A computing power sensing network and its use method and storage medium

By designing a computing power perception network, the problem of low allocation efficiency of computing power resources under the network computing converged architecture is solved, flexible scheduling of computing power resources and efficient utilization of computing resources are realized, and user experience and network resource utilization are improved. Edge server operators provide a level-to-play opportunity.

CN115002681BActive Publication Date: 2024-11-08CHINA MOBILE COMM LTD RES INST +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202110231080.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-03-02
Publication Date
2024-11-08
Estimated Expiration
2041-03-02

AI Technical Summary

Technical Problem

The existing technology cannot effectively acquire and allocate computing resources under the network architecture of network computing integration, resulting in low computing resource utilization and inability to meet the high and stringent QoE requirements.

Method used

A computing power perception network was designed, including computing power service layer, computing network management layer, computing power resource layer and network resource layer. Through distributed microservice architecture and API gateway, the management, scheduling and billing of computing power resources are realized, and the management, scheduling and billing of computing power resources are blocked. The underlying hardware differences provide computing power models and APIs to realize flexible scheduling of computing power resources and optimization of network resources.

Benefits of technology

It realizes the effective allocation of computing resources and flexible scheduling of computing resources, improves user experience and network resource utilization, provides fair competition opportunities, and maximizes operational value through dynamic rate adjustments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115002681B_ABST
    Figure CN115002681B_ABST
Patent Text Reader

Abstract

The present invention discloses a computing power perception network and its use method and storage medium, including: a computing power service layer, used to carry various services and applications; a computing network management layer, used to receive reported computing power resource information and network resource information, and manage computing power resources and network resources; a computing power resource layer, used to provide computing power resources using computing infrastructure; a computing power routing layer, used to dispatch the services required by the terminal to different computing power nodes on demand. The present invention can provide users with a consistent experience; no longer rely on a unified billing server, and can provide fair competition opportunities for different edge server operators.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technologies, and in particular to a computing power-aware network and a method for using the network and a storage medium thereof. Background Art

[0002] With the development trend of cloud computing and edge computing, in the future society, there will be many different scales of computing power at different distances from users, providing users with various personalized services through the global network. From tens of billions of smart terminals, to billions of home gateways around the world, to thousands of edge clouds with computing capabilities brought by the future MEC (Mobile Edge Computing) technology in each city, and dozens of large cloud DCs (Data Centers) in each country, a large amount of ubiquitous computing power will be connected to the Internet from everywhere, forming a development trend of deep integration of computing and network. The computing resources in the network are integrated into every corner of the network, so that every network node can become a resource provider. The user's request can be satisfied by calling the nearest node resources, no longer limited to a specific node, avoiding the waste of connection and network scheduling resources. The traditional network only provides a pipeline for data communication, based on connection, and is subject to a fixed network addressing mechanism. It often cannot meet the higher and more demanding QoE (Quality of Experience) requirements. Therefore, the design of the new generation of network architecture for future networks needs to coordinately consider the needs of the integrated evolution of networks and computing, and realize the global optimization of the network in ubiquitous connectivity and computing power architecture, flexible scheduling of computing resources, and reasonable distribution of services.

[0003] IaaS (Infrastructure as a Service) refers to a service model that provides IT (Internet Technology) infrastructure as a service through the Internet and charges users based on their actual usage or occupation of resources. In this service model, ordinary users do not need to build a data center or other hardware facilities by themselves, but instead obtain computer infrastructure services, including servers, storage, and network services, from IaaS service providers through leasing through the Internet. Existing IaaS services are usually managed, allocated, and charged by the resource operation management center (cloud computing center) or the billing server for its own resources.

[0004] The shortcoming of the existing technology is that it cannot achieve effective allocation of computing resources under the network architecture of network computing fusion. Summary of the invention

[0005] The present invention provides a computing power-aware network and a method for using the network and a storage medium, so as to solve the problem that computing power resources cannot be effectively obtained under a network architecture of network computing fusion.

[0006] The present invention provides the following technical solutions:

[0007] A computing power-aware network includes: a computing power service layer, a computing network management layer, a computing power resource layer, a computing power routing layer, and a network resource layer, wherein:

[0008] The computing power service layer is used to carry various services and applications;

[0009] The computing network management layer is used to receive reported computing resource information and network resource information and manage computing resource and network resource;

[0010] The computing resource layer is used to provide computing resources using computing infrastructure;

[0011] The computing power routing layer is used to dispatch the services required by the terminal to different computing power nodes on demand based on computing power resources and network resources.

[0012] In implementation, the computing power service layer is based on a distributed microservice architecture.

[0013] During implementation, the computing power service layer is further used to support the deconstruction of applications into atomic functional components, which are then uniformly scheduled by API Gateway.

[0014] During implementation, the computing power service layer is further used to realize service decomposition and service scheduling through the API gateway after receiving the data from the end user.

[0015] During implementation, the computing network management layer is further used to bill the terminals based on their use of computing power.

[0016] During implementation, the computing network management layer is further used to abstractly describe and represent computing resources through computing power modeling, form node computing power information, and shield the differences in underlying hardware devices.

[0017] During implementation, the computing network management layer is further used to transmit node computing power information to network nodes through computing power notifications.

[0018] During implementation, the computing network management layer is further used to monitor and manage the performance of computing resources and network resources to achieve computing power operations and network operations.

[0019] During implementation, the computing power resource layer is further used to provide one or a combination of the following functions to meet the diverse computing needs in the edge computing field based on physical computing resources for different applications: computing power model, computing power API, and computing power resource identification.

[0020] During implementation, the computing power routing layer is further used to dispatch the services required by the terminal to different computing power nodes on demand based on the computing power resources provided by the computing power resource layer and the network resources provided by the network resource layer.

[0021] In implementation, the computing power routing layer is further used to include one or a combination of the following functions:

[0022] Computing network routing identification, computing network routing forwarding, computing power routing addressing, and computing network service announcement.

[0023] During implementation, it further includes:

[0024] The network resource layer is used to use the network infrastructure to provide network connections for various communication devices in the network.

[0025] In implementation, the network resource layer is further used to provide network connectivity for a network infrastructure including one or a combination of the following facilities:

[0026] Access network, metropolitan area network, backbone network.

[0027] A method for using a computing power-aware network, comprising:

[0028] After receiving the computing resource request, the computing network management layer broadcasts the computing resource request to the computing nodes of the computing resource layer;

[0029] Receiving computing power resource information and charging information returned by a computing power node, wherein the computing power node is a computing power node that meets the requirements in the computing power resource request;

[0030] The computing power routing layer establishes a connection between the terminal and the computing power node after determining the computing power node based on the computing power resource information and computing power resource requests returned by each computing power node, so as to schedule the terminal's computing business to be executed on the computing power node.

[0031] During implementation, it further includes:

[0032] Receive computing power node return and charging information;

[0033] The computing power routing layer establishes a connection between the terminal and the computing power node after determining the computing power node based on the computing power resource information, charging information and computing power resource request returned by each computing power node, so as to schedule the computing business of the terminal to be executed on the computing power node;

[0034] The computing network management layer charges the terminal based on the computing resource information and charging information returned by the computing nodes in the computing resource layer.

[0035] During implementation, the computing network management layer receives computing resource requests through the resource operation center; and / or charges the terminal through the resource operation center.

[0036] In implementation, the computing resource request includes the node selection strategy;

[0037] The computing power routing layer establishes a connection between the terminal and the computing power node after the ingress routing determines the computing power node based on the computing power resource information and charging information returned by each computing power node and the node selection strategy in the computing power resource request, so as to schedule the computing business of the terminal to be executed on the computing power node;

[0038] or,

[0039] The computing resource request includes the computing resource size and the maximum acceptable latency;

[0040] The computing power routing layer transmits the computing power resource information and charging information returned by each computing power node to the terminal under the condition of maximum acceptable delay. The computing power node information that meets the requirements will reach the terminal. After the terminal determines the computing power node, a connection between the terminal and the computing power node will be established to schedule the terminal's computing business to the computing power node for execution.

[0041] In implementation, the computing power resource request received by the computing network management layer is sent to the computing network management layer after the ingress router receives the computing power resource request; and / or,

[0042] The computing resources required by the terminal are dispatched to different computing nodes on demand through entry routing.

[0043] A computer-readable storage medium stores a computer program for executing the method for using the computing power-aware network.

[0044] The beneficial effects of the present invention are as follows:

[0045] In the technical solution provided by the embodiment of the present invention, since computing resources are provided to users based on the same computing power-aware network, a consistent experience can be provided to users;

[0046] Furthermore, since multiple computing nodes directly provide all information including billing information to end users before computing resources are allocated, users can make independent choices among multiple computing nodes based on their own needs. Therefore, the billing information of computing power nodes can be determined directly through agreement with users, and no longer relies on a unified billing server, which can provide fair competition opportunities for different edge server operators. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0048] Figure 1This is a schematic diagram of the computing power perception network structure in an embodiment of the present invention;

[0049] Figure 2 A schematic diagram of an implementation flow of a method for using a computing power-aware network in an embodiment of the present invention;

[0050] Figure 3 This is a schematic diagram of the workflow of computing resource operation in the first embodiment of the present invention;

[0051] Figure 4 The figure is a schematic diagram of the workflow of computing resource operation in the second embodiment of the present invention. DETAILED DESCRIPTION

[0052] The inventors noticed during the invention process that:

[0053] The existing computing resource operation mode (IaaS service) is usually that the resource operation management center (cloud computing center) or the billing server directly manages, allocates and bills the resources that belong to it. However, the computing nodes in the network may belong to different owners, and their billing methods are also very different. From the overall point of view, it is difficult to achieve effective allocation of computing resources. Since edge computing nodes are usually resource-constrained and can only provide services to a limited number of users, the operation management center or the billing server needs to monitor the load of the node in real time in order to decide whether to continue to allocate its resources to other users in the future. With the increase in the number of computing nodes and the uncertainty of network latency, the cost of real-time monitoring of node load continues to increase. Therefore, future edge computing operations need to consider how to fairly and effectively allocate computing resources of different owners and achieve effective control and billing of computing node loads.

[0054] Based on this, the embodiments of the present invention provide a new network architecture solution for network computing fusion for the future data communication network architecture oriented to computing network fusion. Furthermore, a solution for computing resource operation is provided based on this new network architecture.

[0055] The specific implementation of the present invention will be described below in conjunction with the accompanying drawings.

[0056] Figure 1 This is a schematic diagram of the computing power-aware network structure. As shown in the figure, the network includes: computing power service layer, computing network management layer, computing power resource layer, computing power routing layer and network resource layer, among which:

[0057] The computing power service layer is used to carry various services and applications;

[0058] The computing network management layer is used to receive reported computing resource information and network resource information and manage computing resource and network resource;

[0059] The computing resource layer is used to provide computing resources using computing infrastructure;

[0060] The computing power routing layer is used to dispatch the services required by the terminal to different computing power nodes on demand based on computing power resources and network resources.

[0061] During implementation, the computing network management layer is further used to bill the terminals based on their use of computing power.

[0062] During implementation, the method may further include:

[0063] The network resource layer is used to use the network infrastructure to provide network connections for various communication devices in the network.

[0064] Specifically, in response to the development trend of computing-network integration, the solution proposes a network architecture - CAN (Computing-aware Networking). Computing-aware Networking is a new network architecture with deep integration of computing networks. It is based on ubiquitous network connections and highly distributed computing nodes. Through automated deployment of services, optimal routing and load balancing, it builds a new network infrastructure that can perceive computing power, ensuring that the network can dispatch computing resources at different locations on demand and in real time, thereby improving the utilization of network and computing resources and further enhancing user experience.

[0065] It is predicted that by the end of 2020, there will be more than 50 billion terminals and devices connected to the Internet, of which more than 50% of the data needs to be analyzed, processed and stored at the edge of the network. The interconnection of all things brings about the needs of massive data, computing, connection, storage, processing, etc. The processing of data requires a lot of computing power, and the model of terminals plus data centers is increasingly unable to cope with data processing. With the acceleration of 5G network construction, it is natural for computing power to spread from the cloud and the end to the edge of the network.

[0066] The computing power-aware network is aimed at the computing power needs of the future intelligent society. It emphasizes the deep integration of computing and networks. It can realize the efficient computing power needs of the cloud, edge, and end, and can also achieve high throughput, agile connection, and balanced on-demand data and computing power. It is currently mainly promoted by operators.

[0067] The digital transformation of the industry brings about a deep integration of the physical world and the digital world, which requires ubiquitous computing power and connections. At the same time, with the development of cloud computing and edge computing, computing power of different sizes will be distributed at different distances from users in the future society, helping computing power to gradually move from the center to the edge, providing the ultimate business experience for vertical industry businesses. In addition, with the lightweight computing carrier and the development trend of applications gradually decomposing into services and functions, the network needs to be more flexibly scheduled. The future network needs to perceive, interconnect and coordinate ubiquitous computing power and services. In order to meet the above future network needs, major operators and equipment manufacturers have proposed new network technologies for computing power networks to cope with the development of computing-network integration, and carried out research on computing power networks and the construction of industrial ecosystems.

[0068] The computing power-aware network interconnects dynamically distributed computing resources based on ubiquitous network connections. Through unified management and coordinated scheduling of multi-dimensional resources such as network, storage, and computing power, massive applications can call computing resources in different locations on demand and in real time, achieve global optimization of connections and computing power in the network, and provide a consistent user experience.

[0069] The computing power-aware network in the embodiment of the present invention refers to a network that enables massive applications to call computing resources in different places on demand and in real time, and realizes global optimization of connections and computing power in the network. However, it should also be noted that this network is still evolving. Therefore, no matter what name it has in other places or scenarios, as long as it has such a capable network architecture, the network that is also suitable for the technical solution provided in the embodiment of the present invention is the computing power-aware network referred to in the embodiment of the present invention.

[0070] For the computing network integrating the new network architecture - computing power-aware network, in order to realize the perception, interconnection and coordinated scheduling of ubiquitous computing and services, the computing power-aware network architecture system can be divided into five functional modules from the perspective of logical functions: computing power service layer, computing network management layer, computing power resource layer, computing power routing layer and network resource layer, such as Figure 1 As shown, the computing power routing layer includes a control plane and a forwarding plane.

[0071] Based on the ubiquitous computing resources on the network, the computing resource layer completes the abstraction, modeling, control and management of computing resources, and notifies the computing routing layer through the computing notification module. The computing routing layer comprehensively considers user needs, computing resource conditions and the underlying network resource status, and schedules service applications to appropriate nodes to achieve optimal resource utilization and ensure the ultimate user experience.

[0072] The following is a detailed implementation description by layer. Similarly, it should be noted that there may be other names in other places and scenarios. Therefore, the various names in the embodiments of the present invention are only for the convenience of description, and the names themselves do not contain technical meanings other than the names. For example, the computing power routing layer only refers to this logical layer, and does not have special "routing" and other technical meanings. Similarly, in other environments, layers with the same functions, regardless of how they are named, are used to dispatch the services required by the terminal to different computing power nodes on demand based on computing power resources and network resources. This is the computing power routing layer in the embodiments of the present invention.

[0073] 1. Computing service layer:

[0074] In implementation, the computing power service layer is based on a distributed microservice architecture.

[0075] Microservice Architecture is an architectural concept that aims to decouple solutions by breaking down functionality into discrete services. Think of it as applying many SOLID principles (SRP: Single Responsibility Principle; OCP: Open Closed Principle; LSP: Liskov Substitution Principle; ISP: Interface Segregation Principle; DIP: Dependency Inversion Principle) to the architectural level rather than the classes that get the services. Microservice architecture is an interesting concept. Its main purpose is to decompose functionality into discrete services, thereby reducing the coupling of the system and providing more flexible service support.

[0076] During implementation, the computing power service layer is further used to support the deconstruction of applications into atomic functional components, which are then uniformly scheduled by API Gateway.

[0077] In specific implementation, the computing power service layer is further used to realize service decomposition and service scheduling through the API gateway after receiving the data from the terminal users.

[0078] Specifically, based on the distributed microservice architecture, the computing service layer supports the deconstruction of applications into atomic functional components, which are uniformly scheduled by the API Gateway (API Gateway, API: Application Program Interface).

[0079] The computing power service layer is deployed on the computing power resource layer to carry various services and applications of ubiquitous computing. It can pass user requests for business SLA (Service Level Agreement) including computing power requests and other parameters to the computing power routing layer. In addition, the computing power service layer can also receive data from end users and implement service decomposition, service scheduling and other functions through the API gateway.

[0080] 2. Network management:

[0081] During implementation, the computing network management layer is further used to abstractly describe and represent computing resources through computing power modeling, form node computing power information, and shield the differences in underlying hardware devices.

[0082] In specific implementations, the computing network management layer is further used to transmit node computing power information to network nodes through computing power notifications.

[0083] During implementation, the computing network management layer is further used to monitor and manage the performance of computing resources and network resources to achieve computing power operations and network operations.

[0084] Specifically, it completes computing power operation and computing power service orchestration, and completes the management of computing power resources and network resources, including the perception, measurement and OAM (operation administration and maintenance) management of computing power resources; realizes computing network operation for end users, and the management of computing power routing layer and network resource layer.

[0085] In the face of heterogeneous computing resources, the computing network management layer first abstractly describes and represents computing resources through computing power modeling to form node computing power information and shield the differences in underlying hardware devices. The computing power information can be transmitted to the corresponding network nodes through computing power notifications.

[0086] In addition, it is also possible to monitor and manage the performance of computing resources and network resources, and realize computing power operation and network operation.

[0087] 3. Computing resource layer:

[0088] During implementation, the computing power resource layer is further used to provide one or a combination of the following functions to meet the diverse computing needs in the edge computing field based on physical computing resources for different applications: computing power model, computing power API, and computing power resource identification.

[0089] Specifically, the existing computing infrastructure is used to provide computing resources. The computing infrastructure includes a combination of computing capabilities ranging from single-core CPU to multi-core CPU, to CPU (central processing unit) + GPU (Graphic Processing Unit) + FPGA (Field-Programmable Gate Array);

[0090] In order to meet the diverse computing needs in the edge computing field and for different applications, we provide computing power models, computing power APIs, computing power resource identification and other functions based on physical computing resources.

[0091] 4. Computing power routing layer:

[0092] During implementation, the computing power routing layer is further used to dispatch the services required by the terminal to different computing power nodes on demand based on the computing power resources provided by the computing power resource layer and the network resources provided by the network resource layer.

[0093] In implementation, the computing power routing layer is further used to include one or a combination of the following functions:

[0094] Computing network routing identification, computing network routing forwarding, computing power routing addressing, and computing network service announcement.

[0095] Specifically, the computing power routing layer can include a control plane and a forwarding plane;

[0096] Based on the abstract computing network resource discovery, the network status and computing resource status are comprehensively considered, and the business is flexibly scheduled to different computing power nodes on demand. The computing power routing layer is the core of the computing power-aware network.

[0097] The specific functions mainly include computing network routing identification, computing network routing forwarding, computing power routing addressing, computing network service notification, etc.

[0098] 5. Network resource layer:

[0099] In implementation, the network resource layer is further used to provide network connectivity for a network infrastructure including one or a combination of the following facilities:

[0100] Access network, metropolitan area network, backbone network.

[0101] Specifically, the existing network infrastructure can be used to provide ubiquitous network connections for every corner of the network. The network infrastructure includes access networks, metropolitan area networks, and backbone networks.

[0102] Among them, the computing power resource layer and the network resource layer are the infrastructure layers of the computing power perception network. The computing network management layer and the computing power routing layer are the two core functional modules for realizing the computing power perception functional system. Users and applications access the network through the computing power routing layer.

[0103] The computing power-aware network architecture is based on the five defined functional modules and realizes the perception, control and scheduling of computing network resources.

[0104] Based on the above computing power perception network, an embodiment of the present invention also provides a method for using the computing power perception network, which is described below.

[0105] Figure 2 A flowchart of a method for using a computing power-aware network is provided, as shown in the figure, which may include:

[0106] Step 201: After receiving the computing resource request, the computing network management layer broadcasts the computing resource request to the computing nodes of the computing resource layer;

[0107] Step 202: receiving computing power resource information and charging information returned by a computing power node, wherein the computing power node is a computing power node that meets the requirements in the computing power resource request;

[0108] Step 203: The computing power routing layer establishes a connection between the terminal and the computing power node after determining the computing power node based on the computing power resource information and computing power resource request returned by each computing power node, so as to schedule the computing business of the terminal to be executed on the computing power node.

[0109] During implementation, the method may further include:

[0110] Receive computing power node return and charging information;

[0111] The computing power routing layer establishes a connection between the terminal and the computing power node after determining the computing power node based on the computing power resource information, charging information and computing power resource request returned by each computing power node, so as to schedule the computing business of the terminal to be executed on the computing power node;

[0112] The computing network management layer charges the terminal based on the computing resource information and charging information returned by the computing nodes in the computing resource layer.

[0113] During the implementation of step 202, the computing resource information and charging information returned by the computing node are received, and other required node information may also be received, such as the address and port of the computing node, and may also include public key node information as needed.

[0114] During the implementation of step 203, the computing power routing layer (which may be a routing node that receives the computing power resource request) establishes a connection between the terminal and the computing power node after determining the computing power node based on the computing power resource information and charging information returned by each computing power node and the node selection strategy in the computing power resource request, so as to schedule the computing business of the terminal to be executed on the computing power node.

[0115] Computing power resource information refers to the size of computing power resources that a computing power node can provide. Computing power resources are physical resources.

[0116] The routing node may specifically be: a routing node that receives a computing power resource request, or a routing node that first receives a computing power resource request, or an entry routing node to which a terminal that initiates a computing power resource request belongs.

[0117] During implementation, the computing network management layer receives computing resource requests through the resource operation center; and / or charges the terminal through the resource operation center.

[0118] In implementation, the computing power resource request received by the computing network management layer is sent to the computing network management layer after the ingress router receives the computing power resource request; and / or,

[0119] The computing resources required by the terminal are dispatched to different computing nodes on demand through entry routing.

[0120] In the specific implementation, at least two methods may be included:

[0121] The computing resource request includes a node selection strategy;

[0122] The computing power routing layer establishes a connection between the terminal and the computing power node after the ingress routing determines the computing power node based on the computing power resource information and charging information returned by each computing power node and the node selection strategy in the computing power resource request, so as to schedule the computing business of the terminal to be executed on the computing power node;

[0123] or,

[0124] The computing resource request includes the computing resource size and the maximum acceptable latency;

[0125] The computing power routing layer transmits the computing power resource information and charging information returned by each computing power node to the terminal under the condition of maximum acceptable delay. The computing power node information that meets the requirements will reach the terminal. After the terminal determines the computing power node, a connection between the terminal and the computing power node will be established to schedule the terminal's computing business to the computing power node for execution.

[0126] The above two methods are described below with examples.

[0127] This solution proposes a computing resource operation workflow in the context of the development of computing-network integration, in which computing resources come from edge servers and other facilities that can provide infrastructure as a service (IaaS) to users nearby, collectively referred to as computing nodes in this example.

[0128] The premise is that the computing power node information has been registered to the resource operation center, including but not limited to the computing power node address, node resource information (computing power, bandwidth) and other information.

[0129] Node computing power information: (CPU main frequency, number of CPU cores, GPU main frequency, video memory), etc.

[0130] Embodiment 1:

[0131] In this example, the ingress routing completes the computation node selection. In this example:

[0132] The computing resource request includes a node selection strategy;

[0133] The computing power routing layer establishes a connection between the terminal and the computing power node after the entry routing determines the computing power node based on the computing power resource information and charging information returned by each computing power node and the node selection strategy in the computing power resource request, so as to schedule the terminal's computing business to be executed on the computing power node.

[0134] Figure 3 This is a schematic diagram of the workflow of computing resource operation in Example 1, as shown in the figure, which may include:

[0135] 1: When a user / service needs to use computing resources, it can send a computing resource request to the resource operation center, indicating the required computing resource size and node selection strategy (for example: minimum rate, minimum delay, or minimum rate under a certain delay threshold, etc.); the resource request will first arrive at the entry router.

[0136] 2: The entry router forwards the computing resource request information to the operation center of the computing network management layer and records the user's node selection strategy.

[0137] 3: The operation center broadcasts the computing resource request to the computing nodes registered in the computing resource layer.

[0138] 4: After the computing power node receives the broadcast and the resource request information, if the requested information can be met, it measures the average usage load of the local computing power resources over the previous period of time and calculates the usage rate and returns it to the user.

[0139] 5: The ingress router receives the rate information and obtains the delay during the communication process, then selects the most suitable computing node based on the recorded user selection strategy, and establishes a computing resource usage agreement to confirm the rate information.

[0140] 6: The entry router returns the selected node information to the user.

[0141] 7: The computing power node will feed back the established computing power resource usage agreement to the operation center for record.

[0142] 8: Users can check the agreement records of the operation center at any time and pay the operation according to the agreement.

[0143] Embodiment 2:

[0144] In this example, the user completes the compute node selection, where:

[0145] The computing resource request includes the computing resource size and the maximum acceptable latency;

[0146] The computing power routing layer transmits the computing power resource information and charging information returned by each computing power node to the terminal under the condition of maximum acceptable delay. The computing power node information that meets the requirements will reach the terminal. After the terminal determines the computing power node, a connection between the terminal and the computing power node will be established to schedule the terminal's computing business to the computing power node for execution.

[0147] Figure 4 This is a schematic diagram of the workflow of computing resource operation in Example 2, as shown in the figure, which may include:

[0148] 1: When a user / service needs to use computing resources, it can send a computing resource request to the resource operation center, indicating the required computing resource size and the maximum acceptable latency; the resource request will first arrive at the entry router.

[0149] 2: The ingress router forwards the computing resource request information to the operation center of the computing network management layer and records the user's maximum acceptable delay.

[0150] 3: The operation center broadcasts the computing resource request to the computing nodes registered in the computing resource layer.

[0151] 4: After the computing power node receives the broadcast and the resource request information, if the requested information can be met, it measures the average usage load of the local computing power resources over the previous period of time and calculates the usage rate and returns it to the user.

[0152] 5: The ingress router receives the rate information and obtains the delay during the communication process, then filters the messages of the computing nodes according to the recorded maximum acceptable delay, and returns the information that meets the requirements to the user.

[0153] 6: Users independently select the required computing power nodes based on their own business needs, the returned rate information and link latency, and establish a computing power resource usage agreement with them to confirm the rate information.

[0154] 7: The computing power node will feed back the established computing power resource usage agreement to the operation center for record.

[0155] 8: Users can check the agreement records of the operation center at any time and pay the operation according to the agreement.

[0156] Example of how to calculate the dynamic rate of computing power nodes:

[0157] The computing node calculates the user fee cost based on the local load:

[0158] cost = α * X * Y * Load

[0159] Among them, X represents the basic price of the computing resource per unit of the computing node; Y represents the usage of the computing resource requested by the user; α is the discount parameter, which is set by the computing node owner; Load is the computing node load function, which will increase significantly when the node load is too high, thereby reducing the user's willingness to choose the node. Conversely, when the node load is small, it can attract more users with lower fees. Load can be calculated as follows:

[0160]

[0161] Among them, Load Curr Indicates the average load of the node in the previous 5 minutes, Load Best Indicates the highest load at which the system operates under optimal conditions.

[0162]

[0163] Based on the same inventive concept, an embodiment of the present invention further provides a computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program for executing the above-mentioned method of use.

[0164] For specific implementation, please refer to the implementation of the method of using the computing power-aware network.

[0165] In summary, the embodiments of the present invention provide a computing power-aware network supporting a computing network deep fusion architecture and a method for using the same. The computing power-aware network includes: a computing power service layer, a computing network management layer, a computing power resource layer, a computing power routing layer, and a network resource layer. Specifically:

[0166] Computing service layer: completes the functions of service decomposition and service scheduling, and can send the service requests of business or application to the computing network management layer; the computing service layer is deployed on the computing resource layer;

[0167] Computing network management layer: It is necessary to complete the perception, measurement and OAM management of computing power and network resources. In the face of heterogeneous computing resources, the computing network management layer first uses computing power modeling to uniformly describe and represent computing power resources to form node computing power information; in addition, it is also necessary to monitor the performance of computing power resources and notify the corresponding network nodes of the performance, fault and other information of computing power resources.

[0168] Computing power routing layer: Based on the discovery of abstracted computing resources, the network status and computing resource status are comprehensively considered to flexibly dispatch services to different computing power nodes on demand. Specific functions include computing power routing identification, computing power routing control, computing power status network notification, computing power routing addressing, computing power routing forwarding, etc. The computing power routing layer is used to connect the computing power resource layer and the network resource layer.

[0169] Computing resource layer: Faced with heterogeneous computing resources ubiquitously deployed on the network, it is responsible for collecting computing resource information and reporting it to the computing network management layer.

[0170] The usage method includes: each computing power node calculates the usage fee individually; the user directly establishes an agreement with the computing power node; and the message flow between the user, entry router, computing power node, and operation center.

[0171] Compared with the prior art, the technical effects of this solution are as follows:

[0172] 1. Propose a new computing power-aware network architecture to highly coordinate the network and computing, interconnect ubiquitous computing based on ubiquitous connections, achieve efficient collaboration between cloud, edge, and network, improve the utilization efficiency of computing resources, and provide users with a consistent experience.

[0173] 2. The computing power node billing information is determined through direct agreement with the user and does not rely on a unified billing server, which can provide fair competition opportunities for different edge server operators.

[0174] 3. The server dynamically adjusts the basic rate based on its own load. The rate is reduced when the load is low, and vice versa, the rate is increased; thereby maximizing the operational value of the edge server.

[0175] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) containing computer-usable program code.

[0176] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0177] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0178] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0179] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for using a computing power-aware network, characterized in that: The computing power perception network includes a computing power service layer, a computing network management layer, a computing power resource layer, a computing power routing layer and a network resource layer, wherein: the computing power service layer is used to carry various services and applications; the computing network management layer is used to receive reported computing power resource information and network resource information, and manage computing power resources and network resources; the computing power resource layer is used to provide computing power resources using computing infrastructure; the computing power routing layer is used to dispatch the services required by the terminal to different computing power nodes on demand based on computing power resources and network resources; the computing network management layer is further used to face heterogeneous computing resources, abstractly describe and represent computing power resources through computing power modeling, form node computing power information, and shield the differences in underlying hardware devices; the computing network management layer is further used to monitor and manage the performance of computing power resources and network resources to realize computing power operation and network operation; the computing power resource layer is further used to provide one or a combination of the following functions to meet the diverse computing needs in the edge computing field based on physical computing resources for different applications: computing power model, computing power API, computing power resource identification; The method comprises: After receiving the computing resource request, the computing network management layer broadcasts the computing resource request to the computing nodes of the computing resource layer; Receiving computing power resource information returned by a computing power node, wherein the computing power node is a computing power node that meets the requirements in the computing power resource request; The computing power routing layer establishes a connection between the terminal and the computing power node after determining the computing power node based on the computing power resource information and computing power resource request returned by each computing power node, so as to schedule the computing business of the terminal to be executed on the computing power node; Receive computing power node return and charging information; The computing power routing layer establishes a connection between the terminal and the computing power node after determining the computing power node based on the computing power resource information, charging information and computing power resource request returned by each computing power node, so as to schedule the computing business of the terminal to be executed on the computing power node; The computing network management layer charges the terminal based on the computing resource information and charging information returned by the computing nodes in the computing resource layer.

2. The method according to claim 1, characterized in that The computing network management layer receives computing resource requests through the resource operation center; and / or charges the terminal through the resource operation center.

3. The method according to claim 1, characterized in that The computing resource request includes a node selection strategy; The computing power routing layer establishes a connection between the terminal and the computing power node after the ingress routing determines the computing power node based on the computing power resource information and charging information returned by each computing power node and the node selection strategy in the computing power resource request, so as to schedule the computing business of the terminal to be executed on the computing power node; or, The computing resource request includes the computing resource size and the maximum acceptable latency; The computing power routing layer transmits the computing power resource information and charging information returned by each computing power node to the terminal under the condition of maximum acceptable delay. The computing power node information that meets the requirements will reach the terminal. After the terminal determines the computing power node, a connection between the terminal and the computing power node will be established to schedule the terminal's computing business to the computing power node for execution.

4. The method according to claim 1, characterized in that The computing resource request received by the computing network management layer is sent to the computing network management layer after the ingress router receives the computing resource request; and / or, The computing resources required by the terminal are dispatched to different computing nodes on demand through entry routing.

5. The method according to claim 1, characterized in that The computing power service layer is based on a distributed microservice architecture.

6. The method according to claim 1, characterized in that The computing power service layer is further used to support the deconstruction of applications into atomic functional components, which are then uniformly scheduled by the application programming interface API gateway.

7. The method according to claim 6, characterized in that The computing power service layer is further used to implement service decomposition and service scheduling through the API gateway after receiving data from end users.

8. The method according to claim 1, characterized in that The computing network management layer is further used to bill the terminals based on their use of computing power.

9. The method according to claim 1, characterized in that The computing network management layer is further used to transmit node computing power information to network nodes through computing power notifications.

10. The method according to claim 1, characterized in that The computing power routing layer is further used to dispatch the services required by the terminal to different computing power nodes on demand based on the computing power resources provided by the computing power resource layer and the network resources provided by the network resource layer.

11. The method according to claim 1, characterized in that The computing power routing layer is further used to include one or a combination of the following functions: Computing network routing identification, computing network routing forwarding, computing power routing addressing, and computing network service announcement.

12. The method according to claim 1, characterized in that Further including: The network resource layer is used to use the network infrastructure to provide network connections for various communication devices in the network.

13. The method according to claim 12, characterized in that The network resource layer further provides network connectivity for network infrastructure including one or a combination of the following facilities: Access network, metropolitan area network, backbone network.

14. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program for executing the method according to any one of claims 1 to 13.