations, including acquisition costs, operating costs, maintenance costs, and other economic characteristics related to resources.

[0114] For each resource type, it is crucial to deeply explore and accurately extract its key features and performance indicators. For computing resources, in addition to the number of cores and frequency, cache size, instruction set architecture, parallel processing capabilities, etc. should also be considered. The characteristics of storage resources include not only capacity and read and write speeds, but also data persistence, fault tolerance, and the format and structure of data storage. In terms of network bandwidth, in addition to paying attention to the upper limit and latency, factors such as packet loss rate, jitter, network topology, and routing strategy should also be considered. In addition, for the important feature of energy consumption, the power consumption of the device under different loads, the efficiency of the energy-saving mode, and the overall energy utilization efficiency should be considered.

[0115] The abstract model construction in step S4 includes the following steps:

[0116] S41: Select the model type. Based on resource characteristics and business requirements, determine the appropriate model type. The established model is adapted to the complex needs of computing-network convergence services.

[0117] S42: Define variables, convert the extracted key resource characteristics and performance indicators into variables in the model, and clarify their domain and value range;

[0118] S43: Determine constraints. Establish constraints based on physical limitations of resources, business rules, and quality of service requirements.

[0119] S44: Establishing an objective function, constructing the objective function according to the optimization goal of the computing network;

[0120] S45: Model formalization, using mathematical symbols and logical expressions to formally describe variables, constraints and objective functions to form a complete mathematical or logical model;

[0121] S46: Simplification and approximation: If the model is too complex, reasonable simplification and approximation should be performed to improve the solvability and comprehensibility of the model.

[0122] S47: Verify the model structure to check whether the model structure is reasonable and whether it accurately reflects the relationship between resources and business requirements;

[0123] S48: Initial parameter setting, setting initial values ​​for parameters in the model.

[0124] Based on the extracted rich resource features, appropriate mathematical methods and logical structures are used to construct abstract models. This may involve the use of a hierarchical architecture to divide computing network resources into different layers, such as the physical layer, virtual layer, and application layer, each layer with specific functions and resource allocation rules. Modular design is also a common strategy, encapsulating different types of resources or components with similar functions into independent modules to facilitate management and maintenance, while improving the scalability and flexibility of the model. For example, independent sub-models can be established for the computing module, storage module, and network module, and then integrated into the entire computing network model by defining clear interfaces and interaction rules.

[0125] In step S4, the model is constructed based on the problem and data characteristics, and the three methods of linear regression, Sigmoid function or Softmax function are used to build the model.

[0126] Linear regression equation: y = wx + e

[0127] Where y is the dependent variable, x is the independent variable, w is the coefficient, and e is the error term, which is used to describe the linear relationship between the independent variable and the dependent variable, and can be used to estimate the parameters by methods such as the least squares method;

[0128] Predict changes in network traffic or computing resource demand over time. For example, you can use time as the independent variable and network traffic or resource demand as the dependent variable to build a linear regression model to predict future demand. Analyze the impact of different factors (such as the number of users and business types) on resource consumption, and establish a multivariate linear regression model to determine the weight and impact of each factor.

[0129] Sigmoid function: The value range is between (0,1);

[0130] In resource allocation decisions, the likelihood of resource demand or the probability of resource availability can be represented as the output of a Sigmoid function. For example, based on certain features, it can determine whether a task urgently needs more resources. The output is a probability value between 0 and 1, which can be used to assess network security risks and convert some risk indicators into the probability of risk occurrence.

[0131] Softmax function:

[0132] Classify and allocate resources for different types of network services or computing tasks. For example, based on the characteristics of the tasks, they are classified into high, medium, or low priority, and resource weights are assigned to each category. The selection probability of multiple network topologies or resource configuration schemes is determined, and each scheme is scored based on evaluation metrics. The scores are then converted into selection probabilities using the Softmax function.

[0133] In step S5, resource quantification and measurement for computing resources can be measured in terms of floating-point operations per second, instructions per second, etc. The capacity of storage resources is measured in bytes, the read and write speed is expressed in bytes per second, the network bandwidth is usually measured in bits per second, and the delay is measured in milliseconds. For energy consumption, watts are used as the unit, and the energy consumption distribution under different working conditions is considered.

[0134] Establishing resource association in step S6 includes the following steps:

[0135] S61: Determine resource elements and identify the various resource types that need to be considered;

[0136] S62: Data collection and preprocessing: collecting data on resource usage, performance indicators, business needs, and other aspects, and cleaning, organizing, and standardizing the data;

[0137] S63: Dependency analysis, through correlation analysis, causal analysis and other methods, to determine the direct or indirect dependencies between resources;

[0138] S64: Synergy analysis, studying how resources cooperate with each other to improve overall performance or efficiency;

[0139] S65: Select an association model. According to the analysis results and data characteristics, select an appropriate mathematical model or algorithm to establish the association model;

[0140] S66: Model construction and parameter estimation, using the selected model, input data for training and fitting.

[0141] In-depth analysis of the complex dependencies and synergies between different resource elements is key to building a model that truly reflects the operational status of a computing network. For example, there's a mismatch between access speeds for computing and storage resources. High-speed computing requires correspondingly fast storage access to support data, otherwise it can lead to computing performance bottlenecks. By establishing precise correlation models, we can more accurately predict the impact of resource configuration changes on overall computing network performance, enabling more optimized resource management and allocation.

[0142] The model verification and optimization in step S7 includes the following steps:

[0143] S71: Prepare a validation data set. Separate a portion of independent and representative data from the actual network computing data as the validation data set.

[0144] S72: Set evaluation indicators, determine specific indicators for measuring model performance, and select appropriate indicators based on the specific application and goals of the computing network model;

[0145] S73: Model validation: input the validation dataset into the constructed model to obtain the model output and calculate the performance of the model on the validation dataset based on the set evaluation indicators;

[0146] S74: Result analysis: compare the performance of the model on the validation set with the performance of the expected target or baseline model, analyze the error types and patterns of the model, and identify existing problems and deficiencies;

[0147] S75: Determine the optimization direction. Based on the result analysis, determine the direction in which the model needs to be optimized.

[0148] S76: Optimization implementation: modify and adjust the model accordingly according to the determined optimization direction;

[0149] S77: Iterative optimization. If the optimized model still does not achieve the desired effect, repeat the above steps and perform multiple rounds of optimization and verification until the model performance meets the requirements.

[0150] S78: Final evaluation and confirmation: conduct a comprehensive evaluation of the model after multiple optimizations to ensure that its performance in various indicators and actual scenarios reaches an acceptable level;

[0151] S79: Model deployment and monitoring: deploy the optimized model to the actual computing network system and establish a monitoring mechanism to continuously track the performance and effect of the model in actual operation.

[0152] The verification and optimization process is repeated continuously until the model can accurately reflect the performance and resource utilization of the computing network under various conditions, providing reliable support for resource management decisions.

[0153] In step S8, to achieve seamless interaction and integration between the model and external systems, it is crucial to clearly define the interfaces for interacting with the model. These interfaces should include data input interfaces for receiving resource requests, configuration information, and performance monitoring data from external systems. Data output interfaces should also be defined to provide resource allocation recommendations, performance prediction results, and optimization strategies to external systems. Interface definitions should follow standardized protocols and specifications to ensure compatibility with different computing network management platforms and applications. Security and access control must also be considered in interface design to ensure that only authorized users and systems can access and modify model-related data. Furthermore, a user-friendly interface and programming interface should be provided to facilitate model invocation and secondary development by developers and administrators.

[0154] The document in step S9 should include the original data and conclusions of the demand analysis, the basis and standards for resource classification, the method and process for resource feature extraction, and detailed records of the mathematical formulas, logical relationships and algorithms used. The model verification and optimization steps and interface definitions must also be recorded, including the test data used, verification indicators and specific operations and instructions for optimization. The above are important components of the document so that other developers and users can interact with the model correctly. Through complete documentation work, it will not only facilitate the current team's understanding and use of the model, but also provide valuable reference materials for subsequent maintenance, upgrades and expansions, helping to ensure the sustainability and inheritability of the multi-factor resource abstraction and modeling work of the computing network.

[0155] In summary, the abstraction and modeling of multi-factor resources in a computing network is a complex and systematic project that requires the comprehensive application of multiple technologies and methods, as well as repeated practice and optimization, to build an accurate, practical, and forward-looking model to provide strong support for the efficient operation of the computing network and the optimized resource management.

[0156] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for abstracting and modeling multi-factor resources in a computing network, characterized by: The following steps are involved: S1: Demand analysis: clarify the application scenarios and business requirements of the computing network and determine the required resource elements; S2: Resource classification, classifying various resources in the computing network and further subdividing each category; S3: Resource feature extraction: For each resource type, extract its key features and performance indicators; S4: Abstract model construction, based on the extracted resource characteristics, establish an abstract mathematical model or logical model; S5: Resource quantification and measurement, defining quantitative metrics and units for each resource characteristic, and conducting accurate evaluation and comparison; S6: Establish resource associations, analyze the dependencies and synergies between different resource elements, and reflect the mutual influence between resources by establishing an association model; S7: Model verification and optimization: Use actual computing network data and scenarios to verify and test the constructed model, and optimize and adjust the model based on the verification results; S8: Interface definition, defining the interface for interacting with the model; S9: Documentation: Provide detailed documentation of the abstraction and modeling process, methods, model structure, and parameters.

2. A method for abstracting and modeling multi-factor resources in a computing network according to claim 1, characterized in that: The demand analysis in step S1 includes the following steps: S11: Market research to understand the development trends of the industry in which the computing network is located, the situation of competitors, and potential market opportunities; S12: Determine business objectives and communicate with relevant business departments to clarify the short-term and long-term goals of the business that the computing network will support; S13: User demand collection: Through user interviews, questionnaires, and user feedback, we gather direct user expectations and requirements for the functions, performance, and usability of the computing network. S14: Application scenario definition, which describes in detail the specific scenarios in which the computing network may be applied, including the scenario's environment, conditions, participants, and processes; S15: Existing system assessment: Analyze the existing computing network system or related infrastructure to identify its strengths, weaknesses, and limitations; S16: Clarify key business processes, identifying key business processes related to the computing network, and the resource requirements and dependencies of these processes; S17: Risk and constraint identification: Identify the technical, cost, and time risks and constraints faced in implementing computing network application scenarios and meeting business needs. S18: Prioritization: prioritize different needs and application scenarios based on factors such as business importance and urgency.

3. A method for abstracting and modeling multi-factor resources in a computing network according to claim 1, characterized in that: The resource classification in step S2 is to preliminarily classify the resources in the computing network into major categories, and preliminarily classify the resources into computing resources, network resources, storage resources, energy resources, software resources, security resources and geographical resources. It is also necessary to classify the special resources in the computing network integration business.

4. A method for abstracting and modeling multi-factor resources in a computing network according to claim 1, characterized in that: The resource feature extraction in step S3 includes the following steps: S31: Data collection, collecting detailed technical specifications, performance test reports, usage records and other data related to each resource type; S32: Key feature identification, identifying and processing computing resources, network resources, storage resources, energy resources, software resources, security resources, and geographic resources; S33: Determine performance indicators. Different indicators are determined for different resources, fully integrating the unique characteristics and performance indicators of resources in computing-network convergence business scenarios. S34: Environmental considerations: consider the impact of the physical environment in which the resource is located on its characteristics and performance; S35: Load analysis, analyzing the characteristics and performance of resources under different load conditions; S36: Dynamic characteristics research, focusing on the dynamic change characteristics of resources; S37: Reliability and availability evaluation, identifying characteristics related to reliability; S38: Cost considerations, including acquisition costs, operating costs, maintenance costs, and other economic characteristics related to resources.

5. A method for abstracting and modeling multi-factor resources in a computing network according to claim 1, characterized in that: The abstract model construction in step S4 includes the following steps: S41: Select the model type. Based on resource characteristics and business requirements, determine the appropriate model type. The established model is adapted to the complex needs of computing-network convergence services. S42: Define variables, convert the extracted key resource characteristics and performance indicators into variables in the model, and clarify their domain and value range; S43: Determine constraints. Establish constraints based on physical limitations of resources, business rules, and quality of service requirements. S44: Establishing an objective function, constructing the objective function according to the optimization goal of the computing network; S45: Model formalization, using mathematical symbols and logical expressions to formally describe variables, constraints and objective functions to form a complete mathematical or logical model; S46: Simplification and approximation: If the model is too complex, reasonable simplification and approximation should be performed to improve the solvability and comprehensibility of the model. S47: Verify the model structure to check whether the model structure is reasonable and whether it accurately reflects the relationship between resources and business requirements; S48: Initial parameter setting, setting initial values ​​for parameters in the model.

6. A method for abstracting and modeling multi-factor resources in a computing network according to claim 1, characterized in that: The model construction in step S4 is to use linear regression, Sigmoid function or Softmax function to build the model according to the problem and data characteristics. Linear regression equation: y = wx + e Where y is the dependent variable, x is the independent variable, w is the coefficient, and e is the error term, which is used to describe the linear relationship between the independent variable and the dependent variable, and can be used to estimate the parameters by methods such as the least squares method; Sigmoid function: The value range is between (0,1); Softmax function:

7. A method for abstracting and modeling multi-factor resources in a computing network according to claim 1, characterized in that: In step S5, the resource quantification and measurement for computing resources can be measured in terms of floating-point operations per second, instructions per second, etc. The capacity of storage resources is measured in bytes, the read and write speed is expressed in bytes per second, the network bandwidth is usually measured in bits per second, and the delay is measured in milliseconds. For energy consumption, watts are used as the unit, and the energy consumption distribution under different working conditions is considered.

8. A method for abstracting and modeling multi-factor resources in a computing network according to claim 1, characterized in that: Establishing resource association in step S6 includes the following steps: S61: Determine resource elements and identify the various resource types that need to be considered; S62: Data collection and preprocessing: collecting data on resource usage, performance indicators, business needs, and other aspects, and cleaning, organizing, and standardizing the data; S63: Dependency analysis, through correlation analysis, causal analysis and other methods, to determine the direct or indirect dependencies between resources; S64: Synergy analysis, studying how resources cooperate with each other to improve overall performance or efficiency; S65: Select an association model. According to the analysis results and data characteristics, select an appropriate mathematical model or algorithm to establish the association model; S66: Model construction and parameter estimation, using the selected model, input data for training and fitting.

9. A method for abstracting and modeling multi-factor resources in a computing network according to claim 1, characterized in that: The model verification and optimization in step S7 includes the following steps: S71: Prepare a validation data set. Separate a portion of independent and representative data from the actual network computing data as the validation data set. S72: Set evaluation indicators, determine specific indicators for measuring model performance, and select appropriate indicators based on the specific application and goals of the computing network model; S73: Model validation: input the validation dataset into the constructed model to obtain the model output and calculate the performance of the model on the validation dataset based on the set evaluation indicators; S74: Result analysis: compare the performance of the model on the validation set with the performance of the expected target or baseline model, analyze the error types and patterns of the model, and identify existing problems and deficiencies; S75: Determine the optimization direction. Based on the result analysis, determine the direction in which the model needs to be optimized. S76: Optimization implementation: modify and adjust the model accordingly according to the determined optimization direction; S77: Iterative optimization. If the optimized model still does not achieve the desired effect, repeat the above steps and perform multiple rounds of optimization and verification until the model performance meets the requirements. S78: Final evaluation and confirmation: conduct a comprehensive evaluation of the model after multiple optimizations to ensure that its performance in various indicators and actual scenarios reaches an acceptable level; S79: Model deployment and monitoring: deploy the optimized model to the actual computing network system and establish a monitoring mechanism to continuously track the performance and effect of the model in actual operation.

10. A method for abstracting and modeling multi-factor resources in a computing network according to claim 1, characterized in that: The document in step S9 should include the original data and conclusions of the demand analysis, the basis and standards for resource classification, the method and process for resource feature extraction, and detailed records of the mathematical formulas, logical relationships and algorithms used. The model verification and optimization steps and interface definitions must also be recorded, including the test data used, verification indicators and specific optimization operations and instructions.

Images