Link determination method, device and storage medium

Abstract

The present application provides a link determination method, device and storage medium, which relate to the field of communications and are used to improve the utilization rate of computing resources. The method includes: obtaining the computing resource demand of the terminal and the available computing resources of the cloud resource pool. According to the computing resource demand and the available computing resources, a first quantitative value is determined. The network resources of each alternative link in a plurality of alternative links are obtained, and the alternative link is the transmission path between the cloud resource pool and the terminal. According to the network resources of each alternative link, a plurality of second quantitative values ​​are determined, and one alternative link corresponds to one second quantitative value. According to the first quantitative value, the plurality of second quantitative values, the first preset weight and the second preset weight, a plurality of target values ​​are determined, and one alternative link corresponds to one target value. According to the plurality of target values, a target link is determined from the plurality of alternative links, and the target link is the alternative link corresponding to the minimum value among the plurality of target values.

Description

Technical Field

[0001] The present application relates to the field of communications, and in particular to a method, device, and storage medium for determining a link. Background Art

[0002] With the development of artificial intelligence, more and more businesses require large amounts of computing resources. Cloud computing uses virtualization technology to build a large-capacity computing resource pool. The computing resource pool distributes computing resources to terminals through network links, so that various businesses can obtain the computing resources they need.

[0003] Currently, computing network systems select network links for computing resource transmission based solely on network resources (such as latency), without considering whether the computing resources in the cloud resource pool meet the computing power requirements of the business. This can result in some cloud resource pools being unable to provide sufficient computing resources, leaving some computing resources unused, leading to insufficient utilization of computing resources and reduced computing resource utilization. Summary of the Invention

[0004] The present application provides a link determination method, device and storage medium for improving the utilization of computing resources.

[0005] To achieve the above objectives, this application adopts the following technical solutions:

[0006] In a first aspect, the present application provides a method for determining a link. The method includes: a server may obtain a computing power resource requirement of a terminal and available computing power resources of a cloud resource pool. The server may determine a first quantitative value based on the computing power resource requirement and the available computing power resources, where the first quantitative value indicates the relationship between the computing power resource requirement and the available computing power resources. The server may then obtain network resources for each of a plurality of candidate links, where the candidate links are transmission paths between the cloud resource pool and the terminal. The server may determine multiple second quantitative values ​​based on the network resources of each candidate link, where each candidate link corresponds to a second quantitative value, and the second quantitative value indicates the relationship between the network resources of the candidate link and the network resources of the plurality of candidate links. The server may then determine multiple target values ​​based on the first quantitative value, the plurality of second quantitative values, a first preset weight, and a second preset weight, where the target values ​​indicate the quality of the candidate links, where each candidate link corresponds to a target value. The server may then determine a target link from the plurality of candidate links based on the plurality of target values, where the target link is the candidate link corresponding to the minimum value among the plurality of target values.

[0007] Optionally, the link determination method may further include: the server may obtain a target computing power requirement of the terminal, where the target computing power requirement is a preset multiple of the computing power resource requirement. The server may obtain a first computing power resource and a second computing power resource from the cloud resource pool, where the first computing power resource is the allocated computing power resource in the cloud resource pool, and the second computing power resource is the total computing power resource in the cloud resource pool. The above-mentioned "obtaining available computing power resources in the cloud resource pool" includes: the server may determine the available computing power resources in the cloud resource pool based on the target computing power requirement, the first computing power resource, and the second computing power resource.

[0008] Optionally, the aforementioned "determining the available computing resources of the cloud resource pool based on the target computing power requirement, the first computing power resource, and the second computing power resource" includes: if the difference between the first computing power resource and the second computing power resource is less than the target computing power requirement, the server may determine the available computing resources of the cloud resource pool to be the difference between the first computing power resource and the second computing power resource. If the difference between the first computing power resource and the second computing power resource is greater than or equal to the target computing power requirement, the server may determine the available computing resources of the cloud resource pool to be the target computing power requirement.

[0009] Optionally, the link determination method may also include: the server may obtain N influencing factors, the N influencing factors include: at least one computing power resource indicator and at least one network resource indicator, and N is a positive integer. The server may construct an N-dimensional matrix based on the N influencing factors, and the N-dimensional matrix includes: N×N elements, the elements correspond to any two influencing factors, and the elements are used to indicate the importance between any two influencing factors. Then, the server may determine N first eigenvalues ​​based on the N-dimensional matrix, and the first eigenvalue is used to indicate the degree of influence of the influencing factor on multiple alternative links. Thereafter, the server may determine a second eigenvalue from the N first eigenvalues, and the second eigenvalue is the largest eigenvalue among the N first eigenvalues. The server may perform a consistency check on the N-dimensional matrix based on the second eigenvalue. If the N-dimensional matrix passes the consistency check, the server may determine the normalized eigenvector corresponding to the second eigenvalue, and the normalized eigenvector includes: a first preset weight and a second preset weight.

[0010] Optionally, the computing power resource indicators include at least one of the following: central processing unit (CPU) resources, memory resources, storage resources, and graphics processing unit (GPU) resources; and the network resource indicators include at least one of the following: latency, jitter, and packet loss rate.

[0011] In a second aspect, the present application provides a link determination device, which includes an acquisition module and a processing module.

[0012] The acquisition module is used to obtain the computing power resource demand of the terminal and the available computing power resources of the cloud resource pool. The processing module is used to determine a first quantitative value based on the computing power resource demand and the available computing power resources, and the first quantitative value is used to indicate the relationship between the computing power resource demand and the available computing power resources. The acquisition module is also used to obtain the network resources of each alternative link in a plurality of alternative links, where the alternative link is the transmission path between the cloud resource pool and the terminal. The processing module is also used to determine a plurality of second quantitative values ​​based on the network resources of each alternative link, where one alternative link corresponds to one second quantitative value, and the second quantitative value is used to indicate the relationship between the network resources of the alternative link and the network resources of the plurality of alternative links. The processing module is also used to determine a plurality of target values ​​based on the first quantitative value, the plurality of second quantitative values, the first preset weight, and the second preset weight, where the target value is used to indicate the quality of the alternative link, and one alternative link corresponds to one target value. The processing module is also used to determine a target link from the plurality of alternative links based on the plurality of target values, where the target link is the alternative link corresponding to the minimum value among the plurality of target values.

[0013] Optionally, the acquisition module is further configured to acquire the terminal's target computing power requirement, where the target computing power requirement is a preset multiple of the computing power resource requirement. The acquisition module is further configured to acquire a first computing power resource and a second computing power resource from the cloud resource pool, where the first computing power resource is the allocated computing power resource in the cloud resource pool, and the second computing power resource is the total computing power resource in the cloud resource pool. The processing module is specifically configured to determine the available computing power resources in the cloud resource pool based on the target computing power requirement, the first computing power resource, and the second computing power resource.

[0014] Optionally, the processing module is specifically configured to determine, if the difference between the first computing power resource and the second computing power resource is less than the target computing power requirement, the available computing power resources of the cloud resource pool to be the difference between the first computing power resource and the second computing power resource. The processing module is specifically configured to determine, if the difference between the first computing power resource and the second computing power resource is greater than or equal to the target computing power requirement, the available computing power resources of the cloud resource pool to be the target computing power requirement.

[0015] Optionally, the acquisition module is further used to obtain N influencing factors, and the N influencing factors include: at least one computing power resource indicator and at least one network resource indicator, where N is a positive integer. The processing module is further used to construct an N-dimensional matrix based on the N influencing factors, and the N-dimensional matrix includes: N×N elements, where the elements correspond to any two influencing factors, and the elements are used to indicate the importance between any two influencing factors. The processing module is also used to determine N first eigenvalues ​​based on the N-dimensional matrix, and the first eigenvalue is used to indicate the degree of influence of the influencing factor on multiple alternative links. The processing module is also used to determine a second eigenvalue from the N first eigenvalues, and the second eigenvalue is the largest eigenvalue among the N first eigenvalues. The processing module is also used to perform a consistency check on the N-dimensional matrix based on the second eigenvalue. The processing module is also used to determine the normalized eigenvector corresponding to the second eigenvalue if the N-dimensional matrix passes the consistency check, and the normalized eigenvector includes: a first preset weight and a second preset weight.

[0016] Optionally, the computing power resource indicators include at least one of the following: central processing unit (CPU) resources, memory resources, storage resources, and graphics processing unit (GPU) resources; and the network resource indicators include at least one of the following: latency, jitter, and packet loss rate.

[0017] In a third aspect, the present application provides a link determination apparatus, comprising: a processor and a memory. The processor and the memory are coupled. The memory is configured to store one or more programs, each of which includes computer-executable instructions. When the link determination apparatus is executed, the processor executes the computer-executable instructions stored in the memory to implement the link determination method described in any possible implementation of the first aspect.

[0018] In a fourth aspect, the present application provides a computer-readable storage medium, which stores instructions. When the instructions are executed on a computer, the computer executes the link determination method described in any possible implementation of the first aspect above.

[0019] In a fifth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, enables the computer to implement the link determination method described in any possible implementation manner in the first aspect.

[0020] In the above solution, the technical problems that can be solved and the technical effects achieved by the link determination device, computer equipment, computer storage medium or computer program product can refer to the technical problems and technical effects solved by the first aspect above, and will not be repeated here.

[0021] The technical solution provided by this application provides at least the following beneficial effects: a server can obtain the computing power resource requirements of a terminal and the available computing power resources of a cloud resource pool. Based on the computing power resource requirements and the available computing power resources, the server can determine a first quantized value, where the first quantized value indicates the relationship between the computing power resource requirements and the available computing power resources. Next, the server can obtain the network resources of each candidate link from a plurality of candidate links, where the candidate link is the transmission path between the cloud resource pool and the terminal. Based on the network resources of each candidate link, the server can determine multiple second quantized values, one for each candidate link, where the second quantized value indicates the relationship between the network resources of the candidate link and the network resources of the plurality of candidate links. In this way, the server can provide a unified abstract description of indicators of different dimensions in the cloud resource pool and network path calculation. Next, the server can determine multiple target values ​​based on the first quantized value, the multiple second quantized values, the first preset weight, and the second preset weight. The target values ​​indicate the quality of the candidate links, with one target value corresponding to each candidate link. Subsequently, the server can determine a target link from the plurality of candidate links based on the multiple target values, where the target link is the candidate link corresponding to the minimum value among the multiple target values. In this way, the server combines the computing power resources of the cloud resource pool and the network resources of the alternative links to determine the optimal link. While selecting a link that meets network requirements such as low latency, it can ensure that the cloud resource pool can meet the computing power requirements of the terminal, fully utilize the computing power resources of the cloud resource pool, and improve the utilization rate of computing power resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present application, and together with the specification are used to explain the principles of the present application, and do not constitute an improper limitation on the present application.

[0023] Figure 1 is a schematic diagram showing the architecture of a link determination system according to an exemplary embodiment;

[0024] Figure 2 is a schematic diagram of the architecture of a computing power network management system according to an exemplary embodiment;

[0025] Figure 3 is a flowchart of a method for determining a link according to an exemplary embodiment;

[0026] Figure 4 is a flow chart showing another method for determining a link according to an exemplary embodiment;

[0027] Figure 5 is a schematic diagram showing a hierarchical structure according to an exemplary embodiment;

[0028] Figure 6is a flow chart showing another method for determining a link according to an exemplary embodiment;

[0029] Figure 7 is a structural block diagram of a link determination device according to an exemplary embodiment;

[0030] Figure 8 is a structural diagram of a link determination device according to an exemplary embodiment;

[0031] Figure 9 The present invention is a conceptual partial view of a computer program product according to an exemplary embodiment. DETAILED DESCRIPTION

[0032] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0033] In this document, the character " / " generally indicates an "or" relationship between the preceding and following objects. For example, A / B can be understood as either A or B.

[0034] The terms “first” and “second” in the description and claims of the present application are used to distinguish different objects rather than to describe a specific order of the objects.

[0035] Furthermore, the terms "including," "having," and any variations thereof, as used in the description of this application are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or modules is not limited to the listed steps or modules, but may optionally include other steps or modules not listed, or may optionally include other steps or modules inherent to the process, method, product, or apparatus.

[0036] Additionally, in the embodiments of this application, words such as "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described in this application as "exemplary" or "for example" should not be construed as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "for example" is intended to present concepts in a concrete manner.

[0037] Before introducing the link determination method of the embodiment of the present application in detail, the implementation environment and application scenarios of the embodiment of the present application are first introduced.

[0038] With the development of artificial intelligence, more and more businesses require large amounts of computing resources. As a result, the industry faces new challenges brought about by the mismatch between computing power demand and supply. The speed of computing power supply cannot keep up with the growth rate of computing power demand.

[0039] Cloud computing uses virtualization technology to establish a large-capacity computing resource pool. The computing resource pool distributes computing resources to terminals through network links so that various businesses can obtain the required computing resources.

[0040] However, traditional Internet Protocol (IP) networks lack the ability to perceive computing resources accurately and provide sufficient computing resources for businesses. Furthermore, current computing network systems select network links for transmitting computing resources based solely on network resources (such as latency), without considering whether the computing resources in the cloud resource pool meet the computing power requirements of the business. This can result in some cloud resource pools being unable to provide sufficient computing resources, leaving some computing resources unused, leading to insufficient utilization of computing resources and reduced computing resource utilization.

[0041] To address the above-mentioned issues, an embodiment of the present application provides a link determination method, comprising: a server may obtain the computing power resource requirements of a terminal and the available computing power resources of a cloud resource pool. The server may determine a first quantitative value based on the computing power resource requirements and the available computing power resources. Next, the server may obtain the network resources of each candidate link from a plurality of candidate links, where the candidate link is the transmission path between the cloud resource pool and the terminal. The server may determine multiple second quantitative values ​​based on the network resources of each candidate link, with each candidate link corresponding to one second quantitative value. In this way, the server may provide a unified abstract description of indicators of different dimensions in the cloud resource pool and network path calculation. Then, the server may determine multiple target values ​​based on the first quantitative value, the multiple second quantitative values, the first preset weight, and the second preset weight. The target values ​​are used to indicate the quality of the candidate links, with each candidate link corresponding to one target value. Thereafter, the server may determine a target link from the plurality of candidate links based on the multiple target values, where the target link is the candidate link corresponding to the minimum value among the multiple target values. In this way, the server combines the computing power resources of the cloud resource pool and the network resources of the alternative links to determine the optimal link. While selecting a link that meets network requirements such as low latency, it can ensure that the cloud resource pool can meet the computing power requirements of the terminal, fully utilize the computing power resources of the cloud resource pool, and improve the utilization rate of computing power resources.

[0042] The implementation environment of the embodiments of the present application is introduced below.

[0043] Figure 1This is a schematic diagram of the architecture of a link determination system according to an exemplary embodiment. The architecture includes: a server 101, a cloud platform 102, and a terminal 103. The server 101 and the cloud platform 102 can communicate wired or wirelessly. The server 101 and the terminal 103 can also communicate wired or wirelessly.

[0044] The server 101 can communicate with the cloud platform 102 and the terminal 103. For example, the server 101 can receive computing resource information from the cloud resource pool of the cloud platform 102, and can also receive computing resource requirements from the terminal 103. Furthermore, the server 101 can process the computing resource information and computing resource requirements. The server 101 can also store the computing resource information and computing resource requirements.

[0045] Server 101 is deployed with a computing power network management system. Figure 2 As shown, Figure 2 This is a schematic diagram of the architecture of a computing power network management system according to an exemplary embodiment. Computing power network management system 200, also known as the computing network brain, includes: an application warehouse unit 201, an application management and orchestration unit 202, an application lifecycle management unit 203, a computing network integrated scheduling unit 204, a multi-cloud management and computing power management unit 205, and a network collaborative orchestration unit 206.

[0046] The application warehouse unit 201 may be used to store application information, and the application information may include: application identification and application data.

[0047] The application management and orchestration unit 202 can be used to store specific deployment information and dependency relationships of applications, and can implement rapid deployment of applications in new environments through orchestrated application templates.

[0048] The application lifecycle management unit 203 can be used to manage the application from requirement collection, programming, testing to usage, and can quickly deliver applications to respond to business needs.

[0049] The multi-cloud management and computing power management unit 205 can be used to obtain computing power resource information of the cloud resource pool through the cloud platform.

[0050] The network coordination and orchestration unit 206 can be used to obtain network resources of the network link through a software defined network (SDN) controller. The network coordination and orchestration unit and the SDN controller can communicate wired or wirelessly. The SDN controller is used to manage the network and store network traffic and network resources.

[0051] The computing and network integration scheduling unit 204 can be used to schedule applications and process computing resource information and network resources. The computing resource integration scheduling unit 204 can select cloud resource pools and network links that meet the needs of the terminal and provide the terminal with application services and resources that meet the needs.

[0052] It should be noted that the server can be a single physical server, or a server cluster consisting of multiple servers. Alternatively, the server cluster can be a distributed cluster. Alternatively, the server can be a cloud server. The embodiments of this application do not limit the specific implementation of the server.

[0053] The cloud platform 102 is used to manage at least one cloud resource pool and store computing resource information in the cloud resource pool. The computing resource information may include: total computing resources, allocated computing resources, and available computing resources.

[0054] A terminal (such as terminal 103) can be a device with transceiver functions. The terminal can be deployed on land, including indoors or outdoors, handheld or vehicle-mounted; it can also be deployed on the water surface (such as ships, etc.); it can also be deployed in the air (for example, on airplanes, balloons, and satellites, etc.). The terminal includes a handheld device, vehicle-mounted device, wearable device, or computing device with wireless communication function. Exemplarily, the terminal can be a mobile phone, a tablet computer, or a computer with wireless transceiver function. The terminal device can also be a virtual reality (VR) terminal device, an augmented reality (AR) terminal device, a wireless terminal in industrial control, a wireless terminal in unmanned driving, a wireless terminal in telemedicine, a wireless terminal in smart grid, a wireless terminal in smart city, a wireless terminal in smart home, etc.

[0055] In an embodiment of the present application, the terminal 103 can send computing resource requirements to the server 101.

[0056] The embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0057] like Figure 3 As shown, a method for determining a link provided in an embodiment of the present application includes:

[0058] S301. The server obtains the computing resource requirements of the terminal.

[0059] In the embodiment of the present application, the computing power resource demand is the demand for computing power resource indicators.

[0060] Among them, the computing power resource indicators include at least one of the following: central processing unit (CPU) resources, memory resources, storage resources, and graphics processing unit (GPU) resources.

[0061] For example, if the computing power resource indicators include: CPU resources, memory resources, and storage resources, then the computing power resource requirements include: the required amount of CPU resources, the required amount of memory resources, and the required amount of storage resources.

[0062] In one possible design, the computing resource requirements can be expressed as D = d (X) express.

[0063] For example, if the computing resource requirements include: CPU resource requirements, memory resource requirements, and storage resource requirements, the computing resource requirements can be expressed as D = {d (CPU) , d (MEM) , d (STO)}, where D is used to represent the computing resource requirements of the terminal, d (cPU) Used to indicate the CPU resource demand of the terminal, d (MEM) Used to indicate the memory resource requirement of the terminal, d (STO) Used to indicate the storage resource requirements of the terminal.

[0064] In one possible implementation, the terminal may send a computing resource requirement to the server, and the server may receive the computing resource requirement from the terminal to obtain the computing resource requirement of the terminal.

[0065] In another possible implementation, the server may receive a first input instruction for inputting the computing resource requirement of the terminal. In response to the first input instruction, the server may obtain the computing resource requirement of the terminal.

[0066] S302: The server obtains available computing resources from the cloud resource pool.

[0067] In the embodiment of the present application, the available computing power resources are the available amount of the computing power resource indicators.

[0068] For example, if the computing power resource indicators include: CPU resources, memory resources, and storage resources, then the available computing power resources include: the available amount of CPU resources, the available amount of memory resources, and the available amount of storage resources.

[0069] In some embodiments, before the server obtains the available computing resources from the cloud resource pool, the server may obtain the target computing resources of the terminal.

[0070] The target computing power requirement is a preset multiple of the computing power resource requirement.

[0071] In the embodiment of the present application, the target computing power requirement can be expressed by Formula 1.

[0072] E=η×D Formula 1.

[0073] Among them, E is used to represent the target computing power requirement of the terminal, η is used to represent the preset multiple, and D is used to represent the computing power resource requirement of the terminal.

[0074] It should be noted that in the embodiments of the present application, the preset multiple is not limited. For example, the preset multiple can be 8. For another example, the preset multiple can be 12. For another example, the preset multiple can be 10.

[0075] For example, if the terminal's computing power resource requirements include: 4-core CPU resources, 8G memory resources, and 100G storage resources, and the preset multiplier is 8, then the target computing power requirements include: 32-core CPU resources, 64G memory resources, and 800G storage resources.

[0076] In an embodiment of the present application, before the server obtains the available computing resources of the cloud resource pool, the server may also obtain the first computing resources and the second computing resources of the cloud resource pool.

[0077] Among them, the first computing power resource is the allocated computing power resource in the cloud resource pool, and the second computing power resource is the total computing power resource in the cloud resource pool.

[0078] In one possible design, the first computing resource can be (X) Indicates, where X is used to represent the computing power resource index. The second computing power resource can be represented by λ (x) express.

[0079] For example, if the computing power resource indicators include: CPU resources, memory resources, and storage resources, the first computing power resource can be expressed as {δ (CPU) , δ (MEM) , δ (STO)}, where δ (CPU) Used to indicate the allocated CPU resources in the cloud resource pool, δ (MEM) Used to represent the allocated memory resources in the cloud resource pool, δ (STO) It is used to represent the allocated storage resources in the cloud resource pool. The second computing power resource can be expressed as {λ (CPU) ,λ (MEM) ,λ (STO)}, where λ (CPU) It is used to represent the total CPU resources of the cloud resource pool, λ (MEM) It is used to represent the total memory resources of the cloud resource pool, (STO)Indicates the total storage resources of the cloud resource pool.

[0080] In one possible implementation, the server may determine the available computing resources of the cloud resource pool based on the target computing power requirement, the first computing power resource, and the second computing power resource.

[0081] In an embodiment of the present application, if the difference between the first computing resource and the second computing resource is less than the target computing resource requirement, the server may determine that the available resources are the difference between the first computing resource and the second computing resource. If the difference between the first computing resource and the second computing resource is greater than or equal to the target computing resource requirement, the server may determine that the available resources are the target computing resource requirement.

[0082] In one possible design, the available computing resources can be expressed by Formula 2.

[0083] σ (X) =min{λ (X) -δ (X) ,η×d (X)} Formula 2.

[0084] Among them, σ (X) It is used to represent the available computing power resources in the cloud resource pool, X is used to represent the computing power resource index, and λ (X) Used to represent the first computing power resource of the cloud resource pool, δ (X) It is used to represent the second computing power resource of the cloud resource pool, η is used to represent the preset multiple, and d (X) Used to indicate the computing resource requirements of the terminal.

[0085] It should be noted that in this embodiment of the present application, if the difference between the first computing power resource and the second computing power resource is less than the target computing power requirement, the cloud resource pool can only provide the terminal with available computing power resources. If the difference between the first computing power resource and the second computing power resource is greater than or equal to the target computing power requirement, the cloud resource pool only needs to provide the terminal with computing power resources equal to the target computing power requirement.

[0086] It is understandable that before the server obtains the available computing power resources of the cloud resource pool, the server can obtain the target computing power demand, which is a preset multiple of the computing power resource demand. The server can also obtain the first computing power resource and the second computing power resource of the cloud resource pool, where the first computing power resource is the allocated computing power resource in the cloud resource pool, and the second computing power resource is the total computing power resource in the cloud resource pool. In this way, the server can determine the remaining computing power resources in the cloud resource pool. Afterwards, the server can determine the available computing power resources based on the target computing power demand, the first computing power resource, and the second computing power resource. In this way, the server calculates the available computing power resources based on the total amount and allocated amount of computing power resources in the cloud resource pool, thereby increasing the accuracy of the data.

[0087] In another possible implementation, the cloud resource pool may send available computing resources to the server, and the server may receive the available computing resources from the cloud resource pool to obtain the available computing resources of the cloud resource pool.

[0088] S303. The server determines a first quantized value based on computing resource requirements and available computing resources.

[0089] The first quantified value is used to indicate the relationship between computing power resource requirements and available computing power resources.

[0090] In a possible design, the first quantized value can be expressed by Formula 3.

[0091]

[0092] Among them, index (X) It is used to represent the first quantitative value of the cloud resource pool, X is used to represent the computing power resource index, > (X) It is used to represent the available computing power resources of the cloud resource pool, η is used to represent the preset multiple, and d (X) Used to indicate the computing resource requirements of the terminal.

[0093] It should be noted that, in the embodiment of this application, index (X) The value range of is [0,1]. If the available computing power resources in the cloud resource pool are sufficient, then σ (X) =η×d (X) , then index (X) =0, indicating that the computing power resource indicator in the cloud resource pool meets the terminal's target computing power requirements, which is the most ideal state.

[0094] In an embodiment of the present application, the first quantization value may include multiple first sub-quantization values, and one computing power resource indicator corresponds to one first sub-quantization value.

[0095] For example, if the computing power resource indicators include: CPU resources, memory resources, and storage resources, the server can determine multiple first sub-quantization values ​​according to Formula 3, and the multiple first sub-quantization values ​​include: the first sub-quantization value index corresponding to the CPU resource (CPU) , the first sub-quantization value index corresponding to the memory resource (MEM) , the first sub-quantization value index corresponding to the storage resource (STO) .

[0096] S304: The server obtains network resources of each candidate link among the multiple candidate links.

[0097] Among them, the alternative link is the transmission path between the cloud resource pool and the terminal.

[0098] In a possible implementation, the SDN controller may send the network resources of each candidate link among the multiple candidate links to the server, and the server may receive the network resources of each candidate link among the multiple candidate links from the SDN controller.

[0099] In an embodiment of the present application, the network resource may be a resource amount of a network resource indicator, where the network resource indicator includes at least one of the following: delay, jitter, and packet loss rate.

[0100] For example, if the network resource indicators include: delay, jitter, and packet loss rate, then the network resources include: the resource amount of delay, the resource amount of jitter, and the resource amount of packet loss rate.

[0101] In one possible design, the network resources of the i-th candidate link can be obtained through R i =r (i，Y) Indicates. Where Y is used to represent network resource indicators.

[0102] For example, if the network resources include: the resource amount of delay, the resource amount of jitter, and the resource amount of packet loss rate, the network resource of the i-th candidate link can be expressed as R i ={r (i，DEL) , r (i，JIL) , r (i，PL)}, where r (i，DEL) The amount of resources used to represent the delay of the i-th candidate link, r (i，JIL) The amount of resources used to represent the jitter of the i-th candidate link, r (i，JIL) The amount of resources used to represent the packet loss rate of the i-th candidate link.

[0103] S305: The server determines a plurality of second quantization values ​​according to the network resources of each candidate link.

[0104] One candidate link corresponds to one second quantized value, and the second quantized value is used to indicate the relationship between the network resources of the candidate link and the network resources of multiple candidate links.

[0105] In a possible implementation, for each candidate link, the server may determine the second quantized value of the candidate link according to the network resources of the candidate link and the network resources of the multiple candidate links, so as to determine multiple second quantized values.

[0106] In one possible design, the second quantized value of the i-th candidate link can be expressed by Formula 4.

[0107]

[0108] Among them, index (i，Y) The second quantized value used to represent the i-th candidate link, r (i，Y)It is used to indicate the network resources of the i-th candidate link, and m is used to indicate the number of candidate links.

[0109] In the embodiment of the present application, the second quantization value may include multiple second sub-quantization values, and one network resource indicator corresponds to one second sub-quantization value.

[0110] For example, if the network resource indicators include: delay, jitter, and packet loss rate, the server can determine multiple second sub-quantization values ​​of the i-th candidate link according to formula 4, and the multiple second sub-quantization values ​​include: the second sub-quantization value index corresponding to the delay (i，DEL) , the second sub-quantization value index corresponding to the jitter (i，JIL) , the second sub-quantization value index corresponding to the packet loss rate (i，PL) .

[0111] S306: The server determines multiple target values ​​according to the first quantization value, multiple second quantization values, the first preset weight, and the second preset weight.

[0112] One candidate link corresponds to one target value, and the target value is used to indicate the quality of the candidate link.

[0113] In a possible implementation, for each candidate link, the server may determine a target value of the candidate link according to the first quantized value, the second quantized value, the first preset weight, and the second preset weight to determine multiple target values.

[0114] In one possible design, the target value can be expressed by Formula 5.

[0115] W i =ω1×index (X) +ω2×index (i，Y) Formula 5.

[0116] Among them, W i It is used to represent the target value of the i-th candidate link, ω1 is used to represent the first preset weight, index (X) is used to represent the first quantized value, ω2 is used to represent the second preset weight, index (i，Y) The second quantized value used to represent the i-th candidate link.

[0117] In the embodiment of the present application, the first weight value is the weight value of the computing resource indicator, and the second weight value is the network resource indicator. The first preset weight may include multiple first sub-weights, with each computing resource indicator corresponding to a first sub-weight, and the second preset weight may include multiple second sub-weights, with each network resource indicator corresponding to a second sub-weight.

[0118] For example, if the computing resource indicators include: CPU resources, memory resources, and storage resources, the multiple first sub-weights include: a first sub-weight ω3 corresponding to CPU resources, a first sub-weight ω4 corresponding to memory resources, and a first sub-weight ω5 corresponding to storage resources. If the network resource indicators include: latency, jitter, and packet loss rate, the multiple second sub-weights include: a second sub-weight ω6 corresponding to latency, a second sub-weight ω7 corresponding to jitter, and a second sub-weight ω8 corresponding to packet loss rate.

[0119] In some embodiments, if the computing resource indicators include: CPU resources, memory resources, storage resources, and the network resource indicators include: latency, jitter, and packet loss rate, then the target value of the i-th candidate link can be expressed by Formula 6.

[0120] W i =ω3×index (CPU) +ω4×index (MEM) +ω5+index (STO) +ω6×index (u，DEL) +ω7×index (i，JIL) +ω8×index (i，PL) Formula 6.

[0121] Among them, W i It is used to represent the target value of the i-th candidate link, ω3 is used to represent the first sub-weight corresponding to the CPU resource, index (CPU) It is used to represent the first sub-quantization value corresponding to the CPU resource, and ω4 is used to represent the first sub-weight corresponding to the memory resource. (MEM) It is used to represent the first sub-quantization value corresponding to the memory resource, ω5 is used to represent the first sub-weight corresponding to the storage resource, index (sTO) It is used to represent the first sub-quantization value corresponding to the storage resource, ω6 is used to represent the second sub-weight corresponding to the delay, index (i，DEL) It is used to represent the second sub-quantization value corresponding to the delay of the i-th candidate link, ω7 is used to represent the second sub-weight corresponding to the jitter, index (i，JIL) It is used to represent the second sub-quantization value corresponding to the jitter of the i-th candidate link, ω8 is used to represent the second sub-weight corresponding to the packet loss rate, index (i，PL) Used to represent the second sub-quantization value corresponding to the packet loss rate of the i-th candidate link.

[0122] S307: The server determines a target link from multiple candidate links according to multiple target values.

[0123] The target link is the candidate link corresponding to the minimum value among multiple target values.

[0124] In a possible implementation, the server may compare multiple target values, determine a minimum value from the multiple target values, and use the candidate link corresponding to the minimum value as the target link.

[0125] It should be noted that in the embodiment of the present application, the greater the available computing power resources of the cloud resource pool, the smaller the first quantitative value corresponding to the cloud resource pool. When the available computing power resources meet the target computing power requirements, the first quantitative value is 0, and no weighted accumulation is generated for the calculation of the target value. The smaller the network resources of the alternative link, the better the alternative link. And the smaller the network resources, the smaller the second quantitative value of the alternative link. In other words, the smaller the first quantitative value, the better the cloud resource pool, and the smaller the second quantitative value, the better the alternative link. The smaller the first quantitative value and the second quantitative value, the smaller the target value, so the alternative link corresponding to the minimum value among multiple target values ​​is selected as the target link.

[0126] In an embodiment of the present application, after the server determines a target link from multiple candidate links, the server can send the identifier of the target link to the cloud resource pool. The cloud resource pool can receive the identifier of the target link from the server and query the target link based on the identifier of the target link. The cloud resource pool can then deliver available computing resources to the terminal via the target link.

[0127] It is understood that the server can obtain the computing power resource requirements of the terminal and the available computing power resources of the cloud resource pool. The server can determine a first quantitative value based on the computing power resource requirements and the available computing power resources. The first quantitative value is used to indicate the relationship between the computing power resource requirements and the available computing power resources. Next, the server can obtain the network resources of each candidate link from multiple candidate links. The candidate link is the transmission path between the cloud resource pool and the terminal. The server can determine multiple second quantitative values ​​based on the network resources of each candidate link, with each candidate link corresponding to a second quantitative value. The second quantitative value is used to indicate the relationship between the network resources of the candidate link and the network resources of the multiple candidate links. In this way, the server can provide a unified abstract description of indicators of different dimensions in the cloud resource pool and network path calculation. The server can then determine multiple target values ​​based on the first quantitative value, the multiple second quantitative values, the first preset weights, and the second preset weights. The target values ​​are used to indicate the quality of the candidate links, with each candidate link corresponding to a target value. The server can then determine a target link from the multiple candidate links based on the multiple target values. The target link is the candidate link corresponding to the minimum value among the multiple target values. In this way, the server combines the computing power resources of the cloud resource pool and the network resources of the alternative links to determine the optimal link. While selecting a link that meets network requirements such as low latency, it can ensure that the cloud resource pool can meet the computing power requirements of the terminal, fully utilize the computing power resources of the cloud resource pool, and improve the utilization rate of computing power resources.

[0128] It should be noted that in an embodiment of the present application, there may be multiple cloud resource pools, and the server may execute S302-S307 for each cloud resource pool to determine target values ​​for multiple candidate links for each cloud resource pool. The server may determine the target path from the multiple candidate paths corresponding to each of the multiple cloud resource pools.

[0129] In some embodiments, as Figure 4 As shown, before the server determines the multiple target values ​​(S306) according to the first quantized value, the second quantized value, the first preset weight, and the second preset weight, the link determination method may further include:

[0130] S401: The server obtains N influencing factors.

[0131] Among them, the N influencing factors include: at least one computing power resource indicator and at least one network resource indicator, and N is a positive integer.

[0132] In a possible implementation, the server obtains at least one computing power resource indicator and at least one network resource indicator, and uses the at least one computing power resource indicator and the at least one network resource indicator as N influencing factors.

[0133] For example, if at least one computing resource indicator includes: CPU resources, memory resources, storage resources, and at least one network resource indicator includes: latency, jitter, and packet loss rate, then the N influencing factors include: CPU resources, memory resources, storage resources, latency, jitter, and packet loss rate.

[0134] S402: The server constructs an N-dimensional matrix based on the N influencing factors.

[0135] The N-dimensional matrix includes N×N elements.

[0136] In the embodiment of the present application, an element corresponds to any two influencing factors, and the element is used to indicate the importance between any two influencing factors.

[0137] For example, if influencing factor 1 is CPU resources, influencing factor 2 is memory resources, and element a 12 , the element is used to indicate the importance between CPU resources and memory resources.

[0138] In a possible implementation, the server may construct an N-dimensional matrix using a pairwise comparison method and a 1-9 comparison scaling method.

[0139] For example, if the N influencing factors include: CPU resources, memory resources, storage resources, latency, jitter, and packet loss rate, the N-dimensional matrix can be expressed as

[0140] In one possible design, the server may use a 1-9 comparison scaling method to assign values ​​to elements in an N-dimensional matrix.

[0141] For example, as shown in Table 1, Table 1 shows the meaning of a 1-9 scale.

[0142] Table 1 Meaning of 1-9 scale

[0143]

[0144] That is, when the scale is 1, it means that element i is equally important to element j. When the scale is 3, it means that element i is slightly more important to element j. When the scale is 5, it means that element i is significantly more important to element j. For the introduction of other scales, please refer to the description of the above scales, which will not be repeated here.

[0145] For example, if N influencing factors include: influencing factor 1 is CPU resources, influencing factor 2 is memory resources, influencing factor 3 is storage resources, influencing factor 4 is latency, influencing factor 5 is jitter, and influencing factor 6 is packet loss rate, the N-dimensional matrix can be expressed as Among them, a 11 It is used to indicate the importance between influencing factor 1 and influencing factor 1. 12 It is used to indicate the importance between influencing factor 1 and influencing factor 2. 13 It is used to indicate the importance between influencing factor 1 and influencing factor 3. For the introduction of other elements, please refer to the description of the above elements and will not be repeated here.

[0146] S403: The server determines N first eigenvalues ​​according to the N-dimensional matrix.

[0147] The first characteristic value is used to indicate the degree of influence of the influencing factor on the multiple candidate links.

[0148] In a possible implementation, the server may calculate N first eigenvalues ​​of an N-dimensional matrix.

[0149] In a possible design, the first eigenvalue can be expressed by Formula 7.

[0150]

[0151] in, It is used to represent the first eigenvalue, A is used to represent an N-dimensional matrix, and E is used to represent the identity matrix.

[0152] S404: The server determines a second eigenvalue from the N first eigenvalues.

[0153] The second eigenvalue is the largest eigenvalue among the N first eigenvalues.

[0154] It should be noted that, in the embodiment of the present application, the second eigenvalue is the largest eigenvalue among the N first eigenvalues, indicating that the influencing factor corresponding to the second eigenvalue is the factor with the greatest impact on the multiple candidate links.

[0155] S405: The server performs a consistency check on the N-dimensional matrix according to the second eigenvalue.

[0156] It should be noted that in the embodiment of the present application, when constructing an N-dimensional matrix, although it can more objectively reflect the importance between two influencing factors, the N-dimensional matrix may be an inconsistent matrix. The eigenvectors of the inconsistent matrix cannot reflect the true weights of each influencing factor. Therefore, it is necessary to perform a consistency check on the N-dimensional matrix.

[0157] In a possible implementation, the server may determine the consistency index of the N-dimensional matrix according to the second eigenvalue.

[0158] In the embodiment of the present application, the consistency index can be expressed by Formula 8.

[0159]

[0160] Among them, CI is used to represent the consistency index. It is used to represent the second eigenvalue, and n is used to represent the dimension of an N-dimensional matrix.

[0161] It should be noted that in the embodiment of the present application, if the consistency index is 0, it means that the N-dimensional matrix has complete consistency; if the consistency index is close to 0, it means that the N-dimensional matrix has satisfactory consistency; the larger the consistency index, the more serious the inconsistency of the N-dimensional matrix.

[0162] Then, the server may determine the random consistency value based on the dimension of the N-dimensional matrix and the target correspondence relationship, where the target correspondence relationship is the correspondence relationship between the dimension of the N-dimensional matrix and the random consistency value.

[0163] For example, as shown in Table 2, Table 2 shows a target correspondence relationship, wherein the target correspondence relationship includes: the dimension of the N-dimensional matrix and the random consistency value, where n is used to represent the dimension of the N-dimensional matrix and RI is used to represent the random consistency value.

[0164] Table 2 Target correspondence

[0165] n 1 2 3 4 5 6 7 8 9 RI 0 0 0.58 0.90 1.12 1.24 1.32 1.41 1.45

[0166] That is, when the dimension of the N-dimensional matrix is ​​1, the random consistency value is 0. When the dimension of the N-dimensional matrix is ​​2, the random consistency value is 0. When the dimension of the N-dimensional matrix is ​​3, the random consistency value is 0.58. For the introduction of other dimensions, please refer to the description of the above dimensions and will not be repeated here.

[0167] Afterwards, the server may determine a consistency ratio based on the consistency index and the random consistency value, and may perform consistency checking on the N-dimensional matrix based on the consistency ratio.

[0168] In the embodiment of the present application, the consistency ratio can be expressed by Formula 9.

[0169]

[0170] Among them, CR is used to represent the consistency ratio, CI is used to represent the consistency index, and RI is used to represent the random consistency value.

[0171] In one possible design, if the consistency ratio is greater than or equal to a preset ratio threshold, the server may determine that the N-dimensional matrix fails the consistency check. The server may then reassign values ​​to the elements in the N-dimensional matrix using a 1-9 comparison scaling method and re-execute steps S403-S405.

[0172] In one possible design, if the consistency ratio is less than a preset ratio threshold, the server may determine that the N-dimensional matrix passes the consistency check and may then execute S406.

[0173] S406: The server determines a normalized eigenvector corresponding to the second eigenvalue.

[0174] The normalized feature vector includes: a first preset weight and a second preset weight.

[0175] For example, if N influencing factors include: influencing factor 1 is CPU resources, influencing factor 2 is memory resources, influencing factor 3 is storage resources, influencing factor 4 is latency, influencing factor 5 is jitter, and influencing factor 6 is packet loss rate, the N-dimensional matrix is ​​represented as If the normalized eigenvector is represented as {w1, w2, w3, w4, w5, w6}, then w1 is used to represent the weight value of influencing factor 1 (CPU resources), w2 is used to represent the weight value of influencing factor 2 (memory resources), w3 is used to represent the weight value of influencing factor 3 (storage resources), w4 is used to represent the weight value of influencing factor 4 (latency), w5 is used to represent the weight value of influencing factor 5 (jitter), and w6 is used to represent the weight value of influencing factor 6 (packet loss rate).

[0176] In a possible implementation, the server may determine an eigenvector corresponding to the second eigenvalue based on the second eigenvalue and the N-dimensional matrix. The server may normalize the eigenvector to determine a normalized eigenvector.

[0177] It should be noted that, in the embodiment of the present application, the calculation of the feature vector can refer to conventional technical methods for calculation, which will not be described in detail here. In the embodiment of the present application, for the normalized feature vector, the normalized feature vector can be obtained by {w1, w2, ..., w n},in, i is used to represent the i-th element of the normalized feature vector.

[0178] It is understood that the server can obtain N influencing factors, where the N influencing factors include at least one computing resource indicator and at least one network resource indicator. Based on the N influencing factors, the server can construct an N-dimensional matrix, where the N-dimensional matrix includes N×N elements, each corresponding to any two influencing factors and indicating the relative importance of any two influencing factors. The server can then determine N first eigenvalues ​​based on the N-dimensional matrix, where the first eigenvalue indicates the degree of influence of the influencing factor on multiple candidate links. The server can then determine a second eigenvalue from the N first eigenvalues, where the second eigenvalue is the largest eigenvalue among the N first eigenvalues. In this way, the influencing factor with the greatest impact on the multiple candidate links can be determined. The server can then perform a consistency check on the N-dimensional matrix based on the second eigenvalue. If the N-dimensional matrix passes the consistency check, the server can determine the normalized eigenvector corresponding to the second eigenvalue, where the normalized eigenvector includes a first preset weight and a second preset weight. This allows the server to determine the first and second preset weights using the Analytic Hierarchy Process (AHP), eliminating the need to randomly set the first and second preset weights, thereby improving the accuracy of determining the optimal link.

[0179] It should be noted that in the embodiment of the present application, the method for determining the first preset weight and the second preset weight described in S401-S406 uses the Analytic Hierarchy Process (AHP). When analyzing practical problems, the AHP decomposes the various factors related to the problem into several hierarchical levels from top to bottom according to different attributes. Each factor in the same level is subordinate to or has an influence on the factors in the previous level, and can also dominate or be affected by the factors in the next level.

[0180] In the embodiments of this application, Figure 5 As shown, Figure 5A schematic diagram of a hierarchical structure is shown. Among them, the target layer includes path calculation decisions. The criterion layer includes: computing power resource indicators and network resource indicators. The computing power resource indicators include: CPU resources, memory resources, storage resources, and network resource indicators include: delay, jitter, and packet loss rate. The solution layer includes multiple cloud resource pools (such as: cloud resource pool 1, cloud resource pool 2). The hierarchical analysis method starts from the second layer of the hierarchy (i.e., the criterion layer), and based on the factors of the same layer that are subordinate to or affect the factors of the upper layer, a paired comparison matrix (i.e., an N-dimensional matrix) is constructed using the paired comparison method and the 1-9 comparison scale until the bottom layer. In an embodiment of the present application, the influencing factors of the second layer include: CPU resources, memory resources, storage resources, delay, jitter, and packet loss rate, so the server can execute S302-S306.

[0181] The following describes the link determination method in this application with a specific embodiment. Figure 6 As shown, the link determination method may include the following steps:

[0182] S601. The server receives a resource request message from the terminal through the computing network brain.

[0183] The resource request message includes the target computing resources.

[0184] It should be noted that in the embodiment of the present application, a station processing unit may also be deployed in the link determination system. The station processing unit may receive the resource request message of the terminal and perform billing. The station processing unit may also send the resource request message of the terminal to the computing network brain. The computing network brain may receive the resource request message of the terminal from the station processing unit.

[0185] S602. The server perceives the computing power resource information of the cloud resource pool and the network resources of each candidate link in the multiple candidate links through the computing network brain.

[0186] Among them, computing power resource information includes: total computing power resources, allocated computing power resources and available computing power resources.

[0187] It should be noted that in the embodiment of the present application, the computing network brain can actively perceive the computing resource information of the cloud resource pool through the northbound interface of the multi-cloud management and computing power management unit through the application program interface (API). The computing network brain can actively perceive the network resources of each candidate link in multiple candidate links through the northbound interface of the network collaboration and orchestration unit.

[0188] S603. The server determines multiple target values ​​through the computing network brain according to the target computing resources, the computing resource information of the cloud resource pool, and the network resources of each candidate link in the multiple candidate links.

[0189] Among them, one target value corresponds to one candidate link.

[0190] S604. The server determines a target link from multiple candidate links according to multiple target values ​​through the computing network brain.

[0191] S605. The server sends a scheduling instruction to the cloud platform through the computing network brain.

[0192] The scheduling instruction is used to instruct the cloud platform to send the target scheduling resources to the terminal through the target link. The target scheduling resources are computing resources equal to the target computing resources.

[0193] In an embodiment of the present application, after the computing network brain sends a scheduling instruction to the cloud platform, the cloud platform can receive the scheduling instruction from the computing network brain. In response to the scheduling instruction, the cloud platform can send the target scheduling resources to the terminal via the target link. After the cloud platform sends the target scheduling resources to the terminal, the server can re-perceive the computing power resource information of the cloud resource pool and the network resources of each of the multiple candidate links through the computing network brain.

[0194] In this way, the server can use the computing network brain to adopt an integrated scheduling algorithm to improve the accuracy of the computing network brain's scheduling decisions. While selecting links that meet network requirements such as low latency, it can ensure that the cloud resource pool can meet the computing power requirements of the terminal, fully utilize the computing power resources of the cloud resource pool, and improve the utilization rate of computing power resources.

[0195] The above mainly introduces the solution provided by the embodiment of the present application from the perspective of the method. In order to realize the above functions, the link determination device or electronic device includes a hardware structure and / or software module corresponding to the execution of each function. Those skilled in the art should easily realize that, in combination with the link determination method steps of each example described in the embodiment disclosed in this application, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in a hardware or computer software driven hardware manner depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0196] The present application also provides a link determination device, which can be a computer device, a CPU in the computer device, a module in the computer device for determining a link, or a client in the computer device for determining a link.

[0197] In the embodiment of the present application, the determination of the link can be divided into functional modules or functional units according to the above method example. For example, each functional module or functional unit can be divided according to each function, or two or more functions can be integrated into one processing module. The above integrated module can be implemented in the form of hardware or in the form of software functional modules or functional units. Among them, the division of modules or units in the embodiment of the present application is schematic and is only a logical functional division. In actual implementation, there may be other division methods.

[0198] like Figure 7 FIG. 1 is a schematic diagram of a structure of a link determination device provided in an embodiment of the present application. The link determination device is used to perform Figure 3 、 Figure 4 and Figure 6 The link determination method shown in FIG.

[0199] Acquisition module 701 is configured to acquire the computing power resource requirements of a terminal and the available computing power resources of a cloud resource pool. Processing module 702 is configured to determine a first quantized value based on the computing power resource requirements and the available computing power resources. The first quantized value indicates the relationship between the computing power resource requirements and the available computing power resources. Acquisition module 701 is also configured to acquire the network resources of each candidate link from a plurality of candidate links. A candidate link is a transmission path between a cloud resource pool and a terminal. Processing module 702 is also configured to determine multiple second quantized values ​​based on the network resources of each candidate link, one second quantized value corresponding to each candidate link, and the second quantized value indicates the relationship between the network resources of the candidate link and the network resources of the plurality of candidate links. Processing module 702 is also configured to determine multiple target values ​​based on the first quantized value, the plurality of second quantized values, a first preset weight, and a second preset weight. The target values ​​indicate the quality of the candidate links, one target value corresponding to each candidate link. Processing module 702 is also configured to determine a target link from the plurality of candidate links based on the plurality of target values. The target link is the candidate link corresponding to the minimum value among the plurality of target values.

[0200] Optionally, acquisition module 701 is further configured to acquire a target computing power requirement of the terminal, where the target computing power requirement is a preset multiple of the computing power resource requirement. Acquisition module 701 is further configured to acquire a first computing power resource and a second computing power resource of the cloud resource pool, where the first computing power resource is the allocated computing power resource in the cloud resource pool, and the second computing power resource is the total computing power resource in the cloud resource pool. Processing module 702 is specifically configured to determine the available computing power resources in the cloud resource pool based on the target computing power requirement, the first computing power resource, and the second computing power resource.

[0201] Optionally, processing module 702 is specifically configured to determine, if the difference between the first computing power resource and the second computing power resource is less than the target computing power requirement, the available computing power resources of the cloud resource pool as the difference between the first computing power resource and the second computing power resource. Processing module 702 is specifically configured to determine, if the difference between the first computing power resource and the second computing power resource is greater than or equal to the target computing power requirement, the available computing power resources of the cloud resource pool as the target computing power requirement.

[0202] Optionally, acquisition module 701 is further configured to acquire N influencing factors, where the N influencing factors include at least one computing resource indicator and at least one network resource indicator, where N is a positive integer. Processing module 702 is further configured to construct an N-dimensional matrix based on the N influencing factors, where the N-dimensional matrix includes N×N elements, where an element corresponds to any two influencing factors and indicates the degree of importance between any two influencing factors. Processing module 702 is further configured to determine N first eigenvalues ​​based on the N-dimensional matrix, where the first eigenvalue indicates the degree of influence of the influencing factor on multiple candidate links. Processing module 702 is further configured to determine a second eigenvalue from the N first eigenvalues, where the second eigenvalue is the largest eigenvalue among the N first eigenvalues. Processing module 702 is further configured to perform a consistency check on the N-dimensional matrix based on the second eigenvalue. Processing module 702 is further configured to determine a normalized eigenvector corresponding to the second eigenvalue if the N-dimensional matrix passes the consistency check, where the normalized eigenvector includes a first preset weight and a second preset weight.

[0203] Optionally, the computing power resource indicators include at least one of the following: central processing unit (CPU) resources, memory resources, storage resources, and graphics processing unit (GPU) resources; and the network resource indicators include at least one of the following: latency, jitter, and packet loss rate.

[0204] Figure 8 8 is a hardware structure diagram of a link determination device according to an exemplary embodiment. The link determination device may include a processor 801, and the processor 801 is configured to execute application code to implement the link determination method of the present application.

[0205] The processor 801 may be a central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits for controlling the execution of the program of the present application.

[0206] like Figure 8 As shown, the link determination device may further include a memory 802. The memory 802 is used to store application code for executing the solution of the present application, and the execution is controlled by the processor 801.

[0207] The memory 802 may be a read-only memory (ROM) or other type of static storage device that can store static information and instructions, a random access memory (RAM) or other type of dynamic storage device that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, an optical disc storage (including a compact disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 802 may exist independently and be connected to the processor 801 via the bus 804. The memory 802 may also be integrated with the processor 801.

[0208] like Figure 8 As shown, the link determination apparatus may further include a communication interface 803, wherein the processor 801, the memory 802, and the communication interface 803 may be coupled to each other, for example, via a bus 804. The communication interface 803 is used to exchange information with other devices, for example, to support information exchange between the link determination apparatus and other devices.

[0209] It should be pointed out that Figure 8 The device structure shown in the figure does not constitute a limitation on the determination device of the link, except Figure 8 In addition to the components shown, the link determination device may include more or fewer components than shown, or combine certain components, or arrange the components differently.

[0210] In actual implementation, the functions implemented by the processing module 702 can be Figure 8 The processor 801 shown calls the program code in the memory 802 to implement it.

[0211] The present application also provides a computer-readable storage medium having instructions stored thereon. When the instructions in the computer-readable storage medium are executed by a processor of a computer device, the computer is enabled to execute the link determination method provided in the above-described embodiment. For example, the computer-readable storage medium may be a memory 802 including instructions, and the instructions may be executed by the processor 801 of the computer device to perform the above-described method. Alternatively, the computer-readable storage medium may be a non-transitory computer-readable storage medium, for example, a ROM, RAM, CD-ROM, magnetic tape, floppy disk, or optical data storage device.

[0212] Figure 9 A conceptual partial view of a computer program product provided by an embodiment of the present application is schematically shown. The computer program product includes a computer program for executing a computer process on a computing device.

[0213] In one embodiment, the computer program product is provided using a signal bearing medium 900. The signal bearing medium 900 may include one or more program instructions that, when executed by one or more processors, may provide the above-described Figure 3 、 Figure 4 and Figure 6 Thus, for example, reference to Figure 3 In the embodiment shown in , one or more features of S301 to S307 may be undertaken by one or more instructions associated with the signal bearing medium 900. In addition, Figure 9 The program instructions in also describe example instructions.

[0214] In some examples, the signal-bearing medium 900 may include a computer-readable medium 901, such as, but not limited to, a hard drive, a compact disk (CD), a digital video disk (DVD), a digital tape, a memory, a read-only memory (ROM), or a random access memory (RAM), and the like.

[0215] In some implementations, signal bearing medium 900 may include computer recordable medium 902 such as, but not limited to, memory, read / write (R / W) CD, R / W DVD, or the like.

[0216] In some embodiments, signal bearing medium 900 may include communication medium 903 such as, but not limited to, digital and / or analog communication media (eg, fiber optic cables, waveguides, wired communication links, wireless communication links, etc.).

[0217] The signal bearing medium 900 may be conveyed by a wireless form of communication medium 903. The one or more program instructions may be, for example, computer executable instructions or logic implemented instructions.

[0218] In some examples, such as for Figure 7 The described link determination apparatus may be configured to provide various operations, functions, or actions in response to one or more program instructions via computer-readable medium 901 , computer-recordable medium 902 , and / or communication medium 903 .

[0219] Through the description of the above implementation methods, technical personnel in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete the full classification or partial functions described above.

[0220] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of modules or units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0221] The units described as separate components may or may not be physically separate, and the components shown as units may be one physical unit or multiple physical units, that is, they may be located in one place or distributed in multiple different places. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0222] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0223] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the technical solution of the embodiment of the present application is essentially or the part that contributes to the prior art or the full classification part or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions to enable a device (which can be a single-chip microcomputer, chip, etc.) or a processor (processor) to execute the full classification part or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a ROM, a RAM, a magnetic disk or an optical disk.

[0224] The above are only specific embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any changes or replacements within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A method for determining a link, characterized in that: The method comprises: Obtain the computing power resource requirements of the terminal and the available computing power resources in the cloud resource pool; Determining a first quantized value based on the computing power resource requirement and the available computing power resources, where the first quantized value is used to indicate a relationship between the computing power resource requirement and the available computing power resources; Acquire a network resource of each candidate link among a plurality of candidate links, where the candidate link is a transmission path between the cloud resource pool and the terminal; Determine, according to the network resources of each candidate link, a plurality of second quantized values, where one candidate link corresponds to one second quantized value, and the second quantized value is used to indicate a relationship between the network resources of the candidate link and the network resources of the plurality of candidate links; Determining multiple target values ​​according to the first quantized value, the multiple second quantized values, the first preset weight, and the second preset weight, where the target values ​​are used to indicate the quality of the candidate links, and each candidate link corresponds to one target value; A target link is determined from the multiple candidate links according to the multiple target values, where the target link is the candidate link corresponding to the minimum value among the multiple target values.

2. The method according to claim 1, characterized in that Before obtaining the available computing resources of the cloud resource pool, the method further includes: Obtaining a target computing power requirement of the terminal, where the target computing power requirement is a preset multiple of the computing power resource requirement; Obtain a first computing resource and a second computing resource of the cloud resource pool, where the first computing resource is the allocated computing resource in the cloud resource pool, and the second computing resource is the total computing resource in the cloud resource pool; Obtaining the available computing resources of the cloud resource pool includes: Determine the available computing power resources of the cloud resource pool based on the target computing power requirement, the first computing power resources, and the second computing power resources.

3. The method according to claim 2, characterized in that The determining, according to the target computing power requirement, the first computing power resource, and the second computing power resource, of the available computing power resources of the cloud resource pool includes: If the difference between the first computing power resources and the second computing power resources is less than the target computing power requirement, determining the available computing power resources of the cloud resource pool to be the difference between the first computing power resources and the second computing power resources; If the difference between the first computing power resources and the second computing power resources is greater than or equal to the target computing power requirement, the available computing power resources of the cloud resource pool are determined to be the target computing power requirement.

4. The method according to any one of claims 1 to 3, characterized in that Before determining a plurality of evaluation values ​​according to the first quantized value, the plurality of second quantized values, the first preset weight, and the second preset weight, the method further includes: Obtain N influencing factors, where the N influencing factors include: at least one computing power resource indicator and at least one network resource indicator, where N is a positive integer; Constructing an N-dimensional matrix based on the N influencing factors, the N-dimensional matrix including: N×N elements, each element corresponding to any two of the influencing factors, and each element being used to indicate the importance between any two of the influencing factors; Determining N first eigenvalues ​​according to the N-dimensional matrix, where the first eigenvalues ​​are used to indicate the degree of influence of the influencing factor on the multiple candidate links; Determine a second eigenvalue from the N first eigenvalues, where the second eigenvalue is the largest eigenvalue among the N first eigenvalues; Performing a consistency check on the N-dimensional matrix according to the second eigenvalue; If the N-dimensional matrix passes the consistency check, a normalized eigenvector corresponding to the second eigenvalue is determined, where the normalized eigenvector includes: the first preset weight and the second preset weight.

5. The method according to claim 4, characterized in that The computing power resource indicators include at least one of the following: central processing unit (CPU) resources, memory resources, storage resources, and graphics processing unit (GPU) resources; the network resource indicators include at least one of the following: latency, jitter, and packet loss rate.

6. A link determination device, characterized in that: The device comprises: The acquisition module is used to obtain the computing power resource requirements of the terminal and the available computing power resources of the cloud resource pool; a processing module, configured to determine a first quantized value based on the computing power resource requirement and the available computing power resources, wherein the first quantized value is used to indicate a relationship between the computing power resource requirement and the available computing power resources; The acquisition module is further configured to acquire network resources of each candidate link among a plurality of candidate links, where the candidate link is a transmission path between the cloud resource pool and the terminal; The processing module is further configured to determine a plurality of second quantized values ​​according to the network resources of each candidate link, wherein each candidate link corresponds to one second quantized value, and the second quantized value is used to indicate a relationship between the network resources of the candidate link and the network resources of the plurality of candidate links; The processing module is further configured to determine a plurality of target values ​​based on the first quantized value, the plurality of second quantized values, the first preset weight, and the second preset weight, wherein the target values ​​are used to indicate the quality of the candidate links, and each candidate link corresponds to one target value; The processing module is further configured to determine a target link from the multiple candidate links according to the multiple target values, wherein the target link is the candidate link corresponding to the minimum value among the multiple target values.

7. The device according to claim 6, characterized in that The acquisition module is further configured to acquire a target computing power requirement of the terminal, where the target computing power requirement is a preset multiple of the computing power resource requirement; The acquisition module is further configured to acquire a first computing resource and a second computing resource of the cloud resource pool, wherein the first computing resource is the allocated computing resource in the cloud resource pool, and the second computing resource is the total computing resource in the cloud resource pool; The processing module is specifically used to determine the available computing power resources of the cloud resource pool based on the target computing power requirement, the first computing power resources and the second computing power resources.

8. The device according to claim 7, characterized in that The processing module is specifically configured to determine, if the difference between the first computing power resources and the second computing power resources is less than the target computing power requirement, that the available computing power resources of the cloud resource pool are the difference between the first computing power resources and the second computing power resources; The processing module is specifically used to determine that the available computing power resources of the cloud resource pool are the target computing power requirement if the difference between the first computing power resources and the second computing power resources is greater than or equal to the target computing power requirement.

9. The device according to any one of claims 6 to 8, characterized in that The acquisition module is further configured to acquire N influencing factors, wherein the N influencing factors include: at least one computing power resource indicator and at least one network resource indicator, where N is a positive integer; The processing module is further configured to construct an N-dimensional matrix based on the N influencing factors, wherein the N-dimensional matrix includes: N×N elements, each element corresponding to any two of the influencing factors, and each element indicating the importance between any two of the influencing factors; The processing module is further configured to determine N first eigenvalues ​​according to the N-dimensional matrix, where the first eigenvalues ​​are used to indicate the degree of influence of the influencing factor on the multiple candidate links; The processing module is further configured to determine a second eigenvalue from the N first eigenvalues, where the second eigenvalue is the largest eigenvalue among the N first eigenvalues; The processing module is further configured to perform a consistency check on the N-dimensional matrix according to the second eigenvalue; The processing module is further configured to determine a normalized eigenvector corresponding to the second eigenvalue if the N-dimensional matrix passes the consistency check, wherein the normalized eigenvector includes: the first preset weight and the second preset weight.

10. The device according to claim 9, characterized in that The computing power resource indicators include at least one of the following: central processing unit (CPU) resources, memory resources, storage resources, and graphics processing unit (GPU) resources; the network resource indicators include at least one of the following: latency, jitter, and packet loss rate.

11. A link determination device, characterized in that: include: processor and memory; The processor is coupled to the memory; The memory is used to store one or more programs, which include computer-executable instructions. When the link determination device is running, the processor executes the computer-executable instructions stored in the memory to enable the link determination device to perform the link determination method as described in any one of claims 1 to 5.

12. A computer-readable storage medium storing instructions, characterized in that: When a computer executes the instruction, the computer executes the link determination method according to any one of claims 1 to 5.

Images