WO2023240911A1 - 算力调度系统和方法、控制节点、存储介质、设备 - Google Patents

算力调度系统和方法、控制节点、存储介质、设备 Download PDF

Info

Publication number
WO2023240911A1
WO2023240911A1 PCT/CN2022/130589 CN2022130589W WO2023240911A1 WO 2023240911 A1 WO2023240911 A1 WO 2023240911A1 CN 2022130589 W CN2022130589 W CN 2022130589W WO 2023240911 A1 WO2023240911 A1 WO 2023240911A1
Authority
WO
WIPO (PCT)
Prior art keywords
computing power
information
resource information
target
resource
Prior art date
Application number
PCT/CN2022/130589
Other languages
English (en)
French (fr)
Inventor
王若倪
梁伟
卢毅
Original Assignee
中国电信股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中国电信股份有限公司 filed Critical 中国电信股份有限公司
Publication of WO2023240911A1 publication Critical patent/WO2023240911A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L63/00—Network architectures or network communication protocols for network security
    • H04L63/08—Network architectures or network communication protocols for network security for authentication of entities
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/50—Network services
    • H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L9/00—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols
    • H04L9/40—Network security protocols

Definitions

  • the present disclosure relates to the field of network communication technology, and specifically to a computing power scheduling system and method, control node, storage medium, and equipment.
  • a computing power scheduling system includes: a computing power requester, a computing power supplier and a control node located in a blockchain network.
  • the control node includes an evaluation module; the computing power requester , used to send a computing power resource request including business type information to the evaluation module; the evaluation module is used to evaluate the first computing power resource information of the computing power requester based on the received business type information; the evaluation module is also used to obtain the computing power resource information.
  • the candidate computing power resource information of the computing power supplier and based on the candidate computing power resource information, evaluate the second computing power resource information of the computing power supplier; the control node is used to evaluate the first computing power resource information and the second computing power resource information based on the evaluation module.
  • Computing power resource information determine the target computing power supply equipment, and schedule the target computing power supply equipment to provide target computing power resources to the computing power requester.
  • the computing power resource request includes identification information of the computing power requester
  • the target computing power resource includes identification information of the target computing power supply device.
  • the evaluation module includes a first evaluation sub-module, and the first evaluation sub-module is used to: determine a target computing power evaluation model based on the business type; use the target computing power evaluation model to perform computing power evaluation on the computing power resource request, To obtain the first computing resource information.
  • the candidate computing resource information includes hardware resource information, resource usage information, current period information and/or network resource information of the computing power supplier.
  • the evaluation module also includes a second evaluation sub-module. The second evaluation sub-module Used to: perform mathematical operations on hardware resource information, resource usage information, current period information and/or network resource information to evaluate the second computing power resource information of the computing power supplier.
  • the blockchain network also includes smart contracts
  • the system also includes a consensus node
  • the evaluation module is also used to: send the first computing power resource information and the second computing power resource information to the consensus node
  • the consensus node is used to: Based on the received first computing power resource information and the second computing power resource information, candidate computing power supply equipment is determined; according to the consensus matching rules in the smart contract, the target computing power supply equipment is matched and screened out among the candidate computing power supply equipment.
  • the computing power supplier or consensus node is also used to: settle the computing power resource transaction between the computing power demander and the computing power supplier to obtain the transaction data package, and send the transaction data package to the blockchain Webcast.
  • a computing power scheduling method includes: receiving a computing power resource request from a computing power requester including business type information; and based on the business type information, evaluating the computing power requesting party's first computing power request. computing power resource information; obtain the candidate computing power resource information of the computing power supplier, and evaluate the second computing power resource information of the computing power supplier based on the candidate computing power resource information; based on the first computing power resource information and the second computing power resource Information, determine the target computing power supply equipment, and schedule the target computing power supply equipment to provide target computing power resources to the computing power requester; wherein, the computing power requester and the computing power supplier are located in the same blockchain network.
  • the computing power resource request includes identification information of the computing power requester
  • the target computing power resource includes identification information of the target computing power supply device.
  • evaluating the first computing power resource information of the computing power requester based on the business type information includes: determining a target computing power evaluation model based on the business type; using the target computing power evaluation model to calculate the computing power resource request. Capacity assessment to obtain the first computing power resource information.
  • the candidate computing power resource information includes the computing power supplier's hardware resource information, resource usage information, current period information and/or network resource information
  • the evaluation of the computing power supplier's second computing power resource information includes: Perform mathematical operations on hardware resource information, resource usage information, current period information and/or network resource information to evaluate the second computing power resource information of the computing power supplier.
  • determining the target computing power supply device based on the first computing power resource information and the second computing power resource information includes: determining the candidate computing power supply based on the first computing power resource information and the second computing power resource information. Equipment; according to the consensus matching rules in the smart contract, the target computing power supply equipment is selected from the candidate computing power supply equipment.
  • the method further includes: settling the computing power resource transaction between the computing power demander and the computing power supplier to obtain a transaction data package, and broadcasting the transaction data package to the blockchain network.
  • a control node includes: a first evaluation module, a second evaluation module and a determination module; the first evaluation module is used to receive the computing power of the computing power requester including business type information. resource request; and based on the business type information, evaluate the first computing power resource information of the computing power requester; the second evaluation module is used to obtain the candidate computing power resource information of the computing power supplier, and evaluate the computing power resource information based on the candidate computing power resource information.
  • the second computing power resource information of the power supplier is used to determine the target computing power supply equipment based on the first computing power resource information and the second computing power resource information, so as to schedule the target computing power supply equipment to provide the computing power requester with Target computing resources; among them, computing power requester, computing power supplier and control node are located in the same blockchain network.
  • a computer-readable storage medium is provided, a computer program is stored thereon, and when the computer program is executed by a processor, any one of the above methods is implemented.
  • a computing power scheduling device including: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to execute any of the above by executing the executable instructions. item method.
  • Figure 1 schematically shows one of the structural schematic diagrams of a computing power scheduling system according to some embodiments of the present disclosure.
  • Figure 2 schematically shows a structural diagram of an evaluation module according to some embodiments of the present disclosure.
  • Figure 3 schematically shows the second structural schematic diagram of the computing power scheduling system according to some embodiments of the present disclosure.
  • Figure 4 schematically shows one of the flowcharts of a computing power scheduling method according to some embodiments of the present disclosure.
  • Figure 5 schematically shows the second schematic flowchart of the computing power scheduling method according to some embodiments of the present disclosure.
  • Figure 6 schematically illustrates a structural block diagram of an exemplary control node according to some embodiments of the present disclosure.
  • Figure 7 schematically shows a block diagram of an exemplary computing power scheduling device according to some embodiments of the present disclosure.
  • Example embodiments will now be described more fully with reference to the accompanying drawings.
  • Example embodiments may, however, be embodied in various forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concepts of the example embodiments.
  • the described features, structures or characteristics may be combined in any suitable manner in one or more embodiments.
  • numerous specific details are provided to provide a thorough understanding of embodiments of the disclosure.
  • those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced without one or more of the specific details being omitted, or other methods, components, devices, steps, etc. may be adopted.
  • well-known technical solutions have not been shown or described in detail to avoid obscuring aspects of the disclosure.
  • the purpose of the embodiments of the present disclosure is to provide a computing power scheduling system and method, device, storage medium, and equipment, thereby solving to a certain extent the inability to achieve balanced matching of computing power and computing power resources between the demand side and the supply side in related technologies. Low utilization problem.
  • the evaluation module is set to realize the evaluation of the computing power requested by the computing power requester, and at the same time, based on the candidate computing power resources of the current computing power supplier information, evaluate the computing power resources of the computing power supplier, and then use the control node to balance the computing power resources of the computing power requester and the computing power supplier based on the evaluation results of the evaluation module, and determine the target computing power supply equipment. This achieves unified measurement and balanced scheduling of computing resources and improves utilization of computing resources.
  • this disclosure sets the computing power requester, computing power supplier and control node in the blockchain network, that is, the computing power scheduling process is implemented in the blockchain network, so that all participants in the blockchain network Only then can we know the current and historical computing power resource transactions in a timely manner, realize the sharing of computing power resources, and further ensure the balanced dispatch of computing power.
  • the blockchain network can realize the on-chain storage of computing power resources to facilitate later maintenance and troubleshooting.
  • the computing power scheduling system provided in some embodiments of the present disclosure can be deployed in a blockchain network.
  • the system 100 may include a computing power requester 110, a computing power supplier 120 and a control node 130 located in the blockchain network.
  • the control node 130 includes an evaluation module 131.
  • the computing power scheduling system of the present disclosure can be applied to terminal computing scenarios, network edge computing scenarios and cloud computing scenarios.
  • the computing power requester 110, the computing power supplier 120 and the control node 130 in this disclosure may each be one or more devices with independent addresses and connected to a network with the function of transmitting or receiving data.
  • the device can be a workstation, a mobile terminal, a network user or a personal computer, or it can be a server, printer and other devices connected to the network, which is not limited in this example.
  • the computing power requester 110 is used to send a computing power resource request including business type information to the evaluation module.
  • the computing power resource request may include business data, that is, data to be processed and business type information, and may also include identification information of the computing power requester (such as a digital signature or coded ID, etc.).
  • the computing resource request may also include other information, such as the acquisition time or time limit requirements of the required computing resources, distance requirements, etc. This example does not impose special restrictions on this.
  • the service type information may be determined according to the data form of the service data.
  • the service type information when the service data is an image, the service type information may be an image class; when the service data is a natural language, the service type information may be a natural language class; when the service data is audio data, the service type may be an audio class.
  • the business type information can also be determined based on the amount of business data. For example, when the amount of business data is large (such as greater than a certain threshold or ratio), if it is an image, the business type information can be a complex class; when the amount of business data is small (such as less than a certain threshold or ratio) , the business type information can be a simple class. Similarly, other divisions of business data types can also be made, and this example does not impose special limitations on this.
  • the evaluation module 131 is configured to evaluate the first computing power resource information of the computing power requester 110 based on the received business type information.
  • the first computing power resource information refers to the computing power resource information required by the computing power requester.
  • the first computing resource information can be determined by analyzing and evaluating the business data in the computing resource request.
  • the evaluation module 131 is also used to obtain the candidate computing power resource information of the computing power supplier 120, and evaluate the second computing power resource information of the computing power supplier 120 based on the candidate computing power resource information.
  • the candidate computing power resource information refers to the computing power resource information that the computing power supplier can currently provide.
  • Candidate computing power resource information may include information on the total amount of available resources of the computing power provider and its resource utilization, etc.; information on the total amount of available resources may include hardware resource information such as performance data of CPU and memory, and may also include information related to network resources. (such as network delay, bandwidth, etc.), etc. This example does not limit this.
  • the second computing power resource information refers to computing power resources that can be provided to the computing power requester after evaluation.
  • the second computing power resource information may be comprehensive index information determined based on hardware resource information and/or network resource information, for example, it may be a weighted operation of hardware resource information and/or network resource information.
  • the second computing power resource information may also be level information determined based on hardware resource information and/or network resource information.
  • the second computing power resource information may be determined by setting a mapping relationship between hardware resource information and/or network resource information and levels. .
  • the second computing power resource information may be one or more numerical values, which is not limited in this example.
  • the control node 130 is configured to determine the target computing power supply device based on the first computing power resource information and the second computing power resource information evaluated by the evaluation module 131, and schedule the target computing power supply device to provide the computing power requester 110 with the target computing power. resource.
  • the first computing power resource information and the second computing power resource information can be matched according to preset matching rules to determine the target computing power supply device of the computing power supplier.
  • the target computing power resource may include identification information of the target computing power supply device.
  • the identification information may be a signature (such as a digital signature) or an encoded ID, which is not limited in this example.
  • the fault cause determination method realizes the evaluation of the requested computing power of the computing power requester by setting up an evaluation module, and at the same time, based on the candidate computing power resource information of the current computing power supplier, The computing power resources of the computing power supplier are evaluated, and then the computing power resources of the computing power requester and the computing power supplier are balanced and matched through the control node based on the evaluation results of the evaluation module, and the target computing power supply equipment is determined to achieve computing power.
  • the unified measurement and balanced scheduling of resources can improve the utilization of computing resources.
  • this disclosure sets the computing power requester, computing power supplier and control node in the blockchain network, that is, the computing power scheduling process is implemented in the blockchain network, so that all participants in the blockchain network Only then can we know the current and historical computing power resource transactions in a timely manner, realize the sharing of computing power resources, and further ensure the balanced dispatch of computing power.
  • the blockchain network can realize the on-chain storage of computing power resources to facilitate later maintenance and troubleshooting.
  • the evaluation module 200 includes at least one of a first evaluation sub-module 210 and a second evaluation sub-module 220 .
  • the first evaluation sub-module 210 is configured to: determine a target computing power evaluation model based on the business type; use the target computing power evaluation model to perform computing power evaluation on the computing power resource request to obtain the first computing power resource information.
  • different computing power evaluation models can be used for different business types.
  • the convolutional neural network model can be used; when the business type is natural speech processing, the recursive neural network model can be used; when the business type is audio processing, the recursive neural network model that adds speech recognition function can be used.
  • Network model when the business type is video processing, a convolutional neural network model with temporal feature extraction function can be used.
  • the target computing power can be comprehensively determined based on information such as business type and business data volume, processing difficulty (for example, image processing is more difficult, data type processing is less difficult), processing delay requirements, etc. Evaluate the model.
  • the target computing power evaluation model is a variety of trained neural network models. For example, a correspondence can be established between one or more types of information such as business type and business data volume, processing difficulty, processing delay requirements, etc., and the computing power evaluation model, thereby quickly determining the target computing power. Evaluate the model.
  • the parameters in the target computing power evaluation model can be updated regularly or dynamically, and this example does not limit this.
  • the business data in the computing resource request can be used as the input data of the target computing power evaluation model, and after the computing power evaluation of the model, the first computing power resource information corresponding to the request is output.
  • the first computing power resource information may include required CPU-related information and memory-related information, such as CPU main frequency, first-level cache and memory frequency, and may also include hardware information such as bus bit width and byte constants.
  • the first computing power resource information may also include network resource information such as network delay requirements, and may also include other resource information, which is not limited in this example.
  • the second evaluation sub-module 220 is configured to: perform mathematical operations on hardware resource information, resource usage information, current period information and/or network resource information (that is, perform mathematical operations on hardware resource information, resource usage information, current period information, network resource information). At least one of the mathematical operations) to evaluate the second computing power resource information of the computing power supplier.
  • the hardware resource information may include CPU information, memory information, cache information and/or bus bit width, byte constants and other information, and the resource usage information may include the usage of each hardware resource.
  • the ratio of used memory to total memory is used as memory usage information.
  • the current time period information can include current time information, week information, month information, etc., and can also include working hours and rest periods, peak hours and ordinary hours, etc. This example does not limit this. Different identification information or coding can be used to identify different time periods, and the identification information or coding can be set based on experience.
  • the network resource information may include network transmission delay and/or response delay, and may also include network bandwidth and other information.
  • Mathematical operations can be simple four arithmetic operations or weighted four arithmetic operations, or they can be complex function operations (such as logarithmic functions, trigonometric functions), integrals, differentials, etc., or a combination of various operations. This example does not impose special restrictions on this.
  • the second computing power resource information may be a computing power index value, or may be a related operation result of multiple computing power index values, which is not limited in this example.
  • the total amount of CPU resources of some nodes of the computing power supplier is C res
  • the CPU resource utilization rate is C uti
  • the total amount of memory resources is M res
  • the memory resource utilization rate is M uti
  • the current period information is T
  • ⁇ and ⁇ are preset weight parameters respectively, ⁇ represents the CPU weight, and ⁇ represents the memory resource weight.
  • the values of ⁇ and ⁇ can be any natural numbers and can be set based on experience. ⁇ and ⁇ can also be set differently depending on the business type, and this example does not limit this.
  • the blockchain network also includes smart contracts.
  • the system also includes a consensus node, and the control node includes an evaluation module.
  • the evaluation module is also used to: send the first computing power resource information and the second computing power resource information to the consensus node.
  • the consensus node may be another node different from the control node.
  • a consensus node can be one or more devices with independent addresses and connected to a network capable of transmitting or receiving data.
  • the device can be a workstation, a mobile terminal, a network user or a personal computer, or it can be a server, printer and other devices connected to the network, which is not limited in this example.
  • the evaluation results of the evaluation module can be sent to the consensus node for computing resource matching.
  • a smart contract is a collection of rules recognized by all participants in the blockchain network. Smart contracts can be set up in any node in the blockchain network. For example, smart contracts can be set up in consensus nodes.
  • the consensus node is used to: determine candidate computing power supply equipment based on the received first computing power resource information and second computing power resource information; match and screen out the target computing power supply equipment among the candidate computing power supply equipment according to the consensus matching rules in the smart contract. Power supply equipment.
  • the first computing power resource information may be one or more computing power index values.
  • the second computing power resource information may include the computing power index value of each device/node in the computing power supplier. For example, if the first computing power resource information is Power1 and the second computing power resource information is Power2, then the set formed by the combination of all devices/nodes in Power2 that are larger than Power1 can be used as the candidate computing power supply device, for example, the If the computing power resources corresponding to devices A, B, and C in the second computing power resource information can satisfy Power1, then devices A, B, and C can be used as a device combination of candidate computing power supply devices. You can also set the distance condition between the computing power requester and the computing power supplier to determine the candidate computing power supply device when the computing power is satisfied. You can also set other conditions, which are not limited in this example.
  • the consensus matching rule refers to the consensus computing power matching rule reached by each participant in the blockchain network.
  • consensus matching rules can be set based on computing power resources and/or distance conditions between the computing power requester and the computing power supplier. For example, on the premise that the computing power resources meet the demand (such as greater than or equal to the requesting party's computing power resources), the device/node or device/node combination with the minimum computing power can be selected as the target computing power supply device, so that Reduce waste of computing resources. It can also be set that on the premise of meeting the required computing power resources, the device/node or device/node combination with the minimum distance can be selected as the target computing power supply device, thereby reducing the delay.
  • the target computing power supply device can also be selected by combining computing power size and distance factors. This example does not impose special restrictions on this.
  • the consensus node is also used to: settle the computing power resource transaction between the computing power demander and the computing power supplier to obtain the transaction data package, and send the transaction data package to the blockchain network broadcast.
  • computing power resource transactions refer to the computing power request and supply process between computing power demanders and computing power suppliers in the blockchain network.
  • the identification information of both parties to the transaction (such as name, digital signature, ID), transaction content, transaction time and other information in each computing power request supply process can be packaged to form a transaction data package.
  • Each transaction data packet can be broadcast in the blockchain network.
  • the settlement of the computing power resource transaction can also be carried out on the computing power supplier.
  • This disclosure achieves unified quantification of computing power through the evaluation module, solving the problem that it is difficult to uniformly measure the computing power of different algorithms and resource suppliers; it conducts a unified assessment of the computing power of the computing power demander and the resource supplier to provide computing power. Provides the basis for quantity and optimal matching.
  • This disclosure implements shared scheduling of computing power resources based on blockchain, and realizes shared scheduling based on computing power evaluation, improves resource utilization, and reduces network delay.
  • This disclosure evaluates the computing power of computing power requesters and computing power suppliers respectively, and performs computing resource matching and sharing scheduling on this basis to provide the best match for the demanders and suppliers of computing power resources and improve resource sharing. efficiency.
  • blockchain technology is used to realize the sharing of computing resources and privacy protection.
  • the embodiment of the present invention also provides a computing power scheduling method, which can be applied to the control node of the blockchain network, and can include the following steps S410 to S430.
  • Step S410 Receive the computing power resource request including business type information from the computing power requester; and evaluate the first computing power resource information of the computing power requesting party based on the business type information.
  • Step S420 Obtain the candidate computing power resource information of the computing power supplier, and evaluate the second computing power resource information of the computing power supplier based on the candidate computing power resource information.
  • Step S430 Determine the target computing power supply device based on the first computing power resource information and the second computing power resource information, and schedule the target computing power supply device to provide the target computing power resource to the computing power requester.
  • the computing power resource request includes identification information of the computing power requester
  • the target computing power resource includes identification information of the target computing power supply device.
  • evaluating the first computing power resource information of the computing power requester based on the business type information includes: determining a target computing power evaluation model based on the business type; using the target computing power evaluation model to calculate the computing power resource request. Capacity assessment to obtain the first computing power resource information.
  • the candidate computing power resource information includes hardware resource information, resource usage information, current period information and/or network resource information of the computing power supplier (that is, the candidate computing power resource information includes the hardware of the computing power supplier).
  • resource information, resource usage information, current period information, network resource information evaluate the second computing power resource information of the computing power supplier, including: hardware resource information, resource usage information and current period information and/ or perform mathematical operations on network resource information (that is, perform mathematical operations on at least one of hardware resource information, resource usage information, current period information, and network resource information) to evaluate the second computing power resource information of the computing power supplier .
  • determining the target computing power supply device based on the first computing power resource information and the second computing power resource information includes: determining the candidate computing power supply based on the first computing power resource information and the second computing power resource information. Equipment; according to the consensus matching rules in the smart contract, the target computing power supply equipment is selected from the candidate computing power supply equipment.
  • the method further includes: settling the computing power resource transaction between the computing power demander and the computing power supplier to obtain a transaction data package, and broadcasting the transaction data package to the blockchain network.
  • the blockchain network can include demand nodes (computing power requesters), supply nodes Node (computing power supplier), control node, consensus node, smart contract can be set in any node of the blockchain network, for example, smart contract can be set in the consensus node.
  • the number of nodes of each type can be one or more.
  • the method may include the following steps.
  • Step S501 The computing power requester sends a computing power resource request in the blockchain network.
  • the computing resource request can include the digital signature of the computing power requester and parameters such as required business data and business type.
  • Step S502 The control node receives the computing power resource request and calls the evaluation module to evaluate the computing power of the demand node and the supply node to obtain the evaluation results (first computing power resource information and second computing power resource information).
  • Step S503 The control node sends the evaluation result to the node where the smart contract is located (such as the consensus node).
  • Step S504 The node where the smart contract is located (such as the consensus node) triggers the smart contract to perform computing power matching and calculates the best matching target supply node (target computing power supply device).
  • Step S505 The node where the smart contract is located (such as the consensus node) sends the target supply node information to the demand node.
  • Step S506 The demand node sends a computing resource request to the target supply node.
  • Step S507 The target supply node verifies the computing power resource request. After the verification passes, the process proceeds to step S508.
  • the verification can be to verify the identity information of the demand node, or to verify the request content of the computing resource request, such as security verification, legality verification, etc.
  • Step S508 Provide required target computing power resources to the demand node.
  • Step S509 The target supply node propagates the resource sharing transaction on the blockchain network.
  • Step S510 the consensus node packages the transaction data and uploads the transaction data package to the blockchain network to implement on-chain certificate storage and provide guarantee for the traceability of the transaction.
  • this example embodiment also provides a control node 600.
  • the control node 600 includes: a first evaluation module 610, a second evaluation module 620 and a determination module 630; the first evaluation module 610 is used to receive a computing power request The party's computing power resource request includes business type information; and based on the business type information, evaluates the first computing power resource information of the computing power requester; the second evaluation module 620 is used to obtain the candidate computing power resource information of the computing power supplier, and Based on the candidate computing power resource information, evaluate the second computing power resource information of the computing power supplier; the determination module 630 is used to determine the target computing power supply equipment based on the first computing power resource information and the second computing power resource information to schedule the target
  • the computing power supply device provides target computing power resources to the computing power requester; among them, the computing power requester, the computing power supplier and the control node are located in the same blockchain network.
  • the computing power resource request includes identification information of the computing power requester
  • the target computing power resource includes identification information of the target computing power supply device.
  • the first evaluation module 610 is also configured to: determine a target computing power evaluation model based on the business type; use the target computing power evaluation model to perform computing power evaluation on the computing power resource request to obtain the first computing power evaluation model. human resource information.
  • the candidate computing resource information includes hardware resource information, resource usage information, current period information and/or network resource information of the computing power supplier.
  • the second evaluation module 620 is also used to: evaluate the hardware resources. Information, resource usage information, current period information and/or network resource information are subjected to mathematical operations to evaluate the second computing power resource information of the computing power supplier.
  • the determination module 630 includes a consensus node, which is used to: determine candidate computing power supply devices based on the first computing power resource information and the second computing power resource information; match according to the consensus in the smart contract Rules are used to match and filter out the target computing power supply equipment among the candidate computing power supply equipment.
  • the consensus node is also used to: settle computing resource transactions between computing power demanders and computing power suppliers to obtain transaction data packets, and broadcast the transaction data packets to the blockchain network .
  • this application also provides a computer-readable medium.
  • the computer-readable medium may be included in the device described in the above embodiments; it may also exist separately without being assembled into the device.
  • the above computer-readable medium carries one or more programs. When the above one or more programs are executed by a device, the device implements the method in the following embodiments. For example, the device can implement various steps as shown in Figure 4 and Figure 5, etc.
  • the computer-readable medium shown in the present disclosure may be a computer-readable signal medium or a computer-readable storage medium, or any combination of the above two.
  • the computer-readable storage medium may be, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or any combination thereof. More specific examples of computer readable storage media may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard drive, random access memory (RAM), read only memory (ROM), removable Programmd read-only memory (EPROM or flash memory), fiber optics, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the above.
  • a computer-readable storage medium may be any tangible medium that contains or stores a program for use by or in connection with an instruction execution system, apparatus, or device.
  • a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code therein. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above.
  • a computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device .
  • Program code embodied on a computer-readable medium may be transmitted using any suitable medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the foregoing.
  • a device capable of implementing the above method is also provided.
  • Those skilled in the art will understand that various aspects of the present disclosure may be implemented as systems, methods, or program products. Therefore, various aspects of the present disclosure may be embodied in the following forms, namely: a complete hardware embodiment, a complete software embodiment (including firmware, microcode, etc.), or an embodiment combining hardware and software aspects, which may be collectively referred to herein as "Circuits", “modules” or "systems”.
  • the computing power scheduling device 700 includes a processor 710, a memory 720, an input and output interface 730, and a communication bus 740.
  • the processor 710 is connected to the memory 720 and the input-output interface 730.
  • the processor 710 can be connected to the memory 720 and the input-output interface 730 through the communication bus 740.
  • the processor 710 is configured to support the computing power scheduling device to perform corresponding functions in the computing power scheduling methods in Figures 4 and 5.
  • the processor 710 may be a central processing unit (CPU), a network processor (Network Processor, NP), a hardware chip, or any combination thereof.
  • the above-mentioned hardware chip can be an application-specific integrated circuit (Application-Specific Integrated Circuit, ASIC), a programmable logic device (Programmable Logic Device, PLD) or a combination thereof.
  • the above-mentioned PLD can be a complex programmable logic device (Complex Programmable Logic Device, CPLD), a field-programmable gate array (Field-Programmable Gate Array, FPGA), a general array logic (Generic Array Logic, GAL) or any combination thereof.
  • Memory 720 is used to store program codes and the like.
  • Memory 720 may include volatile memory (VolatileMemory, VM), such as random access memory (Random Access Memory, RAM); memory 720 may also include non-volatile memory (Non-Volatile Memory, NVM), such as read-only memory. (Read-Only Memory, ROM), flash memory (flash memory), hard disk (Hard Disk Drive, HDD) or solid-state drive (Solid-State Drive, SSD); the memory 720 may also include a combination of the above types of memory.
  • volatile memory VolatileMemory, VM
  • RAM Random Access Memory
  • NVM non-volatile Memory
  • NVM non-volatile Memory
  • read-only memory Read-Only Memory
  • flash memory flash memory
  • HDD Hard Disk Drive
  • SSD solid-state drive
  • the input/output interface 730 is used to input or output data.
  • Processor 710 may call the program code described above to perform the following operations:
  • the party provides the target computing power resources; among them, the computing power requester and the computing power supplier are located in the same blockchain network.
  • the computing power resource request includes identification information of the computing power requester
  • the target computing power resource includes identification information of the target computing power supply device.
  • the above-mentioned processor 710 can also evaluate the first computing power resource information of the computing power requester based on the business type information, and perform the following operations: determine the target computing power evaluation model based on the business type; use the target computing power evaluation The model evaluates the computing power resource request to obtain the first computing power resource information.
  • the candidate computing power resource information includes hardware resource information, resource usage information, current period information and/or network resource information of the computing power supplier.
  • the processor 710 may also evaluate the second computing power provider's information. For computing power resource information, perform the following operations: perform mathematical operations on hardware resource information, resource usage information, current period information and/or network resource information to evaluate the computing power supplier's second computing power resource information.
  • the above-mentioned processor 710 can also determine the target computing power supply device based on the first computing power resource information and the second computing power resource information, and perform the following operations: based on the first computing power resource information and the second computing power resource information.
  • Resource information is used to determine the candidate computing power supply equipment; according to the consensus matching rules in the smart contract, the target computing power supply equipment is matched and screened among the candidate computing power supply equipment.
  • the above-mentioned processor 710 can also perform the following operations: settle the computing power resource transaction between the computing power demander and the computing power supplier to obtain the transaction data package, and send the transaction data package to the blockchain network. broadcast.
  • each operation can also correspond to the corresponding description with reference to the method embodiments shown in Figures 4 and 5; the above-mentioned processor 710 can also cooperate with the input and output interface 730 to perform other operations in the above-mentioned method embodiments.
  • the example embodiments described here can be implemented by software, or can be implemented by software combined with necessary hardware. Therefore, the technical solution according to the embodiment of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, U disk, mobile hard disk, etc.) or on the network , including a number of instructions to cause a device to perform a method according to an embodiment of the present disclosure.
  • a non-volatile storage medium which can be a CD-ROM, U disk, mobile hard disk, etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Computer Security & Cryptography (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Hardware Design (AREA)
  • Computing Systems (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

本公开提供一种算力调度系统和方法、控制节点、存储介质、设备；涉及网络通信技术领域。该系统包括：位于区块链网络中的算力请求方、算力供给方和控制节点，控制节点包括评估模块；算力请求方，用于向评估模块发送算力资源请求；评估模块，用于评估算力请求方的第一算力资源信息；同时获取算力供给方的候选算力资源信息，并评估第二算力资源信息；控制节点，用于基于评估结果确定目标算力供给设备，以调度为请求方提供算力。本公开可以解决相关技术中无法实现需求方和供给方的算力均衡匹配及算力资源利用率低的问题。

Description

算力调度系统和方法、控制节点、存储介质、设备
相关申请的交叉引用
本申请是以CN申请号为202210689060.3，申请日为2022年06月16日的申请为基础，并主张其优先权，该CN申请的公开内容在此作为整体引入本申请中。
技术领域
本公开涉及网络通信技术领域，具体而言，涉及一种算力调度系统和方法、控制节点、存储介质、设备。
背景技术
随着5G网络和物联网技术的快速发展，越来越多的设备接入互联网，产生了海量的数据和多样化的网络资源需求，算力资源是非常重要的网络资源，直接影响通信质量和网络时延。
相关技术中，存在部分计算设备由于数据处理量大而出现服务卡顿或网络时延高的现象，同时部分计算设备却空闲的情况。因此，对于需求方的任务，如何评估所需算力，并实现需求方和供给方的算力均衡匹配，以提高算力资源利用率是亟待解决的问题。
需要说明的是，在上述背景技术部分公开的信息仅用于加强对本公开的背景的理解，因此可以包括不构成对本领域普通技术人员已知的现有技术的信息。
发明内容
根据本公开的第一方面，提供了一种算力调度系统，系统包括：位于区块链网络中的算力请求方、算力供给方和控制节点，控制节点包括评估模块；算力请求方，用于向评估模块发送包括业务类型信息的算力资源请求；评估模块，用于基于接收的业务类型信息，评估算力请求方的第一算力资源信息；评估模块，还用于获取算力供给方的候选算力资源信息，并基于候选算力资源信息，评估算力供给方的第二算力资源信息；控制节点，用于基于评估模块评估的第一算力资源信息和第二算力资源信息，确定目标算力供给设备，以调度目标算力供给设备为算力请求方提供目标算力资源。
在一些实施例中，算力资源请求包括算力请求方的标识信息，目标算力资源包括目标算力供给设备的标识信息。
在一些实施例中，评估模块包括第一评估子模块，第一评估子模块用于：基于业务类 型，确定目标算力评估模型；采用目标算力评估模型对算力资源请求进行算力评估，以获得第一算力资源信息。
在一些实施例中，候选算力资源信息包括算力供给方的硬件资源信息、资源使用信息及当前时段信息和/或网络资源信息，评估模块还包括第二评估子模块，第二评估子模块用于：对硬件资源信息、资源使用信息及当前时段信息和/或网络资源信息进行数学运算，以评估算力供给方的第二算力资源信息。
在一些实施例中，区块链网络还包括智能合约，系统还包括共识节点，评估模块还用于：向共识节点发送第一算力资源信息和第二算力资源信息；共识节点用于：基于接收的第一算力资源信息和第二算力资源信息，确定候选算力供给设备；根据智能合约中的共识匹配规则，在候选算力供给设备中匹配筛选出目标算力供给设备。
在一些实施例中，算力供给方或共识节点还用于：对算力需求方和算力供给方进行算力资源交易的结算，以获得交易数据包，并将交易数据包向区块链网络广播。
根据本公开的第二方面，提供了一种算力调度方法，方法包括：接收算力请求方包括业务类型信息的算力资源请求；并基于业务类型信息，评估算力请求方的第一算力资源信息；获取算力供给方的候选算力资源信息，并基于候选算力资源信息，评估算力供给方的第二算力资源信息；基于第一算力资源信息和第二算力资源信息，确定目标算力供给设备，以调度目标算力供给设备为算力请求方提供目标算力资源；其中，算力请求方和算力供给方位于同一区块链网络中。
在一些实施例中，算力资源请求包括算力请求方的标识信息，目标算力资源包括目标算力供给设备的标识信息。
在一些实施例中，基于业务类型信息，评估算力请求方的第一算力资源信息，包括：基于业务类型，确定目标算力评估模型；采用目标算力评估模型对算力资源请求进行算力评估，以获得第一算力资源信息。
在一些实施例中，候选算力资源信息包括算力供给方的硬件资源信息、资源使用信息及当前时段信息和/或网络资源信息，评估算力供给方的第二算力资源信息，包括：对硬件资源信息、资源使用信息及当前时段信息和/或网络资源信息进行数学运算，以评估算力供给方的第二算力资源信息。
在一些实施例中，基于第一算力资源信息和第二算力资源信息，确定目标算力供给设备，包括：基于第一算力资源信息和第二算力资源信息，确定候选算力供给设备；根据智能合约中的共识匹配规则，在候选算力供给设备中匹配筛选出目标算力供给设备。
在一些实施例中，方法还包括：对算力需求方和算力供给方进行算力资源交易的结算，以获得交易数据包，并将交易数据包向区块链网络广播。
根据本公开的第三方面，提供了一种控制节点，控制节点包括：第一评估模块、第二 评估模块和确定模块；第一评估模块用于接收算力请求方包括业务类型信息的算力资源请求；并基于业务类型信息，评估算力请求方的第一算力资源信息；第二评估模块用于获取算力供给方的候选算力资源信息，并基于候选算力资源信息，评估算力供给方的第二算力资源信息；确定模块用于基于第一算力资源信息和第二算力资源信息，确定目标算力供给设备，以调度目标算力供给设备为算力请求方提供目标算力资源；其中，算力请求方、算力供给方和控制节点位于同一区块链网络中。
根据本公开的第四方面，提供一种计算机可读存储介质，其上存储有计算机程序，计算机程序被处理器执行时实现上述任意一项的方法。
根据本公开的第五方面，提供一种算力调度设备，包括：处理器；以及存储器，用于存储处理器的可执行指令；其中，处理器配置为经由执行可执行指令来执行上述任意一项的方法。
应当理解的是，以上的一般描述和后文的细节描述仅是示例性和解释性的，并不能限制本公开。
附图说明
此处的附图被并入说明书中并构成本说明书的一部分，示出了符合本公开的实施例，并与说明书一起用于解释本公开的原理。显而易见地，下面描述中的附图仅仅是本公开的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1示意性示出了根据本公开的一些实施例的算力调度系统的结构示意图之一。
图2示意性示出了根据本公开的一些实施例的评估模块的结构示意图。
图3示意性示出了根据本公开的一些实施例的算力调度系统的结构示意图之二。
图4示意性示出了根据本公开的一些实施例的算力调度方法的流程示意图之一。
图5示意性示出了根据本公开的一些实施例的算力调度方法的流程示意图之二。
图6示意性示出了根据本公开的一些实施例的示例性控制节点的结构框图。
图7示意性示出了根据本公开的一些实施例的示例性算力调度设备框图。
具体实施方式
现在将参考附图更全面地描述示例实施方式。然而，示例实施方式能够以多种形式实施，且不应被理解为限于在此阐述的范例；相反，提供这些实施方式使得本公开将更加全面和完整，并将示例实施方式的构思全面地传达给本领域的技术人员。所描述的特征、结 构或特性可以以任何合适的方式结合在一个或更多实施方式中。在下面的描述中，提供许多具体细节从而给出对本公开的实施方式的充分理解。然而，本领域技术人员将意识到，可以实践本公开的技术方案而省略特定细节中的一个或更多，或者可以采用其它的方法、组元、装置、步骤等。在其它情况下，不详细示出或描述公知技术方案以避免喧宾夺主而使得本公开的各方面变得模糊。
此外，附图仅为本公开的示意性图解，并非一定是按比例绘制。图中相同的附图标记表示相同或类似的部分，因而将省略对它们的重复描述。附图中所示的一些方框图是功能实体，不一定必须与物理或逻辑上独立的实体相对应。可以采用软件形式来实现这些功能实体，或在一个或多个硬件模块或集成电路中实现这些功能实体，或在不同网络和/或处理器装置和/或微控制器装置中实现这些功能实体。
本公开实施例的目的在于提供一种算力调度系统和方法、装置、存储介质、设备，进而在一定程度上解决了相关技术中无法实现需求方和供给方的算力均衡匹配及算力资源利用率低的问题。
本公开示例性实施例可以具有以下部分或全部有益效果：
在本公开示例实施方式所提供的算力调度系统中，一方面，通过设置评估模块实现了对算力请求方的请求算力的评估，同时，基于当前的算力供给方的候选算力资源信息，对算力供给方的算力资源进行评估，再通过控制节点基于评估模块的评估结果，对算力请求方和算力供给方的算力资源进行均衡匹配，确定目标算力供给设备，从而实现算力资源的统一度量和均衡调度，能够提高算力资源利用率。另一方面，本公开将算力请求方、算力供给方和控制节点设置于区块链网络中，即在区块链网络中实现算力的调度过程，使得区块链网络中的所有参与方能够及时获知当前和历史算力资源交易，实现算力资源的共享，可以进一步保证算力的均衡调度。此外，通过区块链网络可以实现算力资源的上链存证，便于后期维护排查。
参考图1，本公开一些实施例中提供的算力调度系统，可以部署于区块链网络中。该系统100可以包括位于区块链网络中的算力请求方110、算力供给方120和控制节点130，控制节点130包括评估模块131。
本公开的算力调度系统可以应用于终端计算场景、网络边缘计算场景和云计算场景。本公开中的算力请求方110、算力供给方120和控制节点130可以分别是一个或多个有独立地址且与具有传送或接收数据功能的网络相连的设备。该设备可以是工作站、移动终端、网络用户或个人计算机，还可以是服务器、打印机和其他与网络连接的设备，本示例对此不做限定。
算力请求方110，用于向评估模块发送包括业务类型信息的算力资源请求。
在本示例实施方式中，算力资源请求可以包括业务数据，即待处理数据和业务类型信 息，还可以包括算力请求方的标识信息(如数字签名或编码ID等)。算力资源请求还可以包括其他信息，如所需算力资源的获取时间或时限要求、距离要求等，本示例对此不做特殊限定。
在本示例实施方式中，业务类型信息可以根据业务数据的数据形式来确定。示例性地，当业务数据为图像时，业务类型信息可以是图像类；当业务数据为自然语言时，业务类型信息可以是自然语言类；当业务数据为音频数据时，业务类型可以是音频类。本示例中，业务类型信息还可以根据业务数据的数据量来确定。示例性地，当业务数据的数据量较大时(如大于某阈值或比例)，为图像时，业务类型信息可以是复杂类；当业务数据的数据量较小时(如小于某阈值或比例)，业务类型信息可以是简单类。类似地，还可以对业务数据类型进行其他划分，本示例对此不做特殊限定。
评估模块131，用于基于接收的业务类型信息，评估算力请求方110的第一算力资源信息。
在本示例实施方式中，第一算力资源信息是指算力请求方所需的算力资源信息。可以通过对算力资源请求中的业务数据进行分析评估，来确定第一算力资源信息。
评估模块131，还用于获取算力供给方120的候选算力资源信息，并基于候选算力资源信息，评估算力供给方120的第二算力资源信息。
在本示例实施方式中，候选算力资源信息是指算力供给方当前能够提供的算力资源信息。候选算力资源信息可以包括算力供给方的可用资源总量信息及其资源利用率情况等；可用资源总量信息可以包括CPU、内存的性能数据等硬件资源信息，还可以包括网络资源相关信息(如网络时延、带宽等)等，本示例对此不做限定。
在本示例实施方式中，第二算力资源信息是指经评估后能为算力请求方提供的算力资源。第二算力资源信息可以是根据硬件资源信息和/或网络资源信息确定的一个综合指标信息，如可以是对硬件资源信息和/或网络资源信息的加权运算。第二算力资源信息也可以是根据硬件资源信息和/或网络资源信息确定的等级信息，例如可以通过设置硬件资源信息和/或网络资源信息与等级的映射关系来确定第二算力资源信息。第二算力资源信息可以是一个或多个数值，本示例对此不做限定。
控制节点130，用于基于评估模块131评估的第一算力资源信息和第二算力资源信息，确定目标算力供给设备，以调度目标算力供给设备为算力请求方110提供目标算力资源。
在本示例实施方式中，可以根据预置的匹配规则，对第一算力资源信息和第二算力资源信息进行匹配，确定算力供给方的目标算力供给设备。在一些实施例中，目标算力资源可以包括目标算力供给设备的标识信息，该标识信息可以是签名(如数字签名)或编码ID等，本示例对此不做限定。
本公开实施例提供的故障原因确定方法，一方面，通过设置评估模块实现了对算力请 求方的请求算力的评估，同时，基于当前的算力供给方的候选算力资源信息，对算力供给方的算力资源进行评估，再通过控制节点基于评估模块的评估结果，对算力请求方和算力供给方的算力资源进行均衡匹配，确定目标算力供给设备，从而实现算力资源的统一度量和均衡调度，能够提高算力资源利用率。另一方面，本公开将算力请求方、算力供给方和控制节点设置于区块链网络中，即在区块链网络中实现算力的调度过程，使得区块链网络中的所有参与方能够及时获知当前和历史算力资源交易，实现算力资源的共享，可以进一步保证算力的均衡调度。此外，通过区块链网络可以实现算力资源的上链存证，便于后期维护排查。
在一些实施例中，参考图2，评估模块200包括第一评估子模块210和第二评估子模块220中的至少一个。
第一评估子模块210用于：基于业务类型，确定目标算力评估模型；采用目标算力评估模型对算力资源请求进行算力评估，以获得第一算力资源信息。
在本示例实施方式中，可以对不同业务类型使用不同的算力评估模型。如当业务类型为图像处理时，可以使用卷积神经网络模型；当业务类型为自然语音处理时，可以使用递归神经网络模型；当业务类型是音频处理时，可以使用增加语音识别功能的递归神经网络模型；当业务类型为视频处理时，可以使用具有时序特征提取功能的卷积神经网络模型。
在本示例实施方式中，可以基于业务类型和业务数据的数据量、处理难易程度(如图像处理难度较高、数据类处理难度较低)、处理时延要求等信息，综合确定目标算力评估模型。目标算力评估模型是经过训练的各类神经网络模型。示例性地，可以将业务类型和业务数据的数据量、处理难易程度、处理时延要求等中的一种或多种信息与算力评估模型之间建立对应关系，从而快速确定目标算力评估模型。目标算力评估模型中的参数可以进行定期或动态更新，本示例对此不做限定。
在本示例实施方式中，可以将算力资源请求中的业务数据作为目标算力评估模型的输入数据，经过模型的算力评估，输出该请求对应的第一算力资源信息。第一算力资源信息可以包括所需CPU相关信息和内存相关信息，如，CPU主频率、一级缓存和内存的频率等信息，还可以包括总线位宽、字节常数等硬件信息。第一算力资源信息还可以包括网络时延要求等网络资源信息，也可以包括其他资源信息，本示例对此不做限定。
第二评估子模块220用于：对硬件资源信息、资源使用信息及当前时段信息和/或网络资源信息进行数学运算(也即，对硬件资源信息、资源使用信息、当前时段信息、网络资源信息中的至少一项进行数学运算)，以评估算力供给方的第二算力资源信息。
在本示例实施方式中，硬件资源信息可以包括CPU信息、内存信息、缓存信息和/或总线位宽、字节常数等信息，资源使用率信息可以包括每个硬件资源的使用情况，如可以采用已用内存与总内存的比值作为内存使用率信息。当前时段信息可以包括当前时刻信 息、星期信息、月份信息等，也可以包括工作时段和休息时段，高峰时段和普通时段等，本示例对此不做限定。可以使用不同标识信息或编码来标识不同时段，该标识信息或编码可以根据经验设置。
在本示例实施方式中，网络资源信息可以包括网络传输时延和/或响应时延，还可以包括网络带宽等信息。数学运算可以是简单四则运算或加权四则运算，也可以是复杂的函数运算(如对数函数、三角函数)、积分、微分等运算，或者各种运算的结合，本示例对此不做特殊限定。本示例中，第二算力资源信息可以是一个算力指标值，也可以是多个算力指标值的相关运算结果，本示例对此不做限定。
示例性地，以硬件资源中的CPU和内存资源为例进行说明。算力供给方的某些节点的CPU资源总量为C res、CPU资源利用率为C uti、内存资源总量为M res、内存资源利用率为M uti、当前时段信息为T，则第二算力资源信息的评估公式可以是：
Power＝(α×C res×(1-C uti)+β×M res×(1-M uti))×T；
式中，α、β分别为预设的权重参数，α表示CPU权重，β表示内存资源权重，α、β的取值可以是任意自然数，可以根据经验进行设置。α、β也可以随业务类型的不同而进行不同设置，本示例对此不做限定。
在一些实施例中，参考图3，区块链网络还包括智能合约，系统除了包括算力请求方、算力供给方、控制节点之外，还包括共识节点，控制节点包括评估模块。评估模块还用于：向共识节点发送第一算力资源信息和第二算力资源信息。
在本示例实施方式中，共识节点可以是不同于控制节点的其他节点。共识节点可以是一个或多个有独立地址且与具有传送或接收数据功能的网络相连的设备。该设备可以是工作站、移动终端、网络用户或个人计算机，还可以是服务器、打印机和其他与网络连接的设备，本示例对此不做限定。
在本示例实施方式中，评估模块的评估结果可以发送给共识节点进行算力资源匹配。智能合约是区块链网络中的各参与方都认可的规则的集合。智能合约可以设置于区块链网络中的任意节点中。例如，智能合约可以设置于共识节点中。
共识节点用于：基于接收的第一算力资源信息和第二算力资源信息，确定候选算力供给设备；根据智能合约中的共识匹配规则，在候选算力供给设备中匹配筛选出目标算力供给设备。
在本示例实施方式中，第一算力资源信息可以是一个或多个算力指标值，同样，第二算力资源信息可以包括算力供给方中每个设备/节点的算力指标值。示例性地，第一算力资源信息为Power1，第二算力资源信息为Power2，则可以将所有大于Power1的Power2中的设备/节点的组合形成的集合作为候选算力供给设备，例如，第二算力资源信息中设备A、B、C对应的算力资源可以满足Power1，则可以将设备A、B、C作为候选算力供给设 备的一个设备组合。也可以在算力满足的情况下，设置算力请求方与算力供给方的距离条件来确定候选算力供给设备，还可以设置其他条件，本示例对此不做限定。
在本示例实施方式中，共识匹配规则是指区块链网络中的各参与方达成共识性算力匹配规则。示例性地，可以根据算力资源和/或算力请求方与算力供给方的距离条件来设置共识匹配规则。例如，可以设置在满足需求的算力资源的前提下(如大于或等于请求方的算力资源)，可以选择算力最小值的设备/节点或设备/节点组合作为目标算力供给设备，从而减少算力资源浪费。也可以设置在满足需求的算力资源的前提下，可以选择距离最小值的设备/节点或设备/节点组合作为目标算力供给设备，从而减少时延。还可以将算力大小和距离因素结合起来选择目标算力供给设备，本示例对此不做特殊限定。
在一些实施例中，参考图3，共识节点还用于：对算力需求方和算力供给方进行算力资源交易的结算，以获得交易数据包，并将交易数据包向区块链网络广播。
在本示例实施方式中，算力资源交易是指区块链网络中的算力需求方和算力供给方之间的算力请求与供给过程。可以将每次算力请求供给过程中的交易双方标识信息(如名称、数字签名、ID)、交易内容和交易时间等信息进行打包形成交易数据包。可以将每次的交易数据包在区块链网络中广播。同样的，该算力资源交易的结算也可以在算力供给方进行。
本公开通过评估模块对算力实现了统一量化，解决了不同算法、资源供给方的算力较难统一度量的问题；对算力需求方和资源供给方的算力进行统一评估，为算力度量和最佳匹配提供基础。
本公开基于区块链的实现了算力资源的共享调度，在算力评估的基础上实现共享调度，提高了资源利用率，降低网络时延。
本公开对算力请求方和算力供给方的算力分别进行评估，在此基础上进行算力资源匹配和共享调度，为算力资源的需求方和供给方提供最佳匹配，提升资源共享效率。同时，使用区块链技术，实现算力资源共享的确权和隐私保护。
本发明实施例还提供了一种算力调度方法，可以应用于区块链网络的控制节点，可以包括以下步骤S410～S430。
步骤S410，接收算力请求方包括业务类型信息的算力资源请求；并基于业务类型信息，评估算力请求方的第一算力资源信息。
步骤S420，获取算力供给方的候选算力资源信息，并基于候选算力资源信息，评估算力供给方的第二算力资源信息。
步骤S430，基于第一算力资源信息和第二算力资源信息，确定目标算力供给设备，以调度目标算力供给设备为算力请求方提供目标算力资源。
在一些实施例中，算力资源请求包括算力请求方的标识信息，目标算力资源包括目标算力供给设备的标识信息。
在一些实施例中，基于业务类型信息，评估算力请求方的第一算力资源信息，包括：基于业务类型，确定目标算力评估模型；采用目标算力评估模型对算力资源请求进行算力评估，以获得第一算力资源信息。
在一些实施例中，候选算力资源信息包括算力供给方的硬件资源信息、资源使用信息及当前时段信息和/或网络资源信息(也即，候选算力资源信息包括算力供给方的硬件资源信息、资源使用信息、当前时段信息、网络资源信息中的至少一项)，评估算力供给方的第二算力资源信息，包括：对硬件资源信息、资源使用信息及当前时段信息和/或网络资源信息进行数学运算(也即，对硬件资源信息、资源使用信息及当前时段信息、网络资源信息中的至少一项进行数学运算)，以评估算力供给方的第二算力资源信息。
在一些实施例中，基于第一算力资源信息和第二算力资源信息，确定目标算力供给设备，包括：基于第一算力资源信息和第二算力资源信息，确定候选算力供给设备；根据智能合约中的共识匹配规则，在候选算力供给设备中匹配筛选出目标算力供给设备。
在一些实施例中，方法还包括：对算力需求方和算力供给方进行算力资源交易的结算，以获得交易数据包，并将交易数据包向区块链网络广播。
举例而言，参考图5，为本公开的一种算力调度方法的具体实现过程，该方法可以在区块链网络中实现，区块链网络可以包括需求节点(算力请求方)、供给节点(算力供给方)、控制节点、共识节点，智能合约可以设置于区块链网络的任意节点中，如智能合约可以设置于共识节点。每类节点的数量均可以是一个或多个。该方法可以包括以下步骤。
步骤S501，算力请求方在区块链网络中发送算力资源请求。
本示例中，算力资源请求可以包含算力请求方的数字签名和所需业务数据、业务类型等参数。
步骤S502，控制节点接收算力资源请求，并调用评估模块对需求节点和供给节点进行算力评估，以获得评估结果(第一算力资源信息和第二算力资源信息)。
步骤S503，控制节点将评估结果发送至智能合约所在节点(如共识节点)。
步骤S504，智能合约所在节点(如共识节点)触发智能合约进行算力匹配，计算最佳匹配的目标供给节点(目标算力供给设备)。
步骤S505，智能合约所在节点(如共识节点)将目标供给节点信息发送给需求节点。
步骤S506，需求节点向目标供给节点发送算力资源请求。
步骤S507，目标供给节点对算力资源请求进行校验，校验通过后，转入步骤S508。
在本示例中，校验可以是对需求节点的身份信息进行校验，还可以是对算力资源请求的请求内容进行校验，如安全性校验、合法性校验等。
步骤S508，向需求节点提供所需目标算力资源。
步骤S509，目标供给节点将该资源共享交易在区块链网络上传播。
步骤S510，共识节点将交易数据进行打包，并将交易数据包上传到区块链网络中，实现上链存证，为交易的可追溯性提供保障。
上述实施例中的算力调度方法中涉及的各个步骤的具体细节已经在对应的算力调度系统中进行了详细的描述，因此此处不再赘述。
参见图6，本示例实施方式中还提供了一种控制节点600，控制节点600包括：第一评估模块610、第二评估模块620和确定模块630；第一评估模块610用于接收算力请求方包括业务类型信息的算力资源请求；并基于业务类型信息，评估算力请求方的第一算力资源信息；第二评估模块620用于获取算力供给方的候选算力资源信息，并基于候选算力资源信息，评估算力供给方的第二算力资源信息；确定模块630用于基于第一算力资源信息和第二算力资源信息，确定目标算力供给设备，以调度目标算力供给设备为算力请求方提供目标算力资源；其中，算力请求方、算力供给方和控制节点位于同一区块链网络中。
在本公开的一些实施例中，算力资源请求包括算力请求方的标识信息，目标算力资源包括目标算力供给设备的标识信息。
在本公开的一些实施例中，第一评估模块610还用于：基于业务类型，确定目标算力评估模型；采用目标算力评估模型对算力资源请求进行算力评估，以获得第一算力资源信息。
在本公开的一些实施例中，候选算力资源信息包括算力供给方的硬件资源信息、资源使用信息及当前时段信息和/或网络资源信息，第二评估模块620还用于：对硬件资源信息、资源使用信息及当前时段信息和/或网络资源信息进行数学运算，以评估算力供给方的第二算力资源信息。
在本公开的一些实施例中，确定模块630包括共识节点，共识节点用于：基于第一算力资源信息和第二算力资源信息，确定候选算力供给设备；根据智能合约中的共识匹配规则，在候选算力供给设备中匹配筛选出目标算力供给设备。
在本公开的一些实施例中，共识节点还用于：对算力需求方和算力供给方进行算力资源交易的结算，以获得交易数据包，并将交易数据包向区块链网络广播。
上述实施例中的控制节点中涉及的各个模块/节点的具体细节已经在对应的算力调度系统中进行了详细的描述，因此此处不再赘述。
作为另一方面，本申请还提供了一种计算机可读介质，该计算机可读介质可以是上述 实施例中描述的设备中所包含的；也可以是单独存在，而未装配入该设备中。上述计算机可读介质承载有一个或者多个程序，当上述一个或者多个程序被一个该设备执行时，使得该设备实现如下述实施例中的方法。例如，设备可以实现如图4和图5所示的各个步骤等。
需要说明的是，本公开所示的计算机可读介质可以是计算机可读信号介质或者计算机可读存储介质或者是上述两者的任意组合。计算机可读存储介质例如可以是——但不限于——电、磁、光、电磁、红外线、或半导体的系统、装置或器件，或者任意以上的组合。计算机可读存储介质的更具体的例子可以包括但不限于：具有一个或多个导线的电连接、便携式计算机磁盘、硬盘、随机访问存储器(RAM)、只读存储器(ROM)、可擦式可编程只读存储器(EPROM或闪存)、光纤、便携式紧凑磁盘只读存储器(CD-ROM)、光存储器件、磁存储器件、或者上述的任意合适的组合。在本公开中，计算机可读存储介质可以是任何包含或存储程序的有形介质，该程序可以被指令执行系统、装置或者器件使用或者与其结合使用。而在本公开中，计算机可读的信号介质可以包括在基带中或者作为载波一部分传播的数据信号，其中承载了计算机可读的程序代码。这种传播的数据信号可以采用多种形式，包括但不限于电磁信号、光信号或上述的任意合适的组合。计算机可读的信号介质还可以是计算机可读存储介质以外的任何计算机可读介质，该计算机可读介质可以发送、传播或者传输用于由指令执行系统、装置或者器件使用或者与其结合使用的程序。计算机可读介质上包含的程序代码可以用任何适当的介质传输，包括但不限于：无线、电线、光缆、RF等等，或者上述的任意合适的组合。
此外，在本公开的示例性实施例中，还提供了一种能够实现上述方法的设备。所属技术领域的技术人员能够理解，本公开的各个方面可以实现为系统、方法或程序产品。因此，本公开的各个方面可以具体实现为以下形式，即：完全的硬件实施例、完全的软件实施例(包括固件、微代码等)，或硬件和软件方面结合的实施例，这里可以统称为“电路”、“模块”或“系统”。
参见图7，图7是本申请实施例提供的一种算力调度设备的结构示意图。如图7所示，该算力调度设备700包括处理器710、存储器720、输入输出接口730以及通信总线740。处理器710连接到存储器720和输入输出接口730，例如处理器710可以通过通信总线740连接到存储器720和输入输出接口730。处理器710被配置为支持该算力调度设备执行图4和图5中算力调度方法中相应的功能。该处理器710可以是中央处理器(Central Processing Unit，CPU)，网络处理器(Network Processor，NP)，硬件芯片或者其任意组合。上述硬件芯片可以是专用集成电路(Application-Specific Integrated Circuit，ASIC)，可编程逻辑器件(Programmable Logic Device，PLD)或其组合。上述PLD可以是复杂可编程逻辑器件(Complex Programmable Logic Device，CPLD)，现场可编程逻辑门阵列(Field-Programmable Gate Array，FPGA)，通用阵列逻辑(Generic Array Logic，GAL)或其任意组合。存储器720用于存储程序代码等。存储器720可以包括易失性存储器(VolatileMemory，VM)，例如随机存取存储器(Random Access Memory，RAM)；存储器720 也可以包括非易失性存储器(Non-Volatile Memory，NVM)，例如只读存储器(Read-Only Memory，ROM)，快闪存储器(flash memory)，硬盘(Hard Disk Drive，HDD)或固态硬盘(Solid-State Drive，SSD)；存储器720还可以包括上述种类的存储器的组合。
该输入输出接口730用于输入或输出数据。
处理器710可以调用上述程序代码以执行以下操作：
接收算力请求方包括业务类型信息的算力资源请求；并基于业务类型信息，评估算力请求方的第一算力资源信息；获取算力供给方的候选算力资源信息，并基于候选算力资源信息，评估算力供给方的第二算力资源信息；基于第一算力资源信息和第二算力资源信息，确定目标算力供给设备，以调度目标算力供给设备为算力请求方提供目标算力资源；其中，算力请求方和算力供给方位于同一区块链网络中。
在一些实施例中，上述算力资源请求包括算力请求方的标识信息，上述目标算力资源包括目标算力供给设备的标识信息。
在一些实施例中，上述处理器710还可以基于业务类型信息，评估算力请求方的第一算力资源信息，执行以下操作：基于业务类型，确定目标算力评估模型；采用目标算力评估模型对算力资源请求进行算力评估，以获得第一算力资源信息。
在一些实施例中，上述候选算力资源信息包括算力供给方的硬件资源信息、资源使用信息及当前时段信息和/或网络资源信息，上述处理器710还可以评估算力供给方的第二算力资源信息，执行以下操作：对硬件资源信息、资源使用信息及当前时段信息和/或网络资源信息进行数学运算，以评估算力供给方的第二算力资源信息。
在一些实施例中，上述处理器710还可以基于第一算力资源信息和第二算力资源信息，确定目标算力供给设备，执行以下操作：基于第一算力资源信息和第二算力资源信息，确定候选算力供给设备；根据智能合约中的共识匹配规则，在候选算力供给设备中匹配筛选出目标算力供给设备。
在一些实施例中，上述处理器710还可以执行以下操作：对算力需求方和算力供给方进行算力资源交易的结算，以获得交易数据包，并将交易数据包向区块链网络广播。
需要说明的是，各个操作的实现还可以对应参照图4和图5所示的方法实施例的相应描述；上述处理器710还可以与输入输出接口730配合执行上述方法实施例中的其他操作。
通过以上的实施例的描述，本领域的技术人员易于理解，这里描述的示例实施例可以通过软件实现，也可以通过软件结合必要的硬件的方式来实现。因此，根据本公开实施例的技术方案可以以软件产品的形式体现出来，该软件产品可以存储在一个非易失性存储介质(可以是CD-ROM，U盘，移动硬盘等)中或网络上，包括若干指令以使得一台设备执行根据本公开实施例的方法。
此外，上述附图仅是根据本公开示例性实施例的方法所包括的处理的示意性说明，而不是限制目的。易于理解，上述附图所示的处理并不表明或限制这些处理的时间顺序。另外，也易于理解，这些处理可以是例如在多个模块中同步或异步执行的。
需要说明的是，尽管在附图中以特定顺序描述了本公开中方法的各个步骤，但是，这并非要求或者暗示必须按照该特定顺序来执行这些步骤，或是必须执行全部所示的步骤才能实现期望的结果。附加的或备选的，可以省略某些步骤，将多个步骤合并为一个步骤执行，以及/或者将一个步骤分解为多个步骤执行等，均应视为本公开的一部分。
应可理解的是，本说明书公开和限定的本公开延伸到文中和/或附图中提到或明显的两个或两个以上单独特征的所有可替代组合。所有这些不同的组合构成本公开的多个可替代方面。本说明书的实施方式说明了已知用于实现本公开的最佳方式，并且将使本领域技术人员能够利用本公开。

Claims (17)

  1. 一种算力调度系统，所述系统包括：位于区块链网络中的算力请求方、算力供给方和控制节点，所述控制节点包括评估模块；
    所述算力请求方，用于向所述评估模块发送包括业务类型信息的算力资源请求；
    所述评估模块，用于基于接收的所述业务类型信息，评估所述算力请求方的第一算力资源信息；
    所述评估模块，还用于获取所述算力供给方的候选算力资源信息，并基于所述候选算力资源信息，评估所述算力供给方的第二算力资源信息；
    所述控制节点，用于基于所述评估模块评估的所述第一算力资源信息和所述第二算力资源信息，确定目标算力供给设备，以调度所述目标算力供给设备为所述算力请求方提供目标算力资源。
  2. 根据权利要求1所述的算力调度系统，其中，所述算力资源请求包括算力请求方的标识信息、所需算力资源的获取时间、时限要求、距离要求中的至少一项，所述目标算力资源包括目标算力供给设备的标识信息。
  3. 根据权利要求1所述的算力调度系统，其中，所述评估模块包括第一评估子模块，所述第一评估子模块用于：
    基于所述业务类型信息，确定目标算力评估模型；
    采用所述目标算力评估模型对所述算力资源请求进行算力评估，以获得第一算力资源信息。
  4. 根据权利要求1或3所述的算力调度系统，其中，所述候选算力资源信息包括算力供给方的硬件资源信息、资源使用信息、当前时段信息、网络资源信息中的至少一项，所述评估模块还包括第二评估子模块，所述第二评估子模块用于：
    对所述硬件资源信息、所述资源使用信息、所述当前时段信息、所述网络资源信息中的至少一项进行数学运算，以评估算力供给方的第二算力资源信息。
  5. 根据权利要求1所述的算力调度系统，其中，所述区块链网络还包括智能合约，所述系统还包括共识节点，所述评估模块还用于：向所述共识节点发送所述第一算力资源信息和所述第二算力资源信息；
    所述共识节点用于：
    基于接收的所述第一算力资源信息和所述第二算力资源信息，确定候选算力供给 设备；
    根据所述智能合约中的共识匹配规则，在所述候选算力供给设备中匹配筛选出目标算力供给设备。
  6. 根据权利要求5所述的算力调度系统，其中，所述算力供给方或所述共识节点还用于：
    对所述算力需求方和所述算力供给方进行算力资源交易的结算，以获得交易数据包，并将所述交易数据包向区块链网络广播。
  7. 一种算力调度方法，所述方法包括：
    接收算力请求方包括业务类型信息的算力资源请求；并基于所述业务类型信息，评估算力请求方的第一算力资源信息；
    获取算力供给方的候选算力资源信息，并基于所述候选算力资源信息，评估算力供给方的第二算力资源信息；
    基于所述第一算力资源信息和所述第二算力资源信息，确定目标算力供给设备，以调度所述目标算力供给设备为算力请求方提供目标算力资源；
    其中，所述算力请求方和所述算力供给方位于同一区块链网络中。
  8. 根据权利要求7所述的算力调度方法，其中，所述算力资源请求包括算力请求方的标识信息、所需算力资源的获取时间、时限要求、距离要求中的至少一项，所述目标算力资源包括目标算力供给设备的标识信息。
  9. 根据权利要求7所述的算力调度方法，其中，所述基于所述业务类型信息，评估算力请求方的第一算力资源信息，包括：
    基于所述业务类型信息，确定目标算力评估模型；
    采用所述目标算力评估模型对所述算力资源请求进行算力评估，以获得第一算力资源信息。
  10. 根据权利要求7或9所述的算力调度方法，其中，所述候选算力资源信息包括算力供给方的硬件资源信息、资源使用信息、当前时段信息、网络资源信息中的至少一项，所述评估算力供给方的第二算力资源信息，包括：
    对所述硬件资源信息、所述资源使用信息及所述当前时段信息、所述网络资源信息中的至少一项进行数学运算，以评估算力供给方的第二算力资源信息。
  11. 根据权利要求7所述的算力调度方法，其中，所述基于所述第一算力资源信息和所述第二算力资源信息，确定目标算力供给设备，包括：
    基于所述第一算力资源信息和所述第二算力资源信息，确定候选算力供给设备；
    根据所述智能合约中的共识匹配规则，在所述候选算力供给设备中匹配筛选出目标算力供给设备。
  12. 根据权利要求7所述的算力调度方法，所述方法还包括：
    对所述算力需求方和所述算力供给方进行算力资源交易的结算，以获得交易数据包，并将所述交易数据包向区块链网络广播。
  13. 根据权利要求7所述的算力调度方法，其中，
    所述业务类型信息根据业务数据的数据形式或者数据量来确定。
  14. 根据权利要求9所述的算力调度方法，其中，基于所述业务类型信息，确定目标算力评估模型包括：
    基于所述业务类型信息和业务数据的数据量、处理难易程度、处理时延要求中的一种或多种信息，综合确定目标算力评估模型。
  15. 一种计算机可读存储介质，其上存储有计算机程序，所述计算机程序被处理器执行时实现权利要求7-11、12-14任一项所述的方法。
  16. 一种算力调度设备，包括：
    处理器；以及
    存储器，用于存储所述处理器的可执行指令；其中，所述处理器配置为经由执行所述可执行指令来执行权利要求7-11、12-14任一项所述的方法。
  17. 一种计算机程序，包括：
    指令，所述指令由处理器执行时使所述处理器执行根据权利要求7-14中任一项所述的方法。
PCT/CN2022/130589 2022-06-16 2022-11-08 算力调度系统和方法、控制节点、存储介质、设备 WO2023240911A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210689060.3 2022-06-16
CN202210689060.3A CN117294702A (zh) 2022-06-16 2022-06-16 算力调度系统和方法、控制节点、存储介质、设备

Publications (1)

Publication Number Publication Date
WO2023240911A1 true WO2023240911A1 (zh) 2023-12-21

Family

ID=89193046

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/130589 WO2023240911A1 (zh) 2022-06-16 2022-11-08 算力调度系统和方法、控制节点、存储介质、设备

Country Status (2)

Country Link
CN (1) CN117294702A (zh)
WO (1) WO2023240911A1 (zh)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108323200A (zh) * 2018-01-25 2018-07-24 深圳前海达闼云端智能科技有限公司 基于区块链的数据训练方法、装置、存储介质及区块链节点
CN110532092A (zh) * 2019-08-21 2019-12-03 云湾科技(嘉兴)有限公司 资源匹配方法、装置及设备
CN112132447A (zh) * 2020-09-21 2020-12-25 江苏省未来网络创新研究院 一种基于区块链的算力网络信任评估与保障算法
CN113918316A (zh) * 2021-08-30 2022-01-11 山东浪潮科学研究院有限公司 基于区块链的5g接入网用户端边缘算力分配方法
CN114095577A (zh) * 2020-07-31 2022-02-25 中国移动通信有限公司研究院 资源请求方法、装置、算力网元节点及算力应用设备

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108323200A (zh) * 2018-01-25 2018-07-24 深圳前海达闼云端智能科技有限公司 基于区块链的数据训练方法、装置、存储介质及区块链节点
CN110532092A (zh) * 2019-08-21 2019-12-03 云湾科技(嘉兴)有限公司 资源匹配方法、装置及设备
CN114095577A (zh) * 2020-07-31 2022-02-25 中国移动通信有限公司研究院 资源请求方法、装置、算力网元节点及算力应用设备
CN112132447A (zh) * 2020-09-21 2020-12-25 江苏省未来网络创新研究院 一种基于区块链的算力网络信任评估与保障算法
CN113918316A (zh) * 2021-08-30 2022-01-11 山东浪潮科学研究院有限公司 基于区块链的5g接入网用户端边缘算力分配方法

Also Published As

Publication number Publication date
CN117294702A (zh) 2023-12-26

Similar Documents

Publication Publication Date Title
US11751004B2 (en) Methods and systems for communication management
CN107360032B (zh) 一种网络流识别方法及电子设备
CN109033404B (zh) 日志数据处理方法、装置和系统
US20150304441A1 (en) Data transfer device and data transfer system using adaptive compression algorithm
US11411865B2 (en) Network resource scheduling method, apparatus, electronic device and storage medium
US10608915B2 (en) Providing dynamic latency in an integration flow
WO2024098872A1 (zh) 任务处理方法、系统、装置、设备及计算机可读存储介质
CN115460216A (zh) 算力资源调度方法和装置、算力资源调度设备、系统
CN113938394B (zh) 监控业务带宽分配方法、装置、电子设备及存储介质
CN113032410A (zh) 数据处理方法、装置、电子设备及计算机存储介质
WO2023240911A1 (zh) 算力调度系统和方法、控制节点、存储介质、设备
Cleland et al. FedComm: Understanding communication protocols for edge-based federated learning
Huang et al. Modeling and analysis on congestion control for data transmission in sensor clouds
CN114448838B (zh) 系统可靠度评估方法
CN115576973A (zh) 一种业务部署方法、装置、计算机设备和可读存储介质
CN111984612B (zh) 一种共享文件的方法、装置和系统
CN114219338A (zh) 基于联合学习的资源分配方法及装置
CN112714037A (zh) 一种线上服务质量的保障性能评估方法、装置及设备
CN110119364B (zh) 一种输入/输出批量提交的方法和系统
CN113626175A (zh) 数据处理的方法和装置
CN116137630B (zh) 网络业务需求的量化处理方法和装置
CN115379243B (zh) Cdn的调度方法、装置、设备及存储介质
CN111092826B (zh) 一种网络接入控制方法、系统及设备
CN116566992B (zh) 边缘计算的动态协同方法、装置、计算机设备和存储介质
CN114201493B (zh) 数据接入方法、装置、设备以及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22946577

Country of ref document: EP

Kind code of ref document: A1