Computing power scheduling methods, devices, electronic equipment and computer-readable storage media

By constructing a minimum spanning tree with the user device as the root node and selecting edge nodes that are close to the user device and whose computing power meets the threshold, the problem of excessive latency in existing computing power scheduling methods is solved, and efficient allocation of computing power resources is achieved.

CN116719644BActive Publication Date: 2026-11-17CHINA MOBILE (SUZHOU) SOFTWARE TECH CO LTD +1
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202310926745.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-07-26
Publication Date
2026-11-17
Estimated Expiration
2043-07-26

AI Technical Summary

Technical Problem

Existing computing power scheduling methods suffer from excessive latency, leading to increased processing time for business requests.

Method used

By constructing a minimum spanning tree with the user equipment as the root node, edge nodes that are close to the user equipment and whose computing power meets the threshold are selected for computing power scheduling. The improved Kruskal algorithm and BFS algorithm are used to dynamically match edge nodes, thereby reducing network latency.

Benefits of technology

It effectively reduces the latency of computing power scheduling, improves the efficiency of computing power resource allocation, and meets the actual business needs of user equipment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116719644B_ABST
    Figure CN116719644B_ABST
Patent Text Reader

Abstract

This application provides a computing power scheduling method, apparatus, electronic device, and computer-readable storage medium, applied in the field of computing power service network technology. The method includes receiving a computing power scheduling request sent by a user equipment; acquiring M edge nodes based on the computing power scheduling request, where M is a positive integer; determining a first weight value for a first edge based on a first network latency and the computing power of the M edge nodes; constructing a first minimum spanning tree based on N edge nodes among the M edge nodes and the computing power threshold of the user equipment, where N is a positive integer; and scheduling the N edge nodes to provide computing power to the user equipment. This method, by constructing a minimum spanning tree rooted at the user equipment and satisfying the user equipment's computing power threshold, schedules computing power based on the N edge nodes closest to the user, which helps reduce the latency of computing power scheduling.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of computing power service network technology, and in particular to a computing power scheduling method, apparatus, electronic device and computer-readable storage medium. Background Technology

[0002] Computing network refers to a new type of information infrastructure that, based on the continuous ubiquitous development of computing power, allocates and flexibly schedules computing and storage resources on demand across the cloud, edge, and endpoints through network means according to business needs. Facing the highly complex computing network environment, integrated orchestration is required at the orchestration management layer. Existing computing orchestration solutions are centralized. Centralized orchestration solutions are computing network orchestration schemes based on centralized scheduling using Software Defined Network (SDN) in data centers. In cloud data centers, multiple distributed application server nodes share business computing and storage requests. Through a centralized SDN controller, unified management and collaborative scheduling of the computing network between the cloud and edge clouds are achieved.

[0003] However, in the existing computing power scheduling method, business requests need to be sent to a centralized control unit, which is deployed in a regional data center. This increases the time required for the business request phase, resulting in excessive latency in the existing computing power scheduling method. Summary of the Invention

[0004] This application provides a computing power scheduling method, apparatus, electronic device, and computer-readable storage medium to solve the problem of excessive latency in existing computing power scheduling methods.

[0005] To solve the above-mentioned technical problems, this application is implemented as follows:

[0006] In a first aspect, embodiments of this application provide a computing power scheduling method, the method comprising:

[0007] Receive computing power scheduling requests sent by user equipment;

[0008] Based on the computing power scheduling request, obtain M edge nodes, where M is a positive integer;

[0009] A first weight value for the first edge is determined based on the first network latency and the computing power of the M edge nodes. A first minimum spanning tree is constructed based on N edge nodes among the M edge nodes and the computing power threshold of the user equipment. The user equipment is the root node of the first minimum spanning tree, the first edge is an edge in the first minimum spanning tree, the first network latency is the network latency between the two nodes forming the first edge, the sum of the computing power of the N edge nodes is greater than the computing power threshold of the user equipment, and N is a positive integer.

[0010] The N edge nodes are scheduled to provide computing power to the user equipment.

[0011] Secondly, embodiments of this application also provide a computing power scheduling device, which includes:

[0012] The first receiving module is used to receive computing power scheduling requests sent by user equipment;

[0013] The first acquisition module is used to acquire M edge nodes based on the computing power scheduling request, where M is a positive integer;

[0014] The first determining module is used to determine the first weight value of the first edge based on the first network latency and the computing power of the M edge nodes, and to construct a first minimum spanning tree based on N edge nodes among the M edge nodes and the computing power threshold of the user equipment, wherein the user equipment is the root node of the first minimum spanning tree, the first edge is an edge in the first minimum spanning tree, the first network latency is the network latency between the two nodes forming the first edge, the sum of the computing power of the N edge nodes is greater than the computing power threshold of the user equipment, and N is a positive integer;

[0015] The first scheduling module is used to schedule the N edge nodes to provide computing power to the user equipment.

[0016] Thirdly, embodiments of this application also provide a computing power scheduling device, including a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the steps of the computing power scheduling method described above.

[0017] Fourthly, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described computing power scheduling method.

[0018] The computing power scheduling method of this application includes receiving a computing power scheduling request sent by a user equipment; acquiring M edge nodes based on the computing power scheduling request, where M is a positive integer; determining a first weight value of a first edge based on a first network latency and the computing power of the M edge nodes; constructing a first minimum spanning tree based on N edge nodes among the M edge nodes and the computing power threshold of the user equipment, where the user equipment is the root node of the first minimum spanning tree, the first edge is an edge in the first minimum spanning tree, the first network latency is the network latency between the two nodes forming the first edge, the sum of the computing power of the N edge nodes is greater than the computing power threshold of the user equipment, and N is a positive integer; and scheduling the N edge nodes to provide computing power to the user equipment. This method, by constructing a minimum spanning tree with the user equipment as the root node and satisfying the user equipment's computing power threshold, schedules computing power based on the N edge nodes closest to the user, which helps reduce the latency of computing power scheduling. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart of the computing power scheduling method provided in the embodiments of this application;

[0021] Figure 2 This is one of the schematic diagrams of the computing power network in the computing power scheduling method provided in the embodiments of this application;

[0022] Figure 3 This is one of the schematic diagrams of the network architecture in the computing power scheduling method provided in the embodiments of this application;

[0023] Figure 4 This is a second schematic diagram of the network architecture of the computing power scheduling method provided in the embodiments of this application;

[0024] Figure 5 This is a flowchart of the minimum spanning tree generation process of the computing power scheduling method provided in the embodiments of this application;

[0025] Figure 6 This is a structural diagram of a computing power scheduling device provided in an embodiment of this application;

[0026] Figure 7 This is a structural diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0027] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0028] This application provides a computing power scheduling method. See also... Figure 1 , Figure 1 This is a flowchart of the computing power scheduling method provided in the embodiments of this application, such as... Figure 1 As shown, the method includes the following steps:

[0029] Step 101: Receive the computing power scheduling request sent by the user equipment;

[0030] In this step, computing power scheduling can refer to the intelligent orchestration and elastic scheduling of computing resources across the entire network, based on the control center of the computing power network. The computing power network combines computing resources with communication networks, providing computing resource services that best meet the needs of user devices through computing power scheduling requests.

[0031] It should be noted that the method provided in this application embodiment can be applied to the computing power service provider side device, which receives the computing power scheduling request sent by the user device.

[0032] Step 102: Based on the computing power scheduling request, obtain M edge nodes, where M is a positive integer;

[0033] In this step, an edge node refers to a service platform built at the network edge close to the user device, which can provide resources such as storage, computing, and networking. The computing power service provider's side device obtains M edge nodes based on the user's computing power scheduling request. These M edge nodes come from at least one local area network (LAN). It should be noted that the computing power service provider's side device will prioritize obtaining edge nodes from the LAN where the user device is located.

[0034] Step 103: Determine the first weight value of the first edge based on the first network latency and the computing power of the M edge nodes, and construct a first minimum spanning tree based on N edge nodes among the M edge nodes and the computing power threshold of the user equipment, wherein the user equipment is the root node of the first minimum spanning tree, the first edge is an edge in the first minimum spanning tree, the first network latency is the network latency between the two nodes forming the first edge, the sum of the computing power of the N edge nodes is greater than the computing power threshold of the user equipment, and N is a positive integer;

[0035] In this step, the computing power service provider's side equipment can construct a dynamic network with the user device as the root node based on the weighted minimum spanning tree algorithm. A connected graph with n vertices is traversed. For each traversed subgraph, all vertices of the original graph are included while maintaining connectivity. The final structure is a tree with n-1 edges, commonly called a spanning tree. The weighted minimum spanning tree is the spanning tree with the minimum weight value of all edges. See also... Figure 2 In this computing network, V0 represents the root node, the user device, and the other nodes represent edge nodes. When the user device continuously sends requests to the edge nodes, the higher the computing power of the edge node to which the request arrives and the lower the network latency, the better the service provided to the user.

[0036] Therefore, the computing power service provider's side equipment uses the user equipment as the root node, and the first network latency between the user equipment and edge nodes, or between edge nodes, and the computing power of the edge nodes as the weight values ​​of the first edge to construct the first minimum spanning tree. It should be noted that, to meet the actual business needs of users, this application improves the traditional Kruskal algorithm for generating minimum spanning trees by incorporating the computing power threshold of the user equipment into the generation process of the first minimum spanning tree. Among M edge nodes, N edge nodes are selected based on the improved Kruskal algorithm to generate the first minimum spanning tree, and the sum of the computing power of these N edge nodes is greater than the computing power threshold of the user equipment.

[0037] Step 104: Schedule the N edge nodes to provide computing power to the user equipment.

[0038] In this step, after N edge nodes are identified, the computing power service provider's side equipment provides the highest quality computing power resource services to user equipment based on the computing power network formed by these N edge nodes.

[0039] In this application, the computing power service provider's side equipment uses the user equipment as the root node of the computing power network. Based on the minimum spanning tree algorithm, the computing power network is constructed with N edge nodes that are close to the user equipment and meet the computing power threshold of the user equipment. The computing power scheduling needs of the user equipment no longer need to be sent to the centralized control unit of the cloud center which is far away. The computing power can be flexibly matched and dynamically scheduled directly based on the aforementioned N edge nodes, thereby reducing the latency of the user equipment in obtaining computing power services.

[0040] Optionally, the computing power of the edge node includes the computing power of the central processing unit of the edge node, the computing power of the graphics card of the edge node, and the computing power of the memory of the edge node;

[0041] The step of determining the first weight value of the first edge based on the first network latency and the computing power of the M edge nodes includes:

[0042] The first weight value is determined based on the first network latency, the central processing unit computing power, the graphics card computing power, and the memory computing power.

[0043] In this embodiment, the computing power of an edge node can be represented as H = {c, g, m}, where c represents the CPU computing power, g represents the GPU computing power, and m represents the memory computing power. A first weight value is determined based on the first network latency, CPU computing power, GPU computing power, and memory computing power. Specifically, this can be expressed using the following calculation formula:

[0044]

[0045] In cases where the computing power of the central processing unit, graphics card, or memory is 0, the first weight value f(t,c,g,m) is set to infinity, thus excluding it from the first minimum spanning tree.

[0046] In the above formula, t represents the first network latency, A1 represents the weight value of the first network latency, A2 represents the weight value of the CPU computing power, A3 represents the weight value of the GPU computing power, and A4 represents the weight value of the memory computing power. The result of A1+A2+A3+A4 is 1, and the specific values ​​of A1, A2, A3, and A4 can be determined according to the actual business needs of the user's equipment. Since a smaller first network latency and a larger computing power are better, the smaller the first weight value f(t,c,g,m), the greater the probability that the first edge corresponding to the first weight value will be included in the minimum spanning tree.

[0047] by Figure 2 For example, V0 represents the root node of the user device, and the other nodes represent edge nodes. Taking V1 as an example, Figure 2 In the context of V1:H = {50, 20, 10}, this indicates that the CPU computing power, GPU computing power, and memory computing power of this edge node are 50, 20, and 10, respectively. T = 8 represents the network latency of 8 kilobytes per second between the user device and the V1 edge node.

[0048] Referring to Table 1, in Table 1, edge represents the first edge between two nodes, and Weight represents the first weight value between the two nodes calculated according to the aforementioned formula. For example, edge[0] represents the edge between V0 and V3, and the first weight value of the edge between V0 and V3 is 1. The first weight value of 1 is based on the first network delay between V0 and V3 being 2, and... Figure 2 The computing power of V3 shown is calculated from H = {80, 40, 100}.

[0049] Each first edge is traversed in ascending order of its corresponding first weight value. It is then determined whether adding the first edge to the first minimum spanning tree would form a cycle. If not, the edge is added to the first minimum spanning tree.

[0050] Table 1

[0051]

[0052] The computing power scheduling method of this application obtains the first weight value of the first edge of the first minimum spanning tree by taking the first network latency, the central processing unit computing power, memory computing power and graphics card computing power of the edge node. This is beneficial for selecting edge nodes with short network latency and high computing power from the user device based on the first weight value.

[0053] Optionally, the user equipment is located in a first local area network, and K of the N edge nodes are edge nodes in the first local area network. The method further includes:

[0054] If the sum of the computing power of the K edge nodes is less than the computing power threshold, obtain NK edge nodes in the second local area network, where K is a positive integer.

[0055] In this embodiment, after receiving the computing power scheduling request from the user device, the computing power service provider-side device first obtains all edge nodes in the local area network where the user device is located. Using the same generation principle as in the aforementioned embodiment, a minimum spanning tree is generated based on all edge nodes (i.e., K edge nodes) in the first local area network where the user device is located. However, if the sum of the computing power of the K edge nodes that construct this minimum spanning tree is less than the computing power threshold of the user device, it means that these K edge nodes do not meet the actual business needs of the user device. In this case, it is necessary to expand the minimum spanning tree, that is, to find more edge nodes in other local area networks to provide computing power services to the user device, until the sum of the computing power of all edge nodes in the final minimum spanning tree can meet the computing power threshold of the user device. The expansion can use the breadth-first search (BFS) algorithm.

[0056] It should be noted that the edge nodes added to the minimum spanning tree also need to meet the requirements of low network latency and high computing power. For example, using... Figure 2Based on all edge nodes, the minimum spanning tree is composed of V0, V2, V3, and V4. However, the sum of the computing power of V2, V3, and V4 is less than the computing power threshold of the user device. At this time, it is necessary to expand the minimum spanning tree. Find edge node V7 in other local area networks. The edge node connected to V7 is V4. The network latency between V7 and V4 is t = 5. The computing power of V7 is H = {80, 40, 100}. Referring to the aforementioned calculation method, calculate the first weight value of the edge formed by V7 and V4, and then determine whether V7 should be included in the aforementioned minimum spanning tree.

[0057] The computing power scheduling method in this application first generates a minimum spanning tree based on the local area network where the user equipment is located. If the sum of computing power of the edge nodes in the minimum spanning tree generated based on the local area network where the user equipment is located does not meet the computing power threshold of the user equipment, the aforementioned minimum spanning tree is expanded until a first minimum spanning tree that meets the computing power threshold of the user equipment is finally obtained. This is beneficial for providing computing power services to users according to their actual business needs.

[0058] Optionally, the second local area network (LAN) is a LAN adjacent to the first LAN.

[0059] In this embodiment, the second local area network is a local area network adjacent to the first local area network. If the minimum spanning tree generated based on the edge nodes of the local area network where the user equipment is located does not meet the computing power threshold of the user equipment, more edge nodes are preferentially obtained from the local area networks adjacent to the local area network where the user equipment is located to expand the aforementioned minimum spanning tree. This helps to reduce the network latency of computing power scheduling.

[0060] Optionally, scheduling the N edge nodes to provide computing power to the user equipment includes:

[0061] Obtain the Internet Protocol (IP) addresses corresponding to the N edge nodes;

[0062] Based on the IP address corresponding to each edge node, the computing power scheduling request is forwarded to each of the N edge nodes respectively;

[0063] The N edge nodes are scheduled to provide computing power to the user equipment.

[0064] In this embodiment, after the computing power service provider's side device determines the N edge nodes that will provide computing power services to the user equipment, it needs to actually utilize these N edge nodes. First, it obtains the Internet Protocol (IP) addresses corresponding to these N edge nodes. Then, based on the IP address of each edge node, it forwards the computing power scheduling request to each of the N edge nodes, and then schedules the N edge nodes to provide computing power to the user equipment. The computing power scheduling method in this embodiment can more accurately schedule each edge node by using the IP addresses corresponding to the N edge nodes, thereby providing computing power services to the user equipment.

[0065] Optionally, forwarding the computing power scheduling request to the N edge nodes according to the IP address corresponding to each edge node includes:

[0066] The IP routing table is encapsulated into a BGP update message using the Border Gateway Protocol (BGP).

[0067] The BGP update message is sent to the Network Function Virtualization (NFV), and the NFV is used to orchestrate Virtual Network Function (VNF) network elements according to the BGP update message.

[0068] The computing power scheduling request is forwarded to the N edge nodes corresponding to the IP routing table through the orchestrated VNF network elements.

[0069] In the embodiments of this application, see Figure 3 and Figure 4 The diagram illustrates how the Border Gateway Protocol Speaker (BGPSpeaker) encapsulates the IP routing table into BGP Update messages. These Update messages are then sent to the Network Function Virtualization Management and Orchestration (NFV MANO). The NFV MANO then re-orchestrates the Virtualized Network Functions (VNFs) based on the Update messages. Finally, user computing power scheduling requests are forwarded through these orchestrated VNFs to the corresponding N edge nodes in the IP routing table. Furthermore, the computing power control table can schedule basic computing units, using Kubernetes (K8S) Point of Delivery (Pod) delivery points as carriers, to edge nodes in the first minimum spanning tree.

[0070] The computing power scheduling method in this application re-orchestrates VNFs through NFV, which is beneficial for building a computing power orchestration network based on edge nodes with user equipment as the core.

[0071] Furthermore, in existing computing power scheduling methods, the NFV MANO center needs to manage the computing power of every edge node, every cloud pool, and even every end device in the entire network. Storage and computing consume a lot of resources of the NFV MANO center. In the computing power scheduling method of this application, the NFV MANO center only needs to manage the edge nodes that are closer to the user, thereby greatly reducing the resource consumption of the NFV MANO center.

[0072] See Figure 5 , Figure 5 This is a second flowchart of a computing power scheduling method provided in the embodiments of this application, such as... Figure 5 As shown, in actual business applications, user devices have a threshold for computing power requirements. Initially, the edge nodes within the local area network (LAN) where the user device resides may not meet this threshold. Therefore, the user device's computing power threshold parameter needs to be incorporated into the minimum spanning tree (MSB) construction process. If the generated MSB fails to meet the computing power threshold, a breadth-first search (BFS) algorithm is needed in other LANs to find other edge nodes. Referring to the aforementioned implementation method, the weight values ​​of the edges formed between edge nodes are calculated. Then, each edge is traversed in ascending order of its corresponding weight value. If a cycle is formed with the original MSB, it is discarded; otherwise, it is added to the MSB. This process is repeated until a first MSB that meets the computing power threshold requirement is obtained.

[0073] For any undirected weighted graph G = {V, E, W}, where V = {V1, V2, ... V} n} represents the nodes of the graph; E = {E1, E2, ..., E} m} represents the edges of the graph; W = {W1, W2, ..., W} m} represents the weight of the edges, and m represents the number of edges. Define variable E` to represent the edges sorted according to W. The Kruskal algorithm formula is:

[0074]

[0075] Among them, E i ' represents the i-th edge after sorting by the first weight value; T represents the minimum spanning tree; the function f(E) i ',T) is used to determine edge E i If the two vertices of ' cannot form a cycle in the already generated minimum spanning tree T, then {V} ii V ijW i Add it to T, otherwise discard it. ii V ij Representing edge E i '' two nodes, W i Representing edge E i The weighted value of '.

[0076] After considering the computing power threshold of user devices, we first define a variable V' to store the value based on T.

[0077] `

[0078] Edge nodes that extend from a node to adjacent local area networks. If V is not empty, it indicates that it is not the first calculation, and `

[0079] The value of V is assigned to V. The specific steps are as follows:

[0080] Step 1: Obtain the initial minimum spanning tree, where V' is initially an empty set.

[0081]

[0082] Step 2: Based on the calculation result T from the previous step, use the BFS algorithm to search for the edge nodes of adjacent local area networks.

[0083] B(V,E',W,M,V')=V+BFS(T(V,E,W,V'),V')E' i ∈E',V'={}

[0084] Step 3: Determine whether all edge nodes meet the computing power threshold M. If they do, stop the calculation and obtain the final minimum spanning tree.

[0085] T final (V,E',W,M,V')=T(V,E',W,V')+B(V,E',W,M,V')

[0086] If the condition is not met, the loop continues.

[0087] See Figure 6 , Figure 6 This is a structural diagram of a computing power scheduling device provided in an embodiment of this application.

[0088] like Figure 6 As shown, the computing power scheduling device 600 includes:

[0089] The first receiving module 601 is used to receive computing power scheduling requests sent by user equipment;

[0090] The first acquisition module 602 is used to acquire M edge nodes based on the computing power scheduling request, where M is a positive integer;

[0091] The first determining module 603 is used to determine the first weight value of the first edge based on the first network latency and the computing power of the M edge nodes, and to construct a first minimum spanning tree based on N edge nodes among the M edge nodes and the computing power threshold of the user equipment, wherein the user equipment is the root node of the first minimum spanning tree, the first edge is an edge in the first minimum spanning tree, the first network latency is the network latency between the two nodes forming the first edge, the sum of the computing power of the N edge nodes is greater than the computing power threshold of the user equipment, and N is a positive integer;

[0092] The first scheduling module 604 is used to schedule the N edge nodes to provide computing power to the user equipment.

[0093] Optionally, the computing power of the edge node includes the computing power of the central processing unit of the edge node, the computing power of the graphics card of the edge node, and the computing power of the memory of the edge node;

[0094] The first determining module 603 includes:

[0095] The first determining unit is configured to determine the first weight value based on the first network latency, the central processing unit computing power, the graphics card computing power, and the memory computing power.

[0096] Optionally, the user equipment is located in a first local area network, and K of the N edge nodes are edge nodes in the first local area network. The device further includes:

[0097] The second acquisition module is used to acquire NK edge nodes in the second local area network when the sum of the computing power of the K edge nodes is less than the computing power threshold, where K is a positive integer.

[0098] Optionally, the second local area network (LAN) is a LAN adjacent to the first LAN.

[0099] Optionally, the first scheduling module 604 includes:

[0100] The first acquisition unit is used to acquire the Internet Protocol IP addresses corresponding to the N edge nodes;

[0101] The first forwarding unit is used to forward the computing power scheduling request to the N edge nodes respectively according to the IP address corresponding to each edge node;

[0102] The first scheduling unit is used to schedule the N edge nodes to provide computing power to the user equipment.

[0103] Optionally, the first forwarding unit includes:

[0104] The first encapsulation subunit is used to encapsulate the IP routing table into a BGP update message using the Border Gateway Protocol (BGP).

[0105] The first sending subunit is used to send the BGP update message to the Network Function Virtualization (NFV), and the NFV is used to orchestrate Virtual Network Function (VNF) network elements according to the BGP update message.

[0106] The first forwarding subunit is used to forward the computing power scheduling request to N edge nodes corresponding to the IP routing table through the orchestrated VNF network elements.

[0107] Please see Figure 7 , Figure 7 This is a structural diagram of an electronic device provided in an embodiment of this application, such as... Figure 7 As shown, the electronic device includes: a processor 700, a memory 720, and a program or instructions stored in the memory 720 and executable on the processor 700. The processor 700 is used to read the program or instructions from the memory 720. The electronic device also includes a bus interface and a transceiver 710.

[0108] Transceiver 710 is used to receive and send data under the control of processor 700.

[0109] Among them, Figure 7 In this context, the bus architecture can include any number of interconnected buses and bridges, specifically linking various circuits together, represented by one or more processors (processor 700) and memory (memory 720). The bus architecture can also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. The bus interface provides an interface. The transceiver 710 can be multiple elements, including transmitters and transceivers, providing a unit for communicating with various other devices over a transmission medium. The processor 700 is responsible for managing the bus architecture and general processing, and the memory 720 can store data used by the processor 700 during operation.

[0110] The processor 700 is used to read programs or instructions from the memory 720 and execute the following steps:

[0111] Receive computing power scheduling requests sent by user equipment;

[0112] Based on the computing power scheduling request, obtain M edge nodes, where M is a positive integer;

[0113] A first weight value for the first edge is determined based on the first network latency and the computing power of the M edge nodes. A first minimum spanning tree is constructed based on N edge nodes among the M edge nodes and the computing power threshold of the user equipment. The user equipment is the root node of the first minimum spanning tree, the first edge is an edge in the first minimum spanning tree, the first network latency is the network latency between the two nodes forming the first edge, the sum of the computing power of the N edge nodes is greater than the computing power threshold of the user equipment, and N is a positive integer.

[0114] The N edge nodes are scheduled to provide computing power to the user equipment.

[0115] Optionally, the computing power of the edge node includes the computing power of the central processing unit of the edge node, the computing power of the graphics card of the edge node, and the computing power of the memory of the edge node;

[0116] Processor 700 is used to read programs or instructions from memory 720 and perform the following steps:

[0117] The step of determining the first weight value of the first edge based on the first network latency and the computing power of the M edge nodes includes:

[0118] The first weight value is determined based on the first network latency, the central processing unit computing power, the graphics card computing power, and the memory computing power.

[0119] Optionally, the processor 700 is configured to read a program or instruction from the memory 720 and perform the following steps:

[0120] If the sum of the computing power of the K edge nodes is less than the computing power threshold, obtain NK edge nodes in the second local area network, where K is a positive integer.

[0121] Optionally, the second local area network (LAN) is a LAN adjacent to the first LAN.

[0122] Optionally, the processor 700 is configured to read a program or instruction from the memory 720 and perform the following steps:

[0123] Obtain the Internet Protocol (IP) addresses corresponding to the N edge nodes;

[0124] Based on the IP address corresponding to each edge node, the computing power scheduling request is forwarded to each of the N edge nodes respectively;

[0125] The N edge nodes are scheduled to provide computing power to the user equipment.

[0126] Optionally, the processor 700 is configured to read a program or instruction from the memory 720 and perform the following steps:

[0127] The IP routing table is encapsulated into a BGP update message using the Border Gateway Protocol (BGP).

[0128] The BGP update message is sent to the Network Function Virtualization (NFV), and the NFV is used to orchestrate Virtual Network Function (VNF) network elements according to the BGP update message.

[0129] The computing power scheduling request is forwarded to the N edge nodes corresponding to the IP routing table through the orchestrated VNF network elements.

[0130] This application also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the various processes of the above-described computing power scheduling method embodiments and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0131] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0132] This application embodiment also provides a chip, the chip including a processor and a communication interface, the communication interface being coupled to the processor, the processor being used to run programs or instructions to implement the above. Figure 1 The various processes of the aforementioned screen distortion processing method embodiments can achieve the same technical effect, and will not be described again here to avoid repetition.

[0133] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0134] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0135] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0136] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A computing power scheduling method, characterized in that, The method includes: Receive computing power scheduling requests sent by user equipment; Based on the computing power scheduling request, obtain M edge nodes, where M is a positive integer; The first weight value of the first edge is determined based on the first network latency between the two nodes forming the first edge and the computing power of the two nodes. A first minimum spanning tree is constructed based on N edge nodes out of the M edge nodes and the computing power threshold of the user equipment. The user equipment is the root node of the first minimum spanning tree, the first edge is an edge in the first minimum spanning tree, the sum of the computing power of the N edge nodes is greater than the computing power threshold of the user equipment, and N is a positive integer. The N edge nodes are scheduled to provide computing power to the user equipment. If the sum of the computing power of the K edge nodes is less than the computing power threshold, obtain NK edge nodes in the second local area network. The user equipment is located in the first local area network, and K edge nodes among the N edge nodes are edge nodes in the first local area network, where K is a positive integer and K≤N.

2. The computing power scheduling method of claim 1, wherein, The computing power of the edge node includes the computing power of the central processing unit of the edge node, the computing power of the graphics card of the edge node, and the computing power of the memory of the edge node; The step of determining the first weight value of the first side based on the first network latency between the two nodes forming the first side and the computing power of the two nodes includes: The first weight value is determined based on the first network latency, the central processing unit computing power, the graphics card computing power, and the memory computing power.

3. The computing power scheduling method of claim 1, wherein, The second local area network is a local area network adjacent to the first local area network.

4. The computing power scheduling method of claim 1, wherein, The scheduling of the N edge nodes to provide computing power to the user equipment includes: Obtain the Internet Protocol (IP) addresses corresponding to the N edge nodes; Based on the IP address corresponding to each edge node, the computing power scheduling request is forwarded to each of the N edge nodes respectively; Control the N edge nodes to execute computing tasks according to the computing power scheduling request.

5. The computing power scheduling method of claim 4, wherein, The step of forwarding the computing power scheduling request to the N edge nodes according to the IP address corresponding to each edge node includes: The IP routing table is encapsulated into BGP update messages using the Border Gateway Protocol (BGP). The BGP update message is sent to the Network Function Virtualization (NFV) orchestrator, which is used to orchestrate Virtual Network Function (VNF) network elements according to the BGP update message. The computing power scheduling request is forwarded to the N edge nodes by the orchestrated VNF network elements according to the IP addresses in the IP routing table.

6. A computing power scheduling device, characterized in that, The device includes: The first receiving module is used to receive computing power scheduling requests sent by user equipment; The first acquisition module is used to acquire M edge nodes based on the computing power scheduling request, where M is a positive integer; The first determining module is used to determine the first weight value of the first edge based on the first network delay between the two nodes forming the first edge and the computing power of the two nodes, and to construct a first minimum spanning tree based on N edge nodes among the M edge nodes and the computing power threshold of the user equipment, wherein the user equipment is the root node of the first minimum spanning tree, the first edge is an edge in the first minimum spanning tree, the sum of the computing power of the N edge nodes is greater than the computing power threshold of the user equipment, and N is a positive integer; The first scheduling module is used to schedule the N edge nodes to provide computing power to the user equipment; The second acquisition module is used to acquire NK edge nodes in the second local area network when the sum of the computing power of the K edge nodes is less than the computing power threshold. The user equipment is located in the first local area network, and K edge nodes among the N edge nodes are edge nodes in the first local area network, where K is a positive integer and K≤N.

7. The computing power scheduling device according to claim 6, characterized in that, The computing power of the edge node includes the computing power of the central processing unit of the edge node, the computing power of the graphics card of the edge node, and the computing power of the memory of the edge node; The first determining module includes: The first determining unit is configured to determine the first weight value based on the first network latency, the central processing unit computing power, the graphics card computing power, and the memory computing power.

8. An electronic device, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the computing power scheduling method as described in any one of claims 1 to 5.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the computing power scheduling method as described in any one of claims 1 to 5.

Citation Information

Patent Citations

  • Concurrent task optimization method based on multi-edge server cooperation

    CN110300168A