Network resource allocation method and device, computer program product and electronic equipment

Abstract

The invention discloses a network resource allocation method and device, a computer program product and electronic equipment. The method relates to the field of big data or other related fields, and comprises the following steps: acquiring resource load data of a network platform in a preset time period, and acquiring network behavior data of a user using the network platform in the preset time period; when the network platform is in the first network state, generating a first resource scheduling strategy according to the resource load data and the network behavior data, and sending the first resource scheduling strategy to the client; when the network platform is in the second network state, the first resource scheduling strategy is adjusted to obtain a second resource scheduling strategy, the second resource scheduling strategy is sent to the client side, and resource scheduling operation is executed at the client side according to the second resource scheduling strategy. Through the method and the device, the technical problems of low resource utilization rate and poor response capability during network resource scheduling in related technologies are solved.

Description

Technical Field

[0001] The present application relates to the field of big data or other related fields, and specifically, to a method, device, computer program product and electronic device for allocating network resources. Background Art

[0002] With the rapid development of Internet technology and the increasing popularity of digital services, efficient management and reasonable scheduling of network resources have become a key technical field to ensure the performance of online services and user experience. In the current highly interconnected network environment, user behavior presents unprecedented uncertainty and dynamic characteristics, which leads to the ever-changing demand for network resources. For example, online video platforms may experience sudden high bandwidth demand due to a large number of users watching in certain periods of time, while e-commerce platforms may encounter a surge in server load during major promotional activities. The network resource demands of different user groups in different time periods vary significantly, which requires the network resource scheduling system to have a high degree of flexibility and responsiveness.

[0003] However, traditional network resource scheduling methods, especially static resource allocation strategies, are difficult to meet the above requirements. These methods usually rely on pre-set resource allocation schemes and lack real-time monitoring and dynamic adjustment mechanisms. In practical applications, static allocation strategies often lead to unbalanced resource allocation. For example, some resources may be idle during off-peak hours, while some resources may be overused during peak hours, which not only wastes resources but also affects service response speed and user satisfaction due to resource shortages. In addition, since the predefined scheduling rules do not take into account the complexity and variability of user behavior, such as behavior patterns driven by specific events (holidays, major news events, etc.), the system is often unable to accurately predict and respond to real-time traffic fluctuations, resulting in network congestion or uneven resource allocation, thereby affecting overall network performance and user experience.

[0004] With regard to the technical problems of low resource utilization and poor response capability when scheduling network resources in related technologies, no effective solution has been proposed so far. Summary of the invention

[0005] The main purpose of the present application is to provide a method, device, computer program product and electronic device for allocating network resources to solve the technical problems of low resource utilization and poor responsiveness in network resource scheduling in related technologies.

[0006] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a method for allocating network resources is provided. The method includes: obtaining resource load data of a network platform within a preset time period, and obtaining network behavior data of users who use the network platform within the preset time period; when the network platform is in a first network state, generating a first resource scheduling strategy according to the resource load data and the network behavior data, and sending the first resource scheduling strategy to a client, wherein, when the client receives the first resource scheduling strategy, a resource scheduling operation is performed according to the first resource scheduling strategy, and the first resource scheduling strategy includes resource allocation strategies of multiple dimensions; when the network platform is in a second network state, the first resource scheduling strategy is adjusted to obtain a second resource scheduling strategy, and the second resource scheduling strategy is sent to the client, wherein, when the client receives the second resource scheduling strategy, a resource scheduling operation is performed according to the second resource scheduling strategy, the resource demand of the second network state is higher than the resource demand of the first network state, and the second resource scheduling strategy includes resource allocation strategies of multiple dimensions.

[0007] Furthermore, generating a first resource scheduling strategy based on resource load data and network behavior data includes: obtaining a resource analysis model, wherein the resource analysis model is used to identify user behavior patterns and network resource usage trends of the network platform in different time periods; inputting resource load data and network behavior data into the resource analysis model, and outputting resource demand analysis results, wherein the resource demand analysis results include user behavior patterns and network resource usage trends within a preset time period; obtaining user distribution data and resource constraints of the network platform, determining a first optimization target based on the user distribution data and resource constraints, processing the resource demand analysis results based on the first optimization target, and generating a first resource scheduling strategy, wherein the resource constraints are used to indicate restrictions that need to be met in resource scheduling.

[0008] Furthermore, the resource analysis model is trained in the following manner: obtaining historical network behavior data and historical resource load data of the network platform within a historical time period, correlating the historical network behavior data with the historical resource load data to obtain correlated data, wherein the correlation method is to match the resource load conditions of the historical network behavior data with the historical resource load data at different time points within the historical time period; extracting feature data of the correlated data to obtain a historical feature set, wherein the historical feature set includes at least one of the following: user access record information, resource usage data, and peak time; obtaining historical behavior results within the historical time period, and training a preset resource analysis model based on the historical feature set and the historical behavior results to obtain a resource analysis model.

[0009] Furthermore, adjusting the first resource scheduling strategy to obtain the second resource scheduling strategy includes: obtaining a resource scheduling model, and using the resource scheduling model to adjust the first resource scheduling strategy to obtain a transition resource scheduling strategy; using a genetic algorithm to adjust the transition resource scheduling strategy to obtain N candidate resource scheduling sub-strategies, where N is a positive integer; screening out a target resource scheduling sub-strategy from the N candidate resource scheduling sub-strategies, constructing a resource scheduling structure according to the target resource scheduling sub-strategy, and using the resource scheduling structure to adjust the transition resource scheduling strategy to obtain the second resource scheduling strategy.

[0010] Furthermore, the first resource scheduling strategy is adjusted using a resource scheduling model to obtain a transition resource scheduling strategy, including: obtaining user demand data, calculating the matching degree between the user demand data and the first resource scheduling strategy, and obtaining scheduling deviation information, wherein the scheduling deviation information includes at least one of the following: resource type, expected allocation quantity, actual demand quantity, and matching deviation value; obtaining resource rules, inputting the resource rules and scheduling deviation information into a dynamic scheduling model, and outputting resource scheduling results; analyzing the resource scheduling results using a network resource analysis tool to generate a first resource allocation node, wherein the first resource allocation node is used to indicate path node information that needs to be adjusted in the network platform, and the path node information includes at least one of the following: server nodes, storage device nodes, and bandwidth allocation nodes; and adjusting the first resource scheduling strategy according to the resource allocation node using a resource optimization algorithm to obtain a transition resource scheduling strategy.

[0011] Furthermore, selecting a target resource scheduling sub-strategy from N candidate resource scheduling sub-strategies includes: obtaining a second optimization target, and determining the weight of each resource indicator in the second optimization target to obtain Y resource indicator weights, wherein the second optimization target includes at least one of the following: a resource utilization indicator and a response time indicator, and Y is a positive integer; for a candidate resource scheduling sub-strategy, obtaining Y resource scheduling indicator values ​​of the candidate resource scheduling sub-strategy, and calculating a scheme score of the candidate resource scheduling sub-strategy based on the Y resource indicator weights and the Y resource scheduling indicator values ​​to obtain a scheme score; extracting the candidate resource scheduling sub-strategy with the highest score from the N candidate resource scheduling sub-strategies based on the N scheme scores, and determining the candidate resource scheduling sub-strategy with the highest score as the target resource scheduling sub-strategy.

[0012] Furthermore, constructing a resource scheduling structure according to the target resource scheduling sub-strategy includes: determining a second resource allocation node according to the target resource scheduling sub-strategy, wherein the second resource allocation node is used to indicate path node information that needs to be adjusted in the network platform, and the path node information includes at least one of the following: a server node, a storage device node, and a bandwidth allocation node; determining a resource flow state between all nodes of the second resource allocation node through a topology analysis tool, wherein the resource flow state is used to indicate a path for resources to flow from one node to another; obtaining user demand data, and determining the priorities of all nodes in the second resource allocation node according to the user demand data and the target resource scheduling sub-strategy; obtaining resource rules, and using the resource rules to construct a scheduling rule set, wherein the scheduling rule set includes resource allocation conditions, node load thresholds, and scheduling trigger conditions; and constructing a resource scheduling structure based on the scheduling rule set, the resource flow state, the second resource allocation node, and the priorities of all nodes.

[0013] Furthermore, the resource scheduling structure is used to adjust the transition resource scheduling strategy to obtain a second resource scheduling strategy, including: inputting the target resource scheduling sub-strategy into a linear programming model and outputting primary resource allocation information, wherein the linear programming model is configured by the resource scheduling structure, and the primary resource allocation information is used to indicate the allocation scheme of network resources between different nodes; obtaining resource usage data and system performance data of the network platform, and determining resource allocation deviation information based on the resource usage data and the system performance data; performing secondary optimization on the resource allocation deviation information based on an incremental optimization algorithm to obtain secondary resource allocation information, wherein the secondary optimization is used to adjust the distribution of resources between different nodes; and generating a second resource scheduling strategy based on the primary resource allocation information and the secondary resource allocation information.

[0014] In order to achieve the above-mentioned purpose, according to another aspect of the present application, a network resource allocation device is provided. The device includes: an acquisition unit, which is used to acquire resource load data of a network platform within a preset time period, and acquire network behavior data of users who use the network platform within the preset time period; a generation unit, which is used to generate a first resource scheduling strategy according to resource load data and network behavior data when the network platform is in a first network state, and send the first resource scheduling strategy to a client, wherein when the client receives the first resource scheduling strategy, a resource scheduling operation is performed according to the first resource scheduling strategy, and the first resource scheduling strategy includes resource allocation strategies of multiple dimensions; an adjustment unit, which is used to adjust the first resource scheduling strategy when the network platform is in a second network state, obtain a second resource scheduling strategy, and send the second resource scheduling strategy to the client, wherein when the client receives the second resource scheduling strategy, a resource scheduling operation is performed according to the second resource scheduling strategy, the resource demand of the second network state is higher than the resource demand of the first network state, and the second resource scheduling strategy includes resource allocation strategies of multiple dimensions.

[0015] According to another aspect of an embodiment of the present invention, a computer storage medium is provided. The computer storage medium is used to store a program. When the program is executed, the device where the computer storage medium is located is controlled to execute a method for allocating network resources.

[0016] According to another aspect of an embodiment of the present invention, an electronic device is provided, comprising one or more processors and a memory; the memory stores computer-readable instructions, and the processor is used to execute the computer-readable instructions, wherein the computer-readable instructions execute a method for allocating network resources when executed.

[0017] According to another aspect of an embodiment of the present invention, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, a method for allocating network resources is executed.

[0018] Through the present application, the following steps are adopted: obtaining resource load data of a network platform within a preset time period, and obtaining network behavior data of users who use the network platform within the preset time period; when the network platform is in a first network state, generating a first resource scheduling strategy according to the resource load data and the network behavior data, and sending the first resource scheduling strategy to a client, wherein when the client receives the first resource scheduling strategy, a resource scheduling operation is performed according to the first resource scheduling strategy, and the first resource scheduling strategy includes resource allocation strategies of multiple dimensions; when the network platform is in a second network state, adjusting the first resource scheduling strategy to obtain a second resource scheduling strategy, and sending the second resource scheduling strategy to the client, wherein when the client receives the second resource scheduling strategy, a resource scheduling operation is performed according to the second resource scheduling strategy, the resource demand of the second network state is higher than the resource demand of the first network state, and the second resource scheduling strategy includes resource allocation strategies of multiple dimensions, which solves the technical problems of low resource utilization and poor response capability in network resource scheduling in related technologies, generates the first resource scheduling strategy by using resource load data and network behavior data, and then performs one-time and secondary resource allocation on the first resource scheduling strategy, and finally generates and executes the second resource scheduling strategy, thereby achieving the technical effect of improving network resource utilization and response efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0020] Figure 1 is a flow chart of a method for allocating network resources according to an embodiment of the present application;

[0021] Figure 2 is a schematic diagram of a network resource allocation system provided according to an embodiment of the present application;

[0022] Figure 3 is a schematic diagram of an optional method for allocating network resources provided according to an embodiment of the present application;

[0023] Figure 4 is a schematic diagram of a network resource allocation device provided according to an embodiment of the present application;

[0024] Figure 5 is a schematic diagram of an electronic device provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0025] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0026] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0027] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0028] It should be noted that the relevant information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for display, data for analysis, etc.) involved in this disclosure are all information and data authorized by the user or fully authorized by all parties. For example, an interface is set between this system and the relevant user or organization. Before obtaining relevant information, it is necessary to send an acquisition request to the aforementioned user or organization through the interface, and obtain relevant information after receiving the consent information fed back by the aforementioned user or organization.

[0029] It should be noted that the collected information used in this application is information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data comply with the relevant laws, regulations and standards of the relevant regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or refuse to use it.

[0030] The present invention is described below in conjunction with preferred implementation steps. Figure 1 is a flow chart of a method for allocating network resources according to an embodiment of the present application, such as Figure 1 As shown, the method comprises the following steps:

[0031] Step S101, obtaining resource load data of a network platform within a preset time period, and obtaining network behavior data of users who use the network platform within the preset time period.

[0032] Faced with the increasingly complex demands in network resource scheduling, the limitations of traditional scheduling technology have become particularly obvious. In order to achieve efficient utilization of network resources, improve service responsiveness and user satisfaction, and reduce costs and performance losses caused by over-allocation or shortage of resources, we can first obtain the resource load data of the network platform within a preset time period, as well as the network behavior data of users using the network platform within the corresponding time period.

[0033] Among them, network behavior data can be data such as web pages visited by users, links clicked, applications used, access time, usage duration, data traffic generated within a preset time period, etc. obtained through technologies such as log recording systems and user tracking systems, and then the collected behavior data is preprocessed to obtain network behavior data; resource load data can refer to the usage of resources such as network servers, storage devices, bandwidth, etc., such as the CPU usage rate, memory occupancy, disk read and write speed, etc. obtained by deploying server monitoring tools for real-time monitoring, or using analyzers to collect and analyze network traffic data within a preset time period (which may include upload and download speeds, traffic peak time, etc.), and can also be the usage of storage devices recorded through storage management systems (which may include storage space occupancy, read and write operation frequency, data transmission speed, etc.), and network platforms can refer to a certain application platform. By analyzing the information of user behavior, the usage of network resources within a specific time period is determined. For example, when users intensively access the resources of a website within a certain time period, it may cause the server load of the website to increase. Through load monitoring tools such as server monitoring systems and bandwidth monitoring systems, real-time collection and analysis are performed to generate resource load data.

[0034] It should be noted that after obtaining the above-mentioned data, the collected raw data needs to be cleaned to remove outliers, missing values ​​and irrelevant information to ensure the accuracy and reliability of the data; at the same time, the data needs to be converted into a unified format for subsequent analysis.

[0035] Step S102: When the network platform is in a first network state, a first resource scheduling strategy is generated based on resource load data and network behavior data, and the first resource scheduling strategy is sent to the client. When the client receives the first resource scheduling strategy, a resource scheduling operation is performed according to the first resource scheduling strategy. The first resource scheduling strategy includes resource allocation strategies in multiple dimensions.

[0036] Specifically, the first network state may refer to a simple, low-dynamic change or relatively stable resource demand scenario. In this state, the first resource scheduling strategy can be determined directly based on the resource load data and the network behavior data. The resource load data and the network behavior data can be input into a resource analysis model for big data analysis to generate a resource demand analysis result. The first resource scheduling strategy is then determined based on the resource demand analysis result. The strategy may include resource allocation strategies in multiple dimensions, for example, resource allocation strategies and priority settings. For example, according to the resource demand analysis results, it is decided to give priority to critical services during peak hours, allocate more servers and bandwidth resources to the video platform, and reduce the resource priority of some non-critical services.

[0037] Step S103, when the network platform is in the second network state, adjust the first resource scheduling strategy to obtain the second resource scheduling strategy, and send the second resource scheduling strategy to the client, wherein, when the client receives the second resource scheduling strategy, perform resource scheduling operations according to the second resource scheduling strategy, the resource demand of the second network state is higher than the resource demand of the first network state, and the second resource scheduling strategy includes resource allocation strategies in multiple dimensions.

[0038] Specifically, the second network state may refer to a complex, highly dynamically changing network environment. In this state, directly adopting the above-mentioned first resource scheduling strategy cannot meet the real-time resource demand, and the first resource scheduling strategy needs to be adjusted, such as performing local resource optimization operations, so as to obtain the second resource scheduling strategy.

[0039] For example, a large video streaming platform faces a surge in user visits during peak hours. For example, from 8 to 10 o'clock every day, the platform's video playback volume suddenly surges, resulting in excessive server load and insufficient bandwidth. During the above peak period, the CPU (Central Processing Unit) utilization rate of some server nodes exceeded 90%, and the bandwidth utilization rate was close to saturation, thus affecting the smoothness of video playback and user experience. In order to optimize resource utilization and improve response speed, it is necessary to obtain resource load data and network behavior data, among which network behavior data can include video playback volume, user stay time, click behavior, geographic location information, etc. Resource load data can include server CPU utilization, bandwidth usage, storage space occupancy, etc.

[0040] At this time, after obtaining the above data, the first resource scheduling strategy can be determined using resource load data and network behavior data, and then the above scheduling strategy can be re-optimized using dynamic scheduling models and other methods to obtain the second resource scheduling strategy. After generating the second resource scheduling strategy, the client first deploys the strategy to the actual network environment through the automated configuration management tool, and deploys it to various resource nodes, such as servers, storage devices, and bandwidth allocation nodes. As the deployment of the second resource scheduling strategy is completed, network resource scheduling begins, coordinating and managing resource allocation and scheduling operations between various resource nodes, and dynamically allocating computing resources, storage resources, and network bandwidth according to the priorities, resource flows, and scheduling rules defined in the optimization plan.

[0041] The network resource allocation method provided in the embodiment of the present application obtains resource load data of a network platform within a preset time period, and obtains network behavior data of users who use the network platform within the preset time period; when the network platform is in a first network state, generates a first resource scheduling strategy according to the resource load data and the network behavior data, and sends the first resource scheduling strategy to a client, wherein, when the client receives the first resource scheduling strategy, a resource scheduling operation is performed according to the first resource scheduling strategy, and the first resource scheduling strategy includes resource allocation strategies of multiple dimensions; when the network platform is in a second network state, the first resource scheduling strategy is adjusted to obtain a second resource scheduling strategy, And send the second resource scheduling strategy to the client, wherein, when the client receives the second resource scheduling strategy, the resource scheduling operation is performed according to the second resource scheduling strategy, the resource demand of the second network state is higher than the resource demand of the first network state, and the second resource scheduling strategy includes resource allocation strategies in multiple dimensions, which solves the technical problems of low resource utilization and poor response capability in network resource scheduling in related technologies, generates the first resource scheduling strategy by utilizing resource load data and network behavior data, and then performs one-time and secondary resource allocation on the first resource scheduling strategy, and finally generates and executes the second resource scheduling strategy, thereby achieving the technical effect of improving network resource utilization and response efficiency.

[0042] Optionally, in the network resource allocation method provided in the embodiment of the present application, generating a first resource scheduling strategy based on resource load data and network behavior data includes: obtaining a resource analysis model, wherein the resource analysis model is used to identify user behavior patterns and network resource usage trends of the network platform in different time periods; inputting resource load data and network behavior data into the resource analysis model, and outputting resource demand analysis results, wherein the resource demand analysis results include user behavior patterns and network resource usage trends within a preset time period; obtaining user distribution data and resource constraints of the network platform, determining a first optimization target based on the user distribution data and resource constraints, processing the resource demand analysis results based on the first optimization target, and generating a first resource scheduling strategy, wherein the resource constraints are used to indicate restriction conditions that need to be met in resource scheduling.

[0043] Specifically, the resource analysis model can be based on machine learning algorithms, such as the Long Short-Term Memory network (LSTM), which is obtained by training a large amount of historical data, and can identify user behavior patterns and network resource usage trends of the network platform in different time periods. The model can process the input data and identify user behavior patterns during the analysis process. For example, it can identify that certain user groups frequently access certain resources within a specific time period, or that user behavior changes under specific external conditions, such as holidays and events. At the same time, the usage trend of network resources can also be analyzed, such as how the number of visits to a certain type of resource, such as video streaming, changes over time, or the peak time of network traffic, and the user behavior analysis results can be obtained through the above analysis.

[0044] In order to generate the first resource scheduling strategy, resource load data and network behavior data can first be input into the resource analysis model to analyze user behavior patterns and network resource usage trends to obtain user behavior analysis results, and then generate resource demand analysis results based on the user behavior analysis results, wherein the resource demand analysis results may include resource demand prediction and resource load assessment. Resource demand prediction is to infer which resources will be used in large quantities in the future by speculating on user behavior trends; resource load assessment is to assess how these predicted resource demands will affect the existing network resource load. For example, if the demand for video streaming services will increase significantly in a certain period of time in the future, then the corresponding network resources, such as servers and bandwidth, will also face higher load pressure.

[0045] Furthermore, it is possible to determine, based on the above resource demand analysis results, through a heuristic algorithm, such as a genetic algorithm, a particle swarm optimization algorithm, to optimally allocate resources to different services or users, that is, to obtain a first resource scheduling strategy, wherein the first resource scheduling strategy may include scheduling strategies in two dimensions: a resource allocation strategy and a priority setting strategy, and the priority setting may be determined by a multi-objective optimization technique based on the importance of the business, the priority of the user, and the degree of resource tension, to ensure that in the case of insufficient resources, key businesses or users can obtain resources first. This embodiment uses a resource analysis model to identify user behavior patterns and resource usage trends, and then generates a first resource scheduling strategy that meets the optimization goal, thereby achieving efficient use and dynamic adjustment of resources, and improving network performance and user experience.

[0046] Optionally, in the network resource allocation method provided in the embodiment of the present application, the resource analysis model is trained in the following manner: obtaining historical network behavior data and historical resource load data of the network platform within a historical time period, associating the historical network behavior data with the historical resource load data to obtain associated data, wherein the association method is to match the resource load conditions of the historical network behavior data with the historical resource load data at different time points within the historical time period; extracting feature data of the associated data to obtain a historical feature set, wherein the historical feature set includes at least one of the following: user access record information, resource usage data and peak time; obtaining historical behavior results within the historical time period, and training a preset resource analysis model based on the historical feature set and the historical behavior results to obtain a resource analysis model.

[0047] Specifically, in order to improve the accuracy of the analysis, we can first obtain historical network behavior data and historical resource load data, and obtain historical behavior results for historical time periods. In order to analyze the relationship between historical network behavior data and historical resource load data, we can associate these two types of data, that is, match the user's specific behavior with the resource usage at the corresponding time point to obtain associated data. Through association analysis, we can understand how user behavior affects the use of network resources in a specific time period. For example, when a large number of users access a certain video streaming service, how the CPU and bandwidth load of the server change. Among them, historical network behavior data may include access records, click-through rates, dwell time information, page jump paths, and other information of multiple users in the historical time period; and historical resource load data include server CPU usage, bandwidth occupancy, storage space usage, etc.

[0048] Furthermore, data preprocessing is performed on historical network behavior data and historical resource load data, and feature extraction is performed to obtain a historical feature set, where the historical feature set includes user access frequency, resource usage pattern and peak time; based on the historical feature set and historical behavior results, a big data analysis model is constructed.

[0049] Then, a resource analysis model is constructed based on the extracted historical feature set and historical behavior results, wherein the resource analysis model can be trained based on machine learning algorithms, such as decision trees, random forests, deep learning models, etc., to identify patterns and regularities in historical data. For example, the random forest model captures the nonlinear relationship between complex user behavior and resource load by integrating multiple decision trees, thereby improving the accuracy of prediction. During the model training process, the parameters and structure of the model are continuously adjusted according to the input historical feature set to improve the prediction accuracy and generalization ability of the model, and finally the construction of the big data analysis model is completed. The trained resource analysis model can effectively predict future user behavior patterns and resource requirements, and support intelligent scheduling of network resources. This embodiment trains the resource analysis model, so that the resource requirements of the network platform in different time periods in the future can be accurately predicted, including the usage of computing resources, storage resources and network bandwidth, to ensure that user needs can be accurately matched when resources are allocated, reduce resource waste, improve resource utilization efficiency, avoid over-configuration of resources, thereby effectively controlling operating costs, and providing a key prediction basis for intelligent resource scheduling.

[0050] Optionally, in the network resource allocation method provided in an embodiment of the present application, adjusting the first resource scheduling strategy to obtain the second resource scheduling strategy includes: obtaining a resource scheduling model, and using the resource scheduling model to adjust the first resource scheduling strategy to obtain a transition resource scheduling strategy; using a genetic algorithm to adjust the transition resource scheduling strategy to obtain N candidate resource scheduling sub-strategies, where N is a positive integer; screening out a target resource scheduling sub-strategy from the N candidate resource scheduling sub-strategies, constructing a resource scheduling structure based on the target resource scheduling sub-strategy, and using the resource scheduling structure to adjust the transition resource scheduling strategy to obtain a second resource scheduling strategy.

[0051] Specifically, after obtaining the first field scheduling strategy, if the network platform is in a complex, highly dynamically changing network environment, it is necessary to adjust the first resource scheduling strategy to obtain a second resource scheduling strategy. The resource scheduling model can be constructed based on complex mathematical models and machine learning algorithms, such as deep reinforcement learning, multi-objective optimization models, etc., and can intelligently adjust the resource allocation strategy while considering multiple constraints (such as resource type, geographical distribution, and business priority). Therefore, the first resource scheduling strategy can be input into the resource scheduling model, so that the model can perform multi-objective optimization based on the current network status, user behavior prediction, and resource constraints, adjust the resource allocation ratio and priority setting in the strategy, and obtain a transitional resource scheduling strategy that aims to balance the supply and demand of resources and improve the rationality and efficiency of resource allocation.

[0052] Furthermore, the genetic algorithm can be used to further adjust the transition resource scheduling strategy. Through the evolution process of the genetic algorithm, multiple candidate resource scheduling sub-strategies are generated. Among them, the genetic algorithm solves complex multi-objective optimization problems by simulating the process of natural selection. That is, in each generation, the algorithm selects according to the evaluation results of each sub-strategy (such as resource utilization, system response time, user satisfaction and other indicators), generates new sub-strategy combinations through crossover operations, and then introduces randomness through mutation to ensure the comprehensiveness of the search space. After multiple rounds of iterations, it gradually converges to multiple candidate resource scheduling sub-strategies.

[0053] Specifically, we first define multiple optimization objectives, such as resource utilization, response time, etc. Then the genetic algorithm generates a population from the transition resource scheduling strategy. Each individual represents a candidate resource scheduling sub-strategy. Through selection, crossover and mutation operations, the algorithm gradually optimizes these candidate resource scheduling sub-strategies in each generation, so that individuals with higher fitness occupy a dominant position in the population. The fitness function is calculated based on the weighted sum of each optimization objective, reflecting the comprehensive performance of each solution. The fitness function is F(x) represents the fitness value of the candidate resource scheduling sub-strategy x. The larger the value, the better the sub-strategy. U(x) is the resource utilization rate, which represents the weighted average resource utilization rate of the candidate resource scheduling sub-strategy x on all resource nodes. T(x) is the response time, which represents the average response time of the candidate resource scheduling sub-strategy x when processing user requests. w1 and w2 are pre-set weights. After multiple generations of evolution, the genetic algorithm generates multiple candidate resource scheduling sub-strategies.

[0054] Then, the candidate resource scheduling sub-strategies are evaluated. For example, a multi-objective optimization method can be used to comprehensively consider indicators such as resource utilization, cost, service quality, and user experience, and a comprehensive score can be calculated for each sub-strategy. Based on the scoring results, the sub-strategy with the best performance is selected as the target resource scheduling sub-strategy. Finally, a network topology analysis method can be used to construct a resource scheduling structure that defines the resource allocation method and scheduling rules between different nodes based on the target resource scheduling sub-strategy. The resource scheduling structure can include resource allocation, bandwidth management, storage device optimization, etc. of the server cluster, which can characterize the distribution of current network resources, the connection relationship between nodes, and the resource flow.

[0055] Furthermore, the transition resource scheduling strategy is fine-tuned based on the resource scheduling structure, and then a second resource scheduling strategy is generated, for example, dynamically adjusting the priority of resources, reallocating specific types of resources (such as increasing server resources for computing-intensive tasks, reducing bandwidth allocation during non-peak hours), and optimizing the distribution of network traffic to ensure that resource allocation not only meets demand but also achieves maximum efficiency. This embodiment uses a resource scheduling model and a genetic algorithm to achieve refined adjustment of resource allocation strategies, ensure reasonable allocation of resources for different services, user groups, and geographical locations, avoid resource waste and over-configuration, and enable the network platform to quickly respond to changes in user demand and fluctuations in the network environment, especially during peak hours or when resources are tight, and to quickly adjust resource allocation to ensure the normal operation of key services and user experience.

[0056] Optionally, in the network resource allocation method provided in the embodiment of the present application, the first resource scheduling strategy is adjusted using a resource scheduling model to obtain a transition resource scheduling strategy, including: obtaining user demand data, calculating the matching degree between the user demand data and the first resource scheduling strategy, and obtaining scheduling deviation information, wherein the scheduling deviation information includes at least one of the following: resource type, expected allocation amount, actual demand amount, and matching deviation value; obtaining resource rules, inputting the resource rules and scheduling deviation information into a dynamic scheduling model, and outputting resource scheduling results; using a network resource analysis tool to analyze the resource scheduling results and generate a first resource allocation node, wherein the first resource allocation node is used to indicate path node information that needs to be adjusted in the network platform, and the path node information includes at least one of the following: server nodes, storage device nodes, and bandwidth allocation nodes; using a resource optimization algorithm to adjust the first resource scheduling strategy according to the resource allocation node to obtain a transition resource scheduling strategy.

[0057] Specifically, before using the resource scheduling model to adjust the first resource scheduling strategy, the resource scheduling model can be first constructed by checking the matching degree between the first resource scheduling strategy and the user demand, that is, after obtaining the user demand data, the matching degree between the first resource scheduling strategy and the user demand data is calculated by a data query statement or a big data processing framework to obtain the scheduling deviation information. Specifically, firstly, the data set of the first resource scheduling strategy and the access and demand data set of the actual user are extracted from the database, and then the difference between the resource allocation in the first resource scheduling strategy and the user demand data is calculated using a difference analysis algorithm, and then the deviation between the two is quantified, thereby generating the scheduling deviation information, and the scheduling deviation information may include a timestamp, a resource type, an expected allocation amount, an actual demand amount, and a calculated matching deviation value.

[0058] Furthermore, based on the scheduling deviation information, a dynamic scheduling determination method determined by a rule engine (i.e., resource rules) is used to construct a dynamic scheduling model, and then the resource rules and scheduling deviation information are input into the dynamic scheduling model, and the resource scheduling result is output, and then the resource allocation node that needs to be optimized is located based on the resource scheduling result, that is, the first resource allocation node is obtained, wherein the rule engine (i.e., resource rules) can determine whether the scheduling plan needs to be adjusted according to the size and type of the scheduling deviation by defining a set of business rules. For example, when the resource load of a node exceeds a set threshold, the rescheduling rule is triggered. If it is determined that adjustment is required, the resource usage and user needs are continuously monitored, and real-time data is collected through real-time monitoring tools. Once the deviation between the actual resource usage and the scheduling plan is monitored to exceed the preset threshold, a response is immediately made according to the rule. When a rule is triggered, the corresponding adjustment strategy is executed, including reallocating resources, switching service nodes, increasing resource redundancy, etc.

[0059] When using a network resource analysis tool to analyze resource scheduling results, a network topology analysis tool can be used to analyze nodes and connection relationships in the network, identify nodes with higher loads or lower resource utilization efficiency in the current scheduling plan, and then combine the resource flow analysis tool to determine the resource allocation and usage path in the network, so as to more accurately locate the first resource allocation node that needs to be adjusted, wherein the first resource allocation node is used to indicate path node information, and the path node information includes at least one of the following: server node, storage device node, and bandwidth allocation node.

[0060] Furthermore, the resource optimization algorithm is used to adjust the first resource scheduling strategy according to the resource allocation node to obtain a transition resource scheduling strategy, wherein the resource optimization algorithm can be a heuristic algorithm, and the heuristic algorithm can be a simulated annealing algorithm. These algorithms are suitable for complex optimization problems. Through random search and evolutionary strategies, they can quickly find a near-optimal resource allocation solution. For example, the genetic algorithm first generates a set of initial solutions (i.e., the first resource scheduling strategy), and then generates a better solution through selection, crossover, and mutation operations. After each iteration, the fitness of each solution, i.e., the effectiveness of resource allocation, is calculated, and the optimal solution is retained for the next generation of evolution. The first resource allocation node is rescheduled through a heuristic scheduling algorithm, and the first resource scheduling strategy is adjusted and locally optimized to obtain a transition resource scheduling strategy, wherein the goal of local optimization is to improve the overall network performance through more refined adjustments in a specific resource allocation node or a local range, and to find the optimal resource configuration method by using a local search algorithm (such as a gradient descent method or a hill climbing algorithm) to optimize and search in a local range. For example, in a server cluster, the load distribution of each server is fine-tuned by the gradient descent method to optimize the overall resource utilization. During the local optimization process, the node load and network traffic are monitored by real-time monitoring tools, and the resource scheduling strategy is adjusted in real time to cope with instantaneous demand changes. This embodiment achieves intelligent and efficient resource allocation by carefully adjusting and optimizing the first resource scheduling strategy, combining the advantages of intelligent models and genetic algorithms, thereby improving network performance and user experience, while effectively controlling operating costs.

[0061] Optionally, in the network resource allocation method provided in an embodiment of the present application, screening out a target resource scheduling sub-strategy from N candidate resource scheduling sub-strategies includes: obtaining a second optimization target, and determining the weight of each resource indicator in the second optimization target, to obtain Y resource indicator weights, wherein the second optimization target includes at least one of the following: a resource utilization indicator and a response time indicator, and Y is a positive integer; for a candidate resource scheduling sub-strategy, obtaining Y resource scheduling indicator values ​​of the candidate resource scheduling sub-strategy, and calculating a scheme score of the candidate resource scheduling sub-strategy based on the Y resource indicator weights and the Y resource scheduling indicator values ​​to obtain a scheme score; extracting the candidate resource scheduling sub-strategy with the highest score from the N candidate resource scheduling sub-strategies based on the N scheme scores, and determining the candidate resource scheduling sub-strategy with the highest score as the target resource scheduling sub-strategy.

[0062] Specifically, after generating multiple candidate resource scheduling sub-strategies, these strategies can be evaluated. For example, a weighted evaluation method can be used to score each strategy. First, the second optimization target can be obtained, which can include resource utilization indicators (such as server CPU utilization, memory utilization, bandwidth occupancy), response time indicators (such as the average response time of user requests), cost control (such as total operating costs), service quality indicators (such as the clarity and smoothness of video streaming), user experience indicators (such as user satisfaction scores), etc. Then, weights are assigned to each resource indicator in the second optimization target, such as resource utilization and response time, where these weights can be set based on business needs or policy priorities.

[0063] Then, according to the performance of each candidate resource scheduling sub-strategy on each indicator, a weighted scheme score is calculated. That is, after determining the resource scheduling index value of each candidate resource scheduling sub-strategy, the score of the candidate resource scheduling sub-strategy is calculated according to the weight of the resource index and the individual resource scheduling index value, thereby obtaining a scheme score. For example, a scheme has a high score of 85 points in resource utilization, and the weight of resource utilization is 0.35. At this time, the score of the strategy on the resource utilization index is 30.25. Finally, the scheme scores of all candidate resource scheduling sub-strategies are compared, and the strategy with the highest score is selected and determined as the target resource scheduling sub-strategy. This embodiment can avoid the imbalance of system performance caused by the optimization of a single indicator by using a weighted method to determine the target resource scheduling sub-strategy, so as to more accurately reflect the comprehensive advantages and disadvantages of the strategy, optimize resource allocation, and reduce overall operating costs.

[0064] Optionally, in the network resource allocation method provided in an embodiment of the present application, constructing a resource scheduling structure according to a target resource scheduling sub-strategy includes: determining a second resource allocation node according to the target resource scheduling sub-strategy, wherein the second resource allocation node is used to indicate path node information that needs to be adjusted in the network platform, and the path node information includes at least one of the following: a server node, a storage device node, and a bandwidth allocation node; determining the resource flow state between all nodes of the second resource allocation node through a topology analysis tool, wherein the resource flow state is used to indicate a path for resources to flow from one node to another; obtaining user demand data, and determining the priorities of all nodes in the second resource allocation node according to the user demand data and the target resource scheduling sub-strategy; obtaining resource rules, and using the resource rules to construct a scheduling rule set, wherein the scheduling rule set includes resource allocation conditions, node load thresholds, and scheduling trigger conditions; and constructing a resource scheduling structure based on the scheduling rule set, the resource flow state, the second resource allocation node, and the priorities of all nodes.

[0065] Specifically, after obtaining the target resource scheduling sub-strategy, a resource scheduling structure that characterizes the distribution of current network resources, the connection relationship between nodes, and the resource flow direction can be constructed based on the strategy and using a network topology analysis tool. First, the second resource allocation node can be determined according to the target resource scheduling sub-strategy, wherein the second resource allocation node can also include a server node, a storage device node, and a bandwidth allocation node, the server node is responsible for processing computing tasks, the storage device node is responsible for data storage and access, and the bandwidth allocation node manages the allocation of network traffic; then the network topology analysis tool defines the connection relationship and resource flow direction between the second resource allocation nodes, wherein the network topology analysis tool is to construct and analyze the nodes and their connection relationships in the network through a topology analysis tool. In this process, the dependency relationship between each node can be identified, and the path for resources to flow from one node to another can be determined. For example, the server node is connected to the storage device through a high-speed link, and the bandwidth allocation node is responsible for allocating network traffic between different servers.

[0066] Furthermore, the priority of each resource allocation node is assigned according to the resource demand (that is, user demand data) and the target resource scheduling sub-strategy. That is, after clarifying the resource flow, the analytic hierarchy process is used to assign a priority to each second resource allocation node. The analytic hierarchy process (AHP) is a multi-criteria decision-making method that can handle complex priority allocation problems. First, the importance factors of resource nodes (such as mission criticality, response time, resource utilization, etc.) are resolved into multi-level evaluation criteria, and then the weight of each criterion is calculated by the paired comparison method, and the priority of each node is calculated comprehensively. For example, the computing power of the server node, the network delay of the bandwidth node, and the data reading speed of the storage node are comprehensively compared to finally determine the priority of each node.

[0067] In order to effectively manage resource scheduling, a rule engine (i.e., resource rules) can be used to build a scheduling rule set, wherein the scheduling rule set includes resource allocation conditions, node load thresholds, and scheduling trigger conditions. The rule engine can automatically handle complex decisions in resource scheduling based on a predefined rule set: first, a set of resource allocation conditions, such as bandwidth utilization, CPU utilization, etc., are defined to determine under what circumstances resources should be reallocated, and then a load threshold is set for each node. For example, when the CPU utilization exceeds 80%, a load balancing operation is triggered, and finally, a scheduling trigger condition is set. For example, when the load of a certain node exceeds the set threshold or the network traffic suddenly increases, resource allocation is automatically adjusted according to the preset rules to ensure that resources can be reasonably used in various situations. Finally, the scheduling rule set, resource flow status, second resource allocation node, and the priority of all nodes can be integrated to form a resource scheduling structure. This embodiment forms a resource scheduling structure through resource allocation node identification, resource flow status analysis, priority setting, and the construction of a scheduling rule set to ensure the reasonable flow and allocation of resources between different nodes, thereby improving the resource utilization efficiency, responsiveness, and user experience of the network platform, while controlling operating costs.

[0068] Optionally, in the network resource allocation method provided in the embodiment of the present application, the resource scheduling structure is used to adjust the transition resource scheduling strategy to obtain the second resource scheduling strategy, including: inputting the target resource scheduling sub-strategy into a linear programming model, and outputting primary resource allocation information, wherein the linear programming model is obtained by configuring the resource scheduling structure, and the primary resource allocation information is used to indicate the allocation scheme of network resources between different nodes; obtaining resource usage data and system performance data of the network platform, and determining resource allocation deviation information based on the resource usage data and the system performance data; performing secondary optimization on the resource allocation deviation information based on an incremental optimization algorithm to obtain secondary resource allocation information, wherein the secondary optimization is used to adjust the distribution of resources between different nodes; and generating a second resource scheduling strategy based on the primary resource allocation information and the secondary resource allocation information.

[0069] Specifically, after obtaining the resource scheduling structure, a one-time resource allocation can be performed, that is, the currently available resources are allocated to each node and task at one time according to the optimal strategy through the static resource allocation algorithm, that is, the target resource scheduling sub-strategy is input into the linear programming model for one-time resource allocation to ensure that the initial resource configuration can meet the needs of most services. At the same time, according to the strategy defined in the resource scheduling structure, the resource requirements of each node or service are calculated, and resources are allocated according to the priority and allocation conditions to form a one-time resource allocation information. It should be noted that in the one-time resource allocation, in order to maximize resource utilization or minimize total cost, the objective function of resource allocation can be defined first, where the objective function can be expressed as Among them, U i is the resource utilization, C j is the cost, x i and j They represent the amount of resources allocated to the i-th node and the cost of the j-th node respectively. Then, the requirements of each resource allocation node, such as server, storage device and bandwidth allocation node, are input into the linear programming model as constraints. Finally, by solving this model, the optimal allocation plan of resources between nodes is calculated to form a one-time resource allocation information.

[0070] After completing the one-time resource allocation, it enters the running state. You can use monitoring tools to track key performance indicators such as CPU usage, memory usage, bandwidth utilization, response time, etc., and then obtain the resource usage data and system performance data of the network platform during the one-time allocation process, as well as the real-time monitoring results of resource consumption. Through this information, you can identify the resource bottlenecks or imbalanced utilization that may exist in the one-time allocation, and then perform secondary resource allocation based on this information, that is, use the deviation analysis method to evaluate the bottlenecks and efficiency problems of resource allocation, and obtain resource allocation deviation information. Among them, the deviation analysis identifies the deficiencies in resource allocation by comparing the actual resource usage with the expected value in the one-time resource allocation information, that is, comparing whether the actual load of each node matches the allocated resources, and finding those nodes with excessive or insufficient resource allocation. For example, the CPU usage of a server may be far beyond expectations, while the resource utilization of another server is low. Through this analysis, resource allocation deviation information is generated.

[0071] In order to further optimize resource utilization, an incremental optimization algorithm is used to adjust these imbalances or inefficiencies, such as reallocating bandwidth, adjusting server load, etc., so as to achieve more refined resource utilization and higher service quality. That is, secondary optimization is performed based on resource allocation deviation information to adjust the distribution of resources among nodes to obtain secondary resource allocation information. Finally, the primary resource allocation information and the secondary resource allocation information are integrated by weighted average method and other methods to generate a second resource scheduling strategy. Among them, incremental optimization, as a gradual adjustment method, can gradually improve system performance through small resource adjustments. When using this algorithm, first identify the resource allocation nodes that need to be optimized, that is, identify the nodes that are bottlenecks due to insufficient initial allocation, or the nodes with excess resources that are not fully utilized, and then use the incremental optimization algorithm to reallocate the resources of these nodes. For example, if the load of some nodes is too high, reduce the resource allocation of other low-load nodes, and reallocate these resources to high-load nodes to balance the load of each node, and then generate secondary resource allocation information through adjustment. This embodiment adjusts the transition resource scheduling strategy by utilizing the resource scheduling structure to obtain the second resource scheduling strategy, thereby realizing self-learning and optimization of the scheduling strategy, improving the intelligence and flexibility of the strategy, and achieving the technical effect of improving network resource utilization and response efficiency.

[0072] The present application also provides a network resource allocation system. Figure 2 It is a schematic diagram of a network resource allocation system provided according to an embodiment of the present application, the system comprising: a real-time information acquisition module, a resource demand analysis module, a transition resource scheduling strategy acquisition module, a second resource scheduling strategy determination module and a network resource scheduling module, wherein the real-time information acquisition module is used to acquire network behavior data in real time and determine resource load data; the resource demand analysis module is used to perform big data analysis based on network behavior data and resource load data, generate network resource demand analysis results, and determine a first resource scheduling strategy; the transition resource scheduling strategy acquisition module is used to construct a dynamic scheduling judgment model by checking the matching degree between the first resource scheduling strategy and user demand data, locate resource allocation nodes and perform secondary scheduling and local resource optimization to obtain a transition resource scheduling strategy; the second resource scheduling strategy determination module is used to optimize the transition resource scheduling strategy, determine the target resource scheduling sub-strategy and construct a scheduling structure, perform one-time resource allocation and secondary resource allocation based on policy internal information, and determine the second resource scheduling strategy; the network resource scheduling module is used to execute network resource scheduling based on the second resource scheduling strategy.

[0073] Furthermore, the system can also input network behavior data and resource load data into the big data analysis model, analyze user behavior patterns and network resource usage trends, and obtain user behavior analysis results; generate network resource demand analysis results based on the user behavior analysis results, and the analysis results include resource demand prediction and resource load evaluation; determine the first resource scheduling strategy based on the network resource demand analysis results, and the first resource scheduling strategy includes resource allocation strategy and priority setting. By obtaining historical user behavior data and historical network resource load data, the historical user behavior data corresponds to the historical network resource load data, and the historical behavior results corresponding to the historical user behavior data and the historical network resource load data are obtained; the historical user behavior data and the historical network resource load data are preprocessed, and feature extraction is performed to obtain a historical feature set, which includes user access frequency, resource usage pattern and peak time; based on the historical feature set and historical behavior results, a big data analysis model is constructed.

[0074] Furthermore, the matching degree between the first resource scheduling strategy and the user demand data is calculated to obtain scheduling deviation information; based on the scheduling deviation information, a dynamic scheduling determination method determined by a rule engine is used to construct a dynamic scheduling determination model to locate resource allocation nodes; the resource allocation nodes are rescheduled by a heuristic scheduling algorithm, the first resource scheduling strategy is adjusted, and a preliminary optimized resource scheduling plan is obtained; the preliminary optimized resource scheduling plan is locally optimized to obtain a transitional resource scheduling strategy. Based on the transitional resource scheduling strategy, the resource allocation strategy is optimized using a multi-objective optimization algorithm to obtain multiple candidate resource allocation schemes; multiple candidate resource allocation schemes are evaluated to determine the first optimal resource allocation strategy; based on the first optimal resource allocation strategy, a resource scheduling structure is constructed. Based on the first optimal resource allocation strategy, resource allocation nodes are determined, and the resource allocation nodes include servers, storage devices, and bandwidth allocation nodes; a network topology analysis method is used to define the connection relationship and resource flow between resource allocation nodes; according to resource demand and the first optimal resource allocation strategy, the priority of each resource allocation node is assigned; a scheduling rule set is constructed, and the rule set includes resource allocation conditions, node load thresholds, and scheduling trigger conditions; resource allocation nodes, resource flows, priorities, and scheduling rules are integrated to form a resource scheduling structure. Use a static resource allocation algorithm to perform a one-time resource allocation on the first optimal resource allocation strategy to obtain one-time resource allocation information; collect resource usage and system performance data, use a deviation analysis method to evaluate the bottleneck and efficiency problems of resource allocation, and obtain resource allocation deviation information; based on the resource allocation deviation information, perform secondary optimization to adjust the distribution of resources among nodes and obtain secondary resource allocation information; combine the one-time resource allocation information with the secondary resource allocation information to generate a second resource scheduling strategy.

[0075] Figure 3is a schematic diagram of an optional network resource allocation method provided according to an embodiment of the present application, and the optional network resource allocation method is applied to the above-mentioned network resource allocation system, such as Figure 3 As shown, the method includes: acquiring network behavior data in real time and determining resource load data; performing big data analysis based on the network behavior data and resource load data, generating network resource demand analysis results, and determining a first resource scheduling strategy; constructing a dynamic scheduling judgment model by checking the matching degree between the first resource scheduling strategy and the user demand data, locating the resource allocation node and performing secondary scheduling and local resource optimization to obtain a transition resource scheduling strategy; optimizing the transition resource scheduling strategy, determining the target resource scheduling sub-strategy and constructing a scheduling structure, performing one-time resource allocation and secondary resource allocation based on internal information of the strategy, and determining a second resource scheduling strategy; and executing network resource scheduling based on the second resource scheduling strategy.

[0076] This embodiment can better manage the allocation and flow of resources between different nodes through the first optimal resource allocation strategy based on the optimized scheduling structure, avoid bottlenecks in resource allocation, ensure that resource allocation not only meets current needs but also reaches the optimal state, and maximize resource utilization efficiency; secondary resource allocation based on internal information of the strategy: the secondary allocation adjusts the first optimal resource allocation strategy in real time according to the actual operating conditions and resource usage feedback, to ensure that resource allocation can dynamically respond to changes in the network environment and user needs.

[0077] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0078] The embodiment of the present application also provides a network resource allocation device. It should be noted that the network resource allocation device of the embodiment of the present application can be used to execute the network resource allocation method provided by the embodiment of the present application. The network resource allocation device provided by the embodiment of the present application is introduced below.

[0079] Figure 4 is a schematic diagram of a network resource allocation device provided according to an embodiment of the present application, such as Figure 4 As shown, the device includes: an acquisition unit 40, a generation unit 41, and an adjustment unit 42.

[0080] The acquisition unit 40 is used to acquire resource load data of the network platform within a preset time period, and to acquire network behavior data of users who use the network platform within the preset time period;

[0081] A generating unit 41 is configured to generate a first resource scheduling strategy according to resource load data and network behavior data when the network platform is in a first network state, and send the first resource scheduling strategy to the client, wherein when the client receives the first resource scheduling strategy, a resource scheduling operation is performed according to the first resource scheduling strategy, and the first resource scheduling strategy includes resource allocation strategies of multiple dimensions;

[0082] The adjustment unit 42 is used to adjust the first resource scheduling strategy when the network platform is in a second network state, obtain a second resource scheduling strategy, and send the second resource scheduling strategy to the client, wherein, when the client receives the second resource scheduling strategy, a resource scheduling operation is performed according to the second resource scheduling strategy, the resource demand of the second network state is higher than the resource demand of the first network state, and the second resource scheduling strategy includes resource allocation strategies in multiple dimensions.

[0083] The network resource allocation device provided in the embodiment of the present application obtains the resource load data of the network platform within a preset time period through the acquisition unit 40, and obtains the network behavior data of the users who use the network platform within the preset time period; the generation unit 41 generates a first resource scheduling strategy according to the resource load data and the network behavior data when the network platform is in a first network state, and sends the first resource scheduling strategy to the client, wherein, when the client receives the first resource scheduling strategy, the resource scheduling operation is performed according to the first resource scheduling strategy, and the first resource scheduling strategy includes resource allocation strategies of multiple dimensions; the adjustment unit 42 adjusts the first resource scheduling strategy when the network platform is in a second network state, and obtains A second resource scheduling strategy is generated, and the second resource scheduling strategy is sent to the client, wherein, when the client receives the second resource scheduling strategy, the resource scheduling operation is performed according to the second resource scheduling strategy, the resource demand of the second network state is higher than the resource demand of the first network state, and the second resource scheduling strategy includes resource allocation strategies in multiple dimensions, which solves the technical problems of low resource utilization and poor response capability in network resource scheduling in related technologies, generates the first resource scheduling strategy by utilizing resource load data and network behavior data, and then performs one-time and secondary resource allocation on the first resource scheduling strategy, and finally generates and executes the second resource scheduling strategy, thereby achieving the technical effect of improving network resource utilization and response efficiency.

[0084] Optionally, in the network resource allocation device provided in the embodiment of the present application, the generation unit 41 includes: a first acquisition module, used to acquire a resource analysis model, wherein the resource analysis model is used to identify user behavior patterns and network resource usage trends of the network platform in different time periods; a first input module, used to input resource load data and network behavior data into the resource analysis model, and output resource demand analysis results, wherein the resource demand analysis results include user behavior patterns and network resource usage trends within a preset time period; a second acquisition module, used to acquire user distribution data and resource constraints of the network platform, determine a first optimization target based on the user distribution data and resource constraints, process the resource demand analysis results based on the first optimization target, and generate a first resource scheduling strategy, wherein the resource constraints are used to indicate restrictions that need to be met in resource scheduling.

[0085] Optionally, in the network resource allocation device provided in the embodiment of the present application, the generation unit 41 includes: a third acquisition module, used to obtain historical network behavior data and historical resource load data of the network platform within a historical time period, and associate the historical network behavior data with the historical resource load data to obtain associated data, wherein the association method is to match the resource load conditions of the historical network behavior data and the historical resource load data at different time points within the historical time period; a first extraction module, used to extract feature data of the associated data to obtain a historical feature set, wherein the historical feature set includes at least one of the following: user access record information, resource usage data and peak time; a fourth acquisition module, used to obtain historical behavior results within the historical time period, and train a preset resource analysis model based on the historical feature set and the historical behavior results to obtain a resource analysis model.

[0086] Optionally, in the network resource allocation device provided in the embodiment of the present application, the adjustment unit 42 includes: a fifth acquisition module, used to acquire a resource scheduling model, and use the resource scheduling model to adjust the first resource scheduling strategy to obtain a transition resource scheduling strategy; a first adjustment module, used to use a genetic algorithm to adjust the transition resource scheduling strategy to obtain N candidate resource scheduling sub-strategies, where N is a positive integer; a screening module, used to screen out a target resource scheduling sub-strategy from the N candidate resource scheduling sub-strategies, construct a resource scheduling structure according to the target resource scheduling sub-strategy, and use the resource scheduling structure to adjust the transition resource scheduling strategy to obtain a second resource scheduling strategy.

[0087] Optionally, in the network resource allocation device provided in the embodiment of the present application, the adjustment unit 42 includes: a sixth acquisition module, used to acquire user demand data, calculate the matching degree between the user demand data and the first resource scheduling strategy, and obtain scheduling deviation information, wherein the scheduling deviation information includes at least one of the following: resource type, expected allocation amount, actual demand amount and matching deviation value; a seventh acquisition module, used to acquire resource rules, input the resource rules and scheduling deviation information into the dynamic scheduling model, and output the resource scheduling result; an analysis module, used to analyze the resource scheduling result using a network resource analysis tool, and generate a first resource allocation node, wherein the first resource allocation node is used to indicate the path node information that needs to be adjusted in the network platform, and the path node information includes at least one of the following: server node, storage device node and bandwidth allocation node; a second adjustment module, used to adjust the first resource scheduling strategy according to the resource allocation node using a resource optimization algorithm to obtain a transition resource scheduling strategy.

[0088] Optionally, in the network resource allocation device provided in the embodiment of the present application, the adjustment unit 42 includes: an eighth acquisition module, used to obtain the second optimization target, and determine the weight of each resource indicator in the second optimization target to obtain Y resource indicator weights, wherein the second optimization target includes at least one of the following: a resource utilization indicator and a response time indicator, and Y is a positive integer; a ninth acquisition module, used to obtain Y resource scheduling indicator values ​​of a candidate resource scheduling sub-strategy for a candidate resource scheduling sub-strategy, and calculate the scheme score of the candidate resource scheduling sub-strategy based on the Y resource indicator weights and the Y resource scheduling indicator values ​​to obtain the scheme score; a second extraction module, used to extract the candidate resource scheduling sub-strategy with the highest score from N candidate resource scheduling sub-strategies based on the N scheme scores, and determine the candidate resource scheduling sub-strategy with the highest score as the target resource scheduling sub-strategy.

[0089] Optionally, in the network resource allocation device provided in the embodiment of the present application, the adjustment unit 42 includes: a first determination module, used to determine the second resource allocation node according to the target resource scheduling sub-strategy, wherein the second resource allocation node is used to indicate the path node information that needs to be adjusted in the network platform, and the path node information includes at least one of the following: server node, storage device node and bandwidth allocation node; a second determination module, used to determine the resource flow state between all nodes of the second resource allocation node through a topology analysis tool, wherein the resource flow state is used to indicate the path of resources flowing from one node to another node; a tenth acquisition module, used to obtain user demand data, and determine the priorities of all nodes in the second resource allocation node according to the user demand data and the target resource scheduling sub-strategy; an eleventh acquisition module, used to obtain resource rules, and use the resource rules to construct a scheduling rule set, wherein the scheduling rule set includes resource allocation conditions, node load thresholds and scheduling trigger conditions; a composition module, used to construct a resource scheduling structure based on the scheduling rule set, resource flow state, the second resource allocation node and the priorities of all nodes.

[0090] Optionally, in the network resource allocation device provided in the embodiment of the present application, the adjustment unit 42 includes: a second input module, used to input the target resource scheduling sub-strategy into a linear programming model, and output primary resource allocation information, wherein the linear programming model is obtained by configuring the resource scheduling structure, and the primary resource allocation information is used to indicate the allocation scheme of network resources between different nodes; a twelfth acquisition module, used to obtain resource usage data and system performance data of the network platform, and determine resource allocation deviation information based on the resource usage data and the system performance data; an optimization module, used to perform secondary optimization on the resource allocation deviation information based on an incremental optimization algorithm to obtain secondary resource allocation information, wherein the secondary optimization is used to adjust the distribution of resources between different nodes; a generation module, used to generate a second resource scheduling strategy based on the primary resource allocation information and the secondary resource allocation information.

[0091] The above-mentioned network resource allocation device includes a processor and a memory. The above-mentioned acquisition unit 40, generation unit 41, adjustment unit 42, etc. are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions.

[0092] The processor includes a kernel, which retrieves the corresponding program unit from the memory. One or more kernels can be set, and the technical problems of low resource utilization and poor response capability in network resource scheduling in related technologies can be solved by adjusting kernel parameters.

[0093] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0094] An embodiment of the present invention provides a computer storage medium, which is used to store a program. When the program is running, the device where the computer storage medium is located is controlled to execute a method for allocating network resources.

[0095] Figure 5 is a schematic diagram of an electronic device provided according to an embodiment of the present application, such as Figure 5 As shown, an embodiment of the present invention provides an electronic device, the electronic device 50 includes a processor, a memory, and a program stored in the memory and executable on the processor, the processor is used to execute computer-readable instructions, wherein the computer-readable instructions execute a method for allocating network resources when they are executed. The device in this article can be a server, a PC, a PAD, a mobile phone, etc.

[0096] The present application also provides a computer program product, including a computer program, which implements the steps of a network resource allocation method in each embodiment of the present application when the computer program is executed by a processor.

[0097] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0098] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0099] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0100] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0101] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0102] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0103] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0104] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0105] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for allocating network resources, characterized in that: include: Obtain resource load data of the network platform within a preset time period, and obtain network behavior data of users who use the network platform within the preset time period; When the network platform is in a first network state, a first resource scheduling strategy is generated according to the resource load data and the network behavior data, and the first resource scheduling strategy is sent to the client, wherein when the client receives the first resource scheduling strategy, a resource scheduling operation is performed according to the first resource scheduling strategy, and the first resource scheduling strategy includes resource allocation strategies of multiple dimensions; When the network platform is in a second network state, the first resource scheduling policy is adjusted to obtain a second resource scheduling policy, and the second resource scheduling policy is sent to the client, wherein when the client receives the second resource scheduling policy, a resource scheduling operation is performed according to the second resource scheduling policy, the resource demand of the second network state is higher than the resource demand of the first network state, and the second resource scheduling policy includes resource allocation strategies in multiple dimensions.

2. The method according to claim 1, characterized in that Generating a first resource scheduling strategy according to the resource load data and the network behavior data includes: Acquire a resource analysis model, wherein the resource analysis model is used to identify user behavior patterns and network resource usage trends of the network platform in different time periods; Inputting the resource load data and the network behavior data into the resource analysis model, and outputting a resource demand analysis result, wherein the resource demand analysis result includes a user behavior pattern and a network resource usage trend within the preset time period; Obtain user distribution data and resource constraints of the network platform, determine a first optimization goal based on the user distribution data and the resource constraints, process the resource demand analysis result based on the first optimization goal, and generate the first resource scheduling strategy, wherein the resource constraints are used to indicate the restriction conditions that need to be met in resource scheduling.

3. The method according to claim 2, characterized in that The resource analysis model is trained in the following way: Acquire historical network behavior data and historical resource load data of the network platform within a historical time period, and associate the historical network behavior data with the historical resource load data to obtain associated data, wherein the association method is to match the resource load conditions of the historical network behavior data with the historical resource load data at different time points within the historical time period; Extracting feature data of the associated data to obtain a historical feature set, wherein the historical feature set includes at least one of the following: user access record information, resource usage data, and peak time; The historical behavior results within the historical time period are obtained, and a preset resource analysis model is trained based on the historical feature set and the historical behavior results to obtain the resource analysis model.

4. The method according to claim 1, characterized in that: The first resource scheduling strategy is adjusted to obtain a second resource scheduling strategy including: Acquire a resource scheduling model, and use the resource scheduling model to adjust the first resource scheduling strategy to obtain a transition resource scheduling strategy; Using a genetic algorithm to adjust the transition resource scheduling strategy to obtain N candidate resource scheduling sub-strategies, where N is a positive integer; A target resource scheduling sub-strategy is selected from the N candidate resource scheduling sub-strategies, a resource scheduling structure is constructed according to the target resource scheduling sub-strategy, and the transition resource scheduling strategy is adjusted using the resource scheduling structure to obtain the second resource scheduling strategy.

5. The method according to claim 4, characterized in that The first resource scheduling strategy is adjusted by using the resource scheduling model to obtain a transition resource scheduling strategy including: Acquire user demand data, calculate the matching degree between the user demand data and the first resource scheduling strategy, and obtain scheduling deviation information, wherein the scheduling deviation information includes at least one of the following: resource type, expected allocation amount, actual demand amount, and matching deviation value; Obtain resource rules, input the resource rules and the scheduling deviation information into a dynamic scheduling model, and output resource scheduling results; Analyze the resource scheduling result by using a network resource analysis tool to generate a first resource allocation node, wherein the first resource allocation node is used to indicate path node information that needs to be adjusted in the network platform, and the path node information includes at least one of the following: a server node, a storage device node, and a bandwidth allocation node; The first resource scheduling strategy is adjusted according to the resource allocation node using a resource optimization algorithm to obtain the transition resource scheduling strategy.

6. The method according to claim 4, characterized in that Screening out a target resource scheduling sub-strategy from the N candidate resource scheduling sub-strategies includes: Obtain a second optimization target, and determine the weight of each resource indicator in the second optimization target to obtain Y resource indicator weights, wherein the second optimization target includes at least one of the following: a resource utilization indicator and a response time indicator, and Y is a positive integer; For a candidate resource scheduling sub-strategy, obtaining Y resource scheduling indicator values ​​of the candidate resource scheduling sub-strategy, and calculating a scheme score of the candidate resource scheduling sub-strategy according to the Y resource indicator weights and the Y resource scheduling indicator values ​​to obtain a scheme score; A candidate resource scheduling sub-strategy with the highest score is extracted from the N candidate resource scheduling sub-strategies according to the scores of the N solutions, and the candidate resource scheduling sub-strategy with the highest score is determined as the target resource scheduling sub-strategy.

7. The method according to claim 4, characterized in that Constructing a resource scheduling structure according to the target resource scheduling sub-strategy includes: Determine a second resource allocation node according to the target resource scheduling sub-strategy, wherein the second resource allocation node is used to indicate path node information that needs to be adjusted in the network platform, and the path node information includes at least one of the following: a server node, a storage device node, and a bandwidth allocation node; Determine the resource flow state between all nodes of the second resource allocation node by using a topology analysis tool, wherein the resource flow state is used to indicate a path for resources to flow from one node to another node; Acquire user demand data, and determine the priorities of all nodes in the second resource allocation node according to the user demand data and the target resource scheduling sub-strategy; Acquire resource rules, and use the resource rules to construct a scheduling rule set, wherein the scheduling rule set includes resource allocation conditions, node load thresholds, and scheduling trigger conditions; The resource scheduling structure is formed based on the scheduling rule set, the resource flow state, the second resource allocation node and the priorities of all nodes.

8. The method according to claim 4, characterized in that The transition resource scheduling strategy is adjusted by using the resource scheduling structure to obtain the second resource scheduling strategy, including: Input the target resource scheduling sub-strategy into a linear programming model, and output primary resource allocation information, wherein the linear programming model is obtained by configuring the resource scheduling structure, and the primary resource allocation information is used to indicate an allocation scheme of network resources between different nodes; Acquiring resource usage data and system performance data of the network platform, and determining resource allocation deviation information according to the resource usage data and the system performance data; Performing secondary optimization on the resource allocation deviation information based on an incremental optimization algorithm to obtain secondary resource allocation information, wherein the secondary optimization is used to adjust the distribution of resources between different nodes; The second resource scheduling strategy is generated according to the primary resource allocation information and the secondary resource allocation information.

9. A network resource allocation device, characterized in that: include: An acquisition unit, used to acquire resource load data of a network platform within a preset time period, and to acquire network behavior data of users using the network platform within the preset time period; A generating unit, configured to generate a first resource scheduling strategy according to the resource load data and the network behavior data when the network platform is in a first network state, and send the first resource scheduling strategy to the client, wherein when the client receives the first resource scheduling strategy, a resource scheduling operation is performed according to the first resource scheduling strategy, and the first resource scheduling strategy includes resource allocation strategies of multiple dimensions; An adjustment unit is used to adjust the first resource scheduling policy when the network platform is in a second network state to obtain a second resource scheduling policy, and send the second resource scheduling policy to the client, wherein when the client receives the second resource scheduling policy, a resource scheduling operation is performed according to the second resource scheduling policy, the resource demand of the second network state is higher than the resource demand of the first network state, and the second resource scheduling policy includes resource allocation strategies in multiple dimensions.

10. An electronic device, characterized in that: It includes one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the network resource allocation method described in any one of claims 1 to 8.

Images