Data processing method and device, storage medium and electronic equipment

Abstract

The invention provides a data processing method, a data processing device, a computer storage medium and electronic equipment, and relates to the technical field of communication. The method comprises the following steps: in response to receiving a service request sent by user equipment for a target service, performing demand analysis on the service request based on a session management function (SMF) network element to obtain a service type demand of the target service; the SMF network element determines a to-be-coordinated network element based on the service type demand; a to-be-coordinated network element receives a service request and a service quality demand parameter corresponding to a target service, and obtains a node state parameter of an initial computing node matched with the to-be-coordinated network element, thereby determining a target computing node from the initial computing node based on the service request, the node state parameter and the service quality demand parameter, and sending the target computing node to the to-be-coordinated network element. And providing services for the target business based on the target computing node. According to the invention, multi-element cooperation of communication, calculation, intelligence and the like can be realized, and then integrated fusion services of communication services, calculation services, intelligent services and the like are provided for users.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of communication, and particularly relates to a data processing method, a data processing device, a computer storage medium and an electronic device. BACKGROUND

[0002] Facing more and more emerging scenarios such as the fusion of artificial intelligence and communication, the fusion of perception and communication, etc., the 6th Generation Mobile Networks (6G) network must go beyond the ability of traditional communication networks, deeply integrate perception, computing, intelligence and other ability elements, and develop towards an integrated network of communication services, perception services, computing services and intelligent services, so as to provide integrated services on demand.

[0003] At present, the existing 5G network is mainly realized through the ability outside the mobile network (for example, through the Internet transmission data or service (Over The Top, OTT) / computing power network, etc.).

[0004] However, in the above method, the mobile network only provides connection and cannot perceive upper-layer intelligent services and computing services, so that the intelligent resources and computing resources in the communication network cannot be fully utilized, and the multi-element collaborative mechanism cannot be realized, resulting in the difficulty in realizing the integrated intelligent algorithm fusion service of communication services, computing services and intelligent services provided by the communication network. SUMMARY

[0005] The present disclosure provides a data processing method, a data processing device, a computer storage medium and an electronic device, so that the intelligent resources and computing resources in the communication network can be fully utilized, and the integrated intelligent algorithm fusion service of communication services, computing services and intelligent services provided by the communication network can be realized.

[0006] In a first aspect, an embodiment of the present disclosure provides a data processing method, which comprises: in response to receiving a service request sent by a user equipment for a target service, performing demand analysis on the service request based on a session management function (SMF) network element to obtain a service type required by the target service; determining a to-be-coordinated network element based on the service type by the SMF network element; receiving, by the to-be-coordinated network element, the service request and a service quality demand parameter corresponding to the target service, and obtaining a node state parameter of an initial computing node matched with the to-be-coordinated network element; determining, by the to-be-coordinated network element, a target computing node from the initial computing node based on the service request, the node state parameter and the service quality demand parameter, so as to provide the target service based on the target computing node.

[0007] In a second aspect, an embodiment of the present disclosure provides a data processing apparatus, comprising: a requirement analysis module configured to perform requirement analysis on a service request based on a session management function (SMF) network element, to obtain a service type required by a target service in response to receiving the service request sent by a user equipment for the target service; a network element determination module configured to perform determination of a to-be-cooperated network element based on the service type by the SMF network element; a parameter acquisition module configured to perform acquisition of a node state parameter of an initial computing node matched with the to-be-cooperated network element by the to-be-cooperated network element, by receiving a service quality requirement parameter corresponding to the target service of the service request; and a computing node determination module configured to perform determination of a target computing node from the initial computing node based on the service request, the node state parameter, and the service quality requirement parameter by the to-be-cooperated network element, to provide the target service based on the target computing node for the target service.

[0008] In a third aspect, an embodiment of the present disclosure provides a computer readable storage medium having a computer program stored thereon, the computer program being executed by a processor to implement the data processing method as above.

[0009] In a fourth aspect, an embodiment of the present disclosure provides an electronic device, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to perform the data processing method as above by executing the executable instructions.

[0010] In a fifth aspect, an embodiment of the present disclosure provides a computer program product comprising a computer program, the computer program being executed by a processor to implement the data processing method as above.

[0011] The technical solution of the present disclosure has the following beneficial effects:

[0012] The data processing method as above, the communication network performs requirement analysis on a service request based on a session management function (SMF) network element in response to receiving the service request sent by a user equipment for a target service, to obtain a service type required by the target service; the SMF network element determines a to-be-cooperated network element based on the service type; the to-be-cooperated network element receives a service quality requirement parameter corresponding to the target service of the service request, and acquires a node state parameter of an initial computing node matched with the to-be-cooperated network element; and the to-be-cooperated network element determines a target computing node from the initial computing node based on the service request, the node state parameter, and the service quality requirement parameter, to provide the target service based on the target computing node for the target service.

[0013] In one aspect, the method enhances the function of the original SMF network element in the communication network, so that it can perform unified demand analysis after receiving a service request, to identify the multi-element coordination demand involved in communication, calculation, sensing, intelligence and other elements, and select the to-be-coordinated network element matched therewith. This not only compatible and provides access methods for more elements added in future communication networks, but also can reuse the communication specifications of the existing communication network, such as registration, SM policy association establishment, N4 session establishment process, and the like, and the cell format definition also follows the 3GPP format, thereby improving the development efficiency and reducing the implementation and landing difficulty, and further reducing the implementation cost and complexity. On the other hand, the method also involves the coordination interaction process between the SMF network element and the to-be-coordinated network element, thereby further expanding the intelligent dimension based on the algorithm network service scheme, and then backward compatible with the existing algorithm network integration technical scheme. In addition, the method can fully utilize the intelligent resources and computing resources in the communication network, improve the resource utilization rate, and realize multi-element fusion sensing, service demand analysis, intelligent algorithm node coordination selection and integrated intelligent algorithm service.

[0014] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0015] The accompanying drawings, which are incorporated into and form part of the specification, illustrate embodiments consistent with the present disclosure and, together with the specification, serve to explain the principles of the present disclosure. Obviously, the drawings in the following description are only some embodiments of the present disclosure, and other drawings can be obtained by those skilled in the art without creative labor based on these drawings.

[0016] Figure 1 The schematic diagram of the network architecture for realizing the intelligent algorithm fusion in the present exemplary embodiment is shown schematically;

[0017] Figure 2 The flowchart of the data processing method in the present exemplary embodiment is shown schematically;

[0018] Figure 3 The bit diagram of the service request carried in the PDU session in the present exemplary embodiment is shown schematically;

[0019] Figure 4 The bit diagram representing the service type in the present exemplary embodiment is shown schematically;

[0020] Figure 5 The bit diagram representing the AI service type in the present exemplary embodiment is shown schematically;

[0021] Figure 6A flow chart illustrating a method for registering an intelligent computing node in the present exemplary embodiment is shown schematically;

[0022] Figure 7 A bit diagram illustrating a registration message accepted by a communication network in the present exemplary embodiment is shown schematically;

[0023] Figure 8 A bit diagram illustrating a bit representing whether a computing service is supported in a registration message in the present exemplary embodiment is shown schematically;

[0024] Figure 9 A bit diagram illustrating a bit representing whether an AI service is supported in a registration message in the present exemplary embodiment is shown schematically;

[0025] Figure 10 A network architecture diagram illustrating another implementation of an intelligent computing converged network in the present exemplary embodiment is shown schematically;

[0026] Figure 11 A data processing apparatus structure diagram in the present exemplary embodiment is shown schematically;

[0027] Figure 12 A structure diagram of an electronic device in the present exemplary embodiment is shown schematically. DETAILED DESCRIPTION

[0028] Example embodiments now will be described more fully hereinafter with reference to the accompanying drawings; however, the example embodiments can be implemented in many different 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 scope of the example embodiments to those skilled in the art. The features, structures, or characteristics described in connection with the embodiments can be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a thorough understanding of embodiments of the disclosure. One skilled in the relevant art will recognize, however, that the techniques described herein can be practiced without one or more of the specific details, or with other methods, components, materials, and so forth. Other instances of known technologies will not be described in detail so as not to obscure the aspects of the present disclosure.

[0029] In addition, the drawings are only schematic illustrations of the present disclosure and are not necessarily drawn to scale. Identical reference numerals in the drawings represent the same or similar parts, and thus repeated descriptions thereof will be omitted. Some of the block diagrams shown in the drawings are functional entities, which do not necessarily have to correspond to physically or logically independent entities. These functional entities can be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0030] The flowcharts shown in the drawings are only exemplary illustrations and do not necessarily include all steps. For example, some steps can be further decomposed, and some steps can be combined or partially combined, so that the actual execution order can be changed according to the actual situation.

[0031] In order to help those skilled in the art better understand the technical solutions of the present disclosure, the related content involved in the technical solutions of the present disclosure will be introduced below.

[0032] 1) General computing power (referred to as general computing): is the basic computing power of a computer, which can handle a variety of common tasks; its core hardware is a central processing unit (Central Processing Unit, CPU) and an open-source flexible RISC-V architecture processor, and its characteristics are strong versatility, high adaptability, but the computing power density is low.

[0033] 2) Intelligent computing power (referred to as intelligent computing): focuses on artificial intelligence and big data analysis, and the technical architecture emphasizes parallel computing capability and data-driven; its core hardware is: GPU (graphics processing unit), ASIC (application-specific integrated circuit), FPGA (field programmable gate array) and other AI chips to provide computing power through the above AI chip acceleration computing platform, mainly used for artificial intelligence training and inference calculation, which can be applied to computer vision, natural language processing, machine learning and other fields.

[0034] 3) Session Management Function (Session Management Function, SMF): is a core component of a communication network (such as a 5G network), mainly responsible for the management and control of user sessions, including session establishment, modification, release, and maintenance of user plane functions (such as IP address allocation, route optimization, traffic steering, etc. ) and other functions; in the embodiments of the present application, the SMF network element function is enhanced to support the identification and selection of communication, computing, intelligence and other multi-element collaborative requirements, and the selection of network elements required for collaboration, etc. For details, please refer to the following description.

[0035] 4) User Equipment (User Equipment, UE): terminal device, which can initiate multi-element service request and transmit service data.

[0036] 5) Artificial Intelligence Control Function (AICF): can perceive the AI capability and state of the intelligence calculation node, can analyze the AI service demand of the UE device and select the intelligence calculation node that can meet the demand in the AI capability. In the embodiments of the present disclosure, the AICF can be deployed as a functional module together with the SMF and PCF, or can exist independently as a core network element.

[0037] 6) Computing Policy Control Function (CPCF): can perceive the computing capability and state of the intelligence calculation node or computing node, can analyze the computing service demand of the UE and select the intelligence calculation node that can meet the demand in the computing capability. In the embodiments of the present disclosure, the CPCF can be deployed as a functional module together with the SMF and PCF, or can exist independently as a core network element.

[0038] 7) Service Node Proxy (SNP): used to manage intelligence calculation nodes, computing nodes or future data service nodes, etc., can register and update the state of the above nodes to the core network control plane (such as AICF / CPCF), and receive multi-element coordination strategies from the core network control plane (such as AICF / CPCF), and control the service start and stop of the above nodes according to the strategy.

[0039] Based on the above, the specific content of the related terms involved in the technical solutions of the present disclosure is introduced, and then the application profile of the technical solutions of the present disclosure will be introduced.

[0040] In the related technical background, with the continuous emergence of more and more emerging industries and scenes, such as the fusion of artificial intelligence and communication, the fusion of perception and communication, etc., the 6G network must surpass the ability range of traditional communication networks, deeply integrate communication, perception, calculation, intelligence and other multi-element capabilities (hereinafter referred to as multi-element), and develop towards the direction of integrated fusion network of communication service, perception service, calculation service and intelligent service, in order to provide on-demand integrated services.

[0041] Among them, the intelligent calculation fusion service (i.e. communication, calculation, intelligent fusion service) is one of the important features of the 6G network, which makes the 6G network not only need to provide communication service capability, but also can provide calculation, intelligent and other services. However, for artificial intelligence (Artificial Intelligence, AI) services and calculation services, the existing 5G network mainly realizes through the capabilities outside the mobile network (such as through the Internet to transmit data or service OTT / computing power network, etc.).

[0042] However, in the above methods, the mobile network only provides connectivity and cannot perceive upper-layer intelligent and computing services. Therefore, it cannot fully utilize the intelligent and computing resources in the communication network, and consequently, it cannot achieve a multi-element collaborative mechanism. This makes it difficult to realize integrated communication, computing, and intelligent services provided by the communication network. Furthermore, the inventors have discovered that in order for 6G networks to achieve the aforementioned integrated communication, computing, and intelligent services, at least the following key challenges need to be considered and addressed: how to perceive multi-element resources such as communication, computing, and intelligence; how to effectively decompose the needs of multi-element integrated services; and how to achieve control and execution coordination among multiple elements to provide integrated services on demand.

[0043] To address the aforementioned technical problems, this disclosure proposes a data processing method applicable to communication network scenarios requiring integrated services encompassing multiple capabilities such as communication, computing, intelligence, and sensing. This method provides a multi-element fusion sensing and collaborative scheduling mechanism within the communication network. Through the sub-element registration of intelligent computing nodes and the interaction process between SMF / AICF / CPCF network elements, it achieves multi-element fusion sensing, service requirement decomposition, collaborative selection of intelligent computing nodes, and integrated services.

[0044] Next, we will combine Figure 1 The network architecture used in the data processing method provided in the embodiments of this disclosure will be described.

[0045] Figure 1 This schematic diagram illustrates a network architecture for realizing the convergence of computing, communication, and artificial intelligence in this exemplary embodiment; see reference. Figure 1 As shown, the communication network architecture includes: the UE equipment used by the user, the base station (which provides the Radio Access Network (RAN)), and the core communication network architecture. The core communication network (i.e., the communication network shown in the following embodiment, also referred to as the core network) includes: Access and Mobility Management Function (AMF), Policy Control Function (PCF), Network Repository Function (NRF), and User Plane Function (UPF).

[0046] It should be noted that the AMF is mainly responsible for access control, mobility management and session management of terminal users (UE devices), and can establish and maintain data transmission sessions of users in cooperation with the SMF. The PCF is a key network element responsible for unified policy management and control, for example, it can provide mobility management policies (such as handover threshold, access restriction) of the UE by interacting with the AMF, and respond to event-triggered policy updates, control packet flow description through SMF, and implement application traffic detection and forwarding rule management. The NRF network element is the registration and discovery center of network functions (Network Function, NF) in the core network, for example, the NRF can receive and store instance information of all core network functions (such as AMF, SMF, PCF, etc.), including NF type, instance address, service capability, state and associated network slice information, etc.

[0047] At the same time, the communication network also contains a series of core network elements or functional modules in the embodiments of the present disclosure: an enhanced SMF network element, an AI control function AICF network element, a computing power policy control function CPCF network element, and a plurality of service node proxies SNP (corresponding to SNP A, SNP B and SNP C contained in the Figure 1 The communication network architecture also provides intelligent computing nodes (for example, intelligent computing node A, intelligent computing node C) and computing power nodes (for example, computing power node B). Among them, the intelligent computing node is used to provide intelligent computing services and AI service capabilities, and the computing power node is used to provide general computing services such as logical operation and video encoding and decoding, but does not provide intelligent computing services. Moreover, the node information of the intelligent computing node is registered to the AICF network element and the CPCF network element through the SNP, that is, the AI service capability is registered to the AICF network element, and the intelligent computing service is registered to the CPCF network element. The node information of the computing power node is registered to the CPCF network element through the SNP, and the AICF network element cannot perceive the computing power node.

[0048] It should be explained that the number of intelligent computing nodes and computing power nodes provided in the network architecture is only exemplary, and more or less belongs to the scope involved in the embodiments of the present disclosure, and the embodiments of the present disclosure do not make an exhaustive enumeration again.

[0049] In one optional embodiment of this disclosure, in addition to the functions in the existing network structure, the SMF network element's function is enhanced. Specifically, the SMF network element can determine multi-element coordination needs based on the service request sent by the UE device and select the network element that needs to provide the coordination service. Taking the intelligent computing service type as an example, the network element providing the coordination service may include, for example, AICF and CPCF network elements. The AICF network element can decompose the "intelligent + computing" related needs in the service request, perform initial node selection based on the "intelligent" related needs, obtain initial node selection information, and forward the "computing" related needs and initial node selection information to the CPCF network element. The CPCF network element can then select nodes based on the "computing" related needs. The service node proxy SNP (e.g., SNP A, SNP B, and SNP C) registers and updates the node information of the intelligent computing node and the computing power node with the corresponding AICF and CPCF network elements, respectively, and controls the start and stop of the intelligent computing node and computing power node providing related services according to the instructions of the AICF and CPCF network elements.

[0050] For example, in one optional embodiment, in response to receiving a service request sent by a user equipment (i.e., a UE) for a target service, the communication network performs requirement parsing on the service request based on the SMF network elements in the communication network to obtain the service type of the target service; the SMF network elements can then determine the matching network elements to be coordinated based on the service type; furthermore, the network elements to be coordinated receive the service request, the service quality requirement parameters corresponding to the target service, and obtain the node status parameters of the initial computing nodes that match the network elements to be coordinated, and can then determine the target computing node from the initial computing nodes based on the service request, the node status parameters, and the service quality requirement parameters, so as to provide the corresponding service for the target service based on the target computing node.

[0051] The following describes the methods provided by exemplary embodiments of this application in conjunction with the application scenarios described above and with reference to the accompanying drawings. It should be noted that the application scenarios described above are only shown to facilitate understanding of the spirit and principles of this application, and the embodiments of this application are not limited in any way in this respect.

[0052] Figure 2 A flowchart illustrating a data processing method in this exemplary embodiment is shown below. Figure 2 As shown, this data processing method is based on Figure 1 The overall communication network / core network shown is used as the execution entity for explanation. The specific implementation process of this method includes the following steps S201 to S204:

[0053] Step S201: The communication network responds to the service request sent by the user equipment for the target service, and performs requirement parsing on the service request based on the SMF network element to obtain the service type of the target service.

[0054] In step S202, the SMF network element in the communication network determines the to-be-cooperated network element based on the service type.

[0055] In step S203, the to-be-cooperated network element in the communication network receives the service request and the quality-of-service requirement parameter corresponding to the target service.

[0056] In step S204, the to-be-cooperated network element in the communication network determines the target computing node from the initial computing node based on the service request, the quality-of-service requirement parameter, and the node state parameter of the initial computing node matched by the to-be-cooperated network element, to provide the target service for the target service based on the target computing node.

[0057] In Figure 2 In the provided technical solution, on the one hand, the method enhances the function of the original SMF network element in the communication network, so that it can perform unified demand analysis after receiving a service request, to identify the multi-element cooperation demand involved from the communication, computing, sensing, intelligence, and other elements, and select the to-be-cooperated network element matched therewith. This not only compatiblely provides the access method of more elements added in the future communication network, but also can reuse the communication specification of the existing communication network, such as the registration, SM policy association establishment, N4 session establishment process, and the like, and the information element format definition also follows the 3GPP format, thereby improving the development efficiency and reducing the implementation and landing difficulty, and further reducing the implementation cost and complexity. On the other hand, the method also involves the cooperation interaction process between the SMF network element and the to-be-cooperated network element, thereby further expanding the intelligent dimension based on the algorithm network service scheme, and backward compatible with the existing algorithm network fusion technical solution. In addition, the method can fully utilize the intelligent resources and computing resources in the communication network, thereby realizing multi-element fusion sensing, service demand analysis, intelligent algorithm node cooperation selection, and integrated intelligent algorithm service.

[0058] The specific implementation of each step in the embodiment shown in the following will be described in detail: Figure 2 The specific implementation of each step in the embodiment shown in the following will be described in detail:

[0059] In step S201, the communication network responds to the service request sent by the user equipment for the target service, and performs demand analysis on the service request based on the SMF network element to obtain the service type demand of the target service.

[0060] In the embodiment, the target service can be a service that needs to be provided with a single-capability element or a multi-capability element (hereinafter also referred to as a multi-element) in communication, sensing, computing, and AI intelligence by the communication network. In actual application scenarios, it is usually a multi-element integrated service, and therefore, the target service in the embodiment of the present disclosure can be particularly for a multi-element integrated service.

[0061] For example, a user equipment (UE device) is combined with Figure 1 As shown, the UE device can initiate a service request for a multi-element fusion service (i.e., a target service) to the communication network through a wireless access network, so that the communication network selects a target computing node matching the target service from the computing nodes of the communication network through subsequent steps to provide services for the UE device.

[0062] In the related technical solutions, the communication network cannot perceive the multi-element and effectively decompose the demand for multi-element fusion services, so it cannot provide multi-element collaborative services. In order to solve the above technical problems, in the present embodiment, the SMF network element in the existing communication network is enhanced in function, so that after receiving the service request sent by the user equipment, it can analyze the demand of the service request to determine the service type (Service Type) required by the target service, and then decompose to obtain the multi-element collaborative demand involved in the target service.

[0063] In an optional embodiment, the service type includes one of the following services or types of fusion services: artificial intelligence service (AI service), computing service, data service, perception service, and security service.

[0064] For example, after the SMF network element analyzes the demand of the service request, it can be determined whether the service type required by the target service is one of the AI service, computing service, data service, perception service, and security service or a fusion of multiple service types.

[0065] For example, the service request sent by the UE device can include, in addition to the service type (for example, refer to octet x+5 in Figure 3 8 bits in octet x+5, which represents the service type Service type required by the service request, which can be an AI service, a computing service, a data service, a perception service, a security service, or a fusion service), the service request can also include: service identification (for example, refer to Service ID corresponding to octet x+4 in Figure 3 Container ID x in Figure 3 Container ID x, where x usually represents a variable or a specific ID value), the length of the content of Container ID x (corresponding to Length of ContainerID x contents in Figure 3 ).

[0066] For example, when the UE device requests to establish a protocol data unit (PDU) session to realize data transmission to the communication network (e.g., a 6G network), the specific implementation can refer to Figure 3As shown, the UE device carries the service request in octet x+1 to y of the "Protocol Configuration Options" IE of the PDU session establishment request message. It should be noted that the direction is: MS to network direction, i.e. mobile station to network direction; and the container identifier is: Container identifier = FF00H - FFFFH.

[0067] Meanwhile, the following will introduce the service type corresponding to octet x+5 in Figure 4 , in Figure 3 . Figure 4 The service type is schematically illustrated by the bit positions shown.

[0068] Referring to Figure 4 , when Service type = 0000, it represents the service type as an AI service; when Service type = 0001, it represents the service type as a computing service; when Service type = 0010, it represents the service type as a data service; when Service type = 0011, it represents the service type as a perception service; and when Service type = 0100, it represents the service type as a security service. The service type also provides the above-mentioned multi-element fusion service, for example, when Service type = 1000, it represents the service type as a computing and data service fusion service type; and other arbitrary multi-element fusion service types can be customized by the developer according to the actual network demand.

[0069] In addition to the naming of the service type from the service perspective in the above-mentioned embodiment, the naming of the service type can also be based on the capability elements involved in the above-mentioned service demand. For example, since the AI service demand involves communication, artificial intelligence and computing service, it can be determined that the service type corresponding to the target service is a computing and perception collaborative type demand; for another example, the perception service demand involves communication, perception service and computing service, it can be determined that the service type corresponding to the target service is a computing and perception collaborative type demand; for another example, the computing service involves communication and computing service, it can be determined that the service type corresponding to the target service is a computing collaborative type demand, and so on.

[0070] Through this embodiment, when the UE device interacts with the communication network, the service request can be carried in the created PDU session, and the service type contained in the service request can enable the SMF network element to determine the service type required by the target service, thereby improving the logicality of data interaction.

[0071] In an optional embodiment, for the AI service type, when the SMF network element determines that the service type required by the target service is an AI service, it also needs to determine the specific AI service type. Among them, the AI service type includes: AI training task, AI inference task, etc.

[0072] For example, continuing to refer to Figure 3 , the AI service type can be set through the "AI service type" in octet x+6 of the "Protocol Configuration Options" IE of the PDU session establishment request message. For details, please refer to Figure 5 .

[0073] According to Figure 5 , the AI service type (octets x+6, specifically the first and second bit positions), when AI service type = 00, represents that the AI service type is an AI training task; when AI service type = 01, represents that the AI service type is an AI inference task.

[0074] In step S202, the SMF network element in the communication network determines the to-be-coordinated network element based on the service type requirement.

[0075] For example, after the SMF network element determines the service type of the target service, the SMF network element can determine the to-be-coordinated network element according to the service type.

[0076] In an optional embodiment of the present disclosure, when performing step S202, the SMF network element determines the to-be-coordinated network element based on the service type, including: the SMF network element determines that the to-be-coordinated network element includes an artificial intelligence control function (AICF) network element and a computing power policy control function (CPCF) network element in response to the service type containing an artificial intelligence service.

[0077] Among them, the computing node is specifically defined based on the specific service type, for example, the computing node corresponding to the AI service is specialized as a wisdom computing node, and the computing node corresponding to only the computing service is specialized as a computing power node.

[0078] For example, in this embodiment, when the determined service type is an AI service, the determined to-be-coordinated network element includes: an AICF network element and a CPCF network element. At this time, the initial wisdom computing node in the application scenario corresponding to the initial computing node matched by the to-be-coordinated network element is a node for providing artificial intelligence services (also known as AI services) and computing services, such as FPGA (field programmable gate array), ASIC (application specific integrated circuit), tensor processing unit (TPU), etc.

[0079] It needs to be explained that the initial intelligence node can be an intelligence node existing in the communication network and successfully registered in the AICF network element and the CPCF network element.

[0080] The AI service capabilities provided by the above-mentioned intelligence node will be exemplarily described below:

[0081] The AI service types supported by the AI service include: AI inference, AI training, etc.; the AI service state includes: enabled, ended, abnormal, etc.

[0082] Among them, the AI inference capability: model / algorithm name and version, main parameters such as the number of parallel threads, batch size, complexity, floating-point precision, supported inference business types such as text / image / speech / video, inference performance such as throughput, accuracy, inference latency under different parameters or software and hardware configurations, etc.

[0083] For AI training capability: model / algorithm information name and version, main parameters such as complexity, learning rate, number of parallel threads, sampling ratio, feature ratio, early stopping round, supported training business types such as text / image / speech / video, training performance such as regression RMSE / MAE / R2, classification Accuracy / Recall, training latency under different parameters or software and hardware configurations, etc.

[0084] In an optional embodiment, the SMF network element parses that the service type corresponding to the target task only contains computing services, and the SMF network element can determine that the to-be-coordinated network element contains a computing power policy control function CPCF network element.

[0085] When the service type determined in this embodiment is a computing service, the corresponding to-be-coordinated network element contains: a CPCF network element. At this time, the initial computing node matched by the to-be-coordinated network element is specialized as an initial computing power node, and the initial computing power node represents a node for providing general computing services such as logical operation and video encoding and decoding, such as a central processing unit CPU, an image processing unit GPU.

[0086] It needs to be noted that the above-mentioned computing power node does not support AI services, for example, the computing power node only has CPU resources for providing general computing services such as logical operation and video encoding and decoding. It can be understood that when the intelligence node does not support AI services, it degenerates into a computing power node.

[0087] For example, continue with Figure 3The embodiment shown by the bit number determines different service types, assuming that the service type Service type = 0000, that is, when representing an AI service, the SMF network element can determine that the service type corresponding to the target task needs to be coordinated by communication, intelligence, and computing capability elements, and therefore, the determined to-be-coordinated network elements include the AICF network element and the CPCF network element.

[0088] For example, assuming that the service type Service type = 0001, that is, when representing a computing service, the SMF network element determines that communication and computing services are coordinated, and the determined to-be-coordinated network elements include the CPCF network element.

[0089] It should be noted that the above embodiment is an example of a service type that requires the coordination of communication, intelligence, and computing. For other service types, the computing nodes can be determined according to the involved capability elements, for example, the computing nodes corresponding to the perception element can also be perception sensors, and the specific computing nodes can be determined according to the actual service type.

[0090] Further, when determining the to-be-coordinated network elements, in order to enable the to-be-coordinated network elements to determine the available computing nodes and schedule the corresponding matching computing nodes to process the target service, the process of pre-registering the computing nodes in the corresponding to-be-coordinated network elements can be described in detail in combination with the following embodiments.

[0091] Embodiment 1:

[0092] When using the above AICF network element and CPCF network element, the computing nodes existing in the communication network also need to be pre-registered in the corresponding coordinated network elements, such as the AICF network element and the CPCF network element. For example, the computing nodes included in the communication network are intelligent computing nodes, which will be described in detail in combination with Figure 6 .

[0093] Figure 6 A flowchart of an example method for registering intelligent computing nodes in the present example embodiment is shown schematically, please refer to Figure 6 In an optional embodiment, the method for registering intelligent computing nodes in the communication network includes the following steps S601-S604:

[0094] Step S601, sending a first registration request for an initial intelligent computing node to the AICF network element through the SNP, the first registration request containing first node information of the initial intelligent computing node, the first node information at least containing a first service capability parameter of a provided artificial intelligence service.

[0095] Step S602, the AICF network element sends first notification information to the SMF network element, and forwards the first notification information to the access and mobility management function AMF network element through the SMF network element, the AMF network element determines that the communication network provides artificial intelligence services based on the first notification message, and forwards the first notification message to the user equipment through the AMF network element, so that the user equipment determines that the communication network provides artificial intelligence services.

[0096] And, step S603 is performed, a second registration request for the initial intelligence calculation node is sent to the CPCF network element through the SNP, the second registration request contains second node information of the initial intelligence calculation node, and the second node information at least contains second service capability parameters of the provided computing service.

[0097] Step S604, the CPCF network element sends second notification information to the SMF network element, and forwards the second notification information to the AMF network element through the SMF network element, the AMF network element determines that the communication network provides computing services based on the second notification message, and forwards the second notification message to the user equipment through the AMF network element, so that the user equipment determines that the communication network provides computing services.

[0098] For example, for the scenario where the service type of the target service is AI service, the initial computing node corresponding to S203 above is specifically an initial intelligence calculation node. For the intelligence calculation node supporting AI service, since it can provide AI service capability and computing capability, it is necessary to register the node information of the initial intelligence calculation node to the AICF network element and the CPCF network element through the service node proxy SNP. Specifically, the AI service capability is registered in the AICF network element, and the computing capability is registered in the CPCF network element.

[0099] For the specific steps of registering the AI service capability provided by the intelligence calculation node to the AICF network element, the following steps are taken:

[0100] First, the service node proxy SNP sends a first registration request to the AICF network element, and the first registration request includes first node information of each initial intelligence calculation node in the communication network. For example, the first node information includes the node identifier, address information and AI service capability of each initial intelligence calculation node.

[0101] Then, the AICF network element sends first notification information of the AI service capability to the SMF network element, so that the SMF network element knows that the communication network can provide and support AI services.

[0102] Then, in the case of the above embodiment, the SMF network element continues to send the first notification information of the communication network with AI service capability to the AMF network element, so as to continue to make the AMF network element know that the communication network can provide and support AI services. The AMF network element finally sends the first notification information to the UE device registered in the communication network.

[0103] Through the above embodiment, the UE device can know in advance that the communication network can provide and support the AI service, and then the UE device can initiate a task according to the capability elements that the communication network can provide.

[0104] At the same time or after the above embodiment, the specific steps of registering the computing service capability provided by the intelligent calculation node to the CPCF network element are as follows:

[0105] First, the service node proxy SNP also needs to send a second registration request to the CPCF network element, and the second registration request at least contains a second service capability parameter of the provided computing service. The second service capability parameter may be, for example, a computing resource type, a computing resource size, a load condition, a communication delay, and the like. Of course, it can be understood that in addition to the second service capability parameter, the second registration request can also contain the node identifier, address information, and the like of the initial intelligent calculation node.

[0106] Then, the CPCF network element sends second notification information of the computing service capability to the SMF network element, indicating that the communication network can provide and support the computing service to the SMF network element.

[0107] Then, the SMF network element continues to send second notification information of the computing service capability to the AMF network element, indicating that the communication network can provide and support the computing service to the AMF network element. Finally, the AMF network element informs the UE device registered to the communication network of the second notification information.

[0108] Through the embodiment, the UE device can know in advance that the communication network can provide and support the computing service capability, and then the UE device can initiate a task according to the capability elements that the communication network can provide.

[0109] It needs to be explained that in this embodiment, the computing capability is one of the basic capability elements of the communication network, and in the services involving artificial intelligence and perception, the computing service is involved, but not the type of only providing artificial intelligence service and perception service without involving computing service.

[0110] Embodiment 2:

[0111] Taking the computing service type (or computing service type) as an example again, the computing node in the communication network is specifically a computing power node, and the node information of the initial computing power node needs to be registered to the CPCF network element through the SNP.

[0112] For example, the SNP sends a third registration request for the initial computing power node to the CPCF network element, and the third registration request contains third node information of the initial computing power node.

[0113] The third node information includes a third service capability parameter of the provided computing service, and the third service capability parameter includes a node identifier, address information, and a computing service capability (i.e., the third service capability parameter) of the initial computing power node. For example, the third service capability parameter can be a computing resource type, a computing resource size, a load condition, a communication delay, and the like.

[0114] Then, the CPCF network element sends third notification information to the SMF network element, so that the SMF network element obtains that the current communication network supports and provides the computing service.

[0115] The third notification information is further forwarded to the AMF network element via the SMF network element, so that the AMF network element determines that the current communication network supports and provides the computing service based on the third notification message, and continues to send the third notification message of the computing service to the UE device.

[0116] In the above embodiments 1 and 2, in an optional embodiment, in response to the communication network not providing the service corresponding to the service type, the SMF network element sends a rejection response message to the user equipment via the AMF network element.

[0117] The rejection response message indicates the response information of rejecting to provide the service corresponding to the service type.

[0118] Taking the AI service as an example, when the initial intelligent computing node does not register the node information of the initial intelligent computing node into the AICF network element and the CPCF network element respectively, or the registration information is cancelled, the AICF network element needs to notify the SMF network element, indicating that the communication network no longer supports the AI service, and then the SMF network element notifies the AMF network element, indicating to the AMF network element that the communication network no longer supports the AI service, and finally sends a notification to the UE device through the AMF network element, so that the UE device obtains that the communication network no longer supports the AI service.

[0119] Similarly, when there is no registration information of the intelligent computing node or the computing power node in the CPCF network element, the CPCF network element needs to notify the SMF network element, indicating that the network no longer supports the computing service. Further, the SMF network element needs to notify the AMF network element, indicating to the AMF network element that the communication network no longer supports the computing service, and finally sends a notification to the UE device through the AMF network element, so that the UE device obtains that the communication network no longer supports the computing service.

[0120] In this case, when the UE device sends a service request of the AI service, the SMF network element sends a rejection response message to the user equipment via the AMF network element.

[0121] It should be noted that when the UE device is registered to the communication network, refer to Figure 7As shown, the communication network indicates the capability in octet 6 of the "5GS network feature support" IE receiving the message, so that the UE determines whether to support the AI service capability (for example Figure 7 As shown, octet 6 corresponds to the AIS represented by the 6th bit Figure 7 As shown, octet 6 corresponds to the CS represented by the 5th bit

[0122] For example, continuing to combine Figure 8 , referring to the 5th bit in octet 6 can determine whether the communication network provides and supports the computing service. For example, when the value of the 5th bit is 0, it represents that the communication network does not provide and support the computing service; when the value of the 5th bit is 1, it represents that the communication network provides and supports the computing service, that is, the computing node or the intelligent computing node is deployed in the network.

[0123] For another example, in combination with Figure 9 , referring to the 6th bit in octet 6 can determine whether the communication network provides and supports the AI service. For example, when the value of the 6th bit is 0, it represents that the communication network does not provide and support the AI service; when the value of the 6th bit is 1, it represents that the communication network provides and supports the AI service, that is, the intelligent computing node is deployed in the network.

[0124] It needs to be explained that there is no case in the communication network that only provides the AI service capability without providing the computing service capability, so this kind of situation is not described. And for other service types, octet 6 can be used to represent other bits as shown Figure 7 .

[0125] Further, when the initial intelligent computing node and the initial computing node in the communication network change the service provided, the above-mentioned embodiment process needs to be continued to update.

[0126] However, in actual communication scenarios, some events that change frequently are usually involved, for example: AI service state, etc., and some events that change infrequently are also involved, for example: AI service type, AI inference capability, AI training capability, address information, etc. Different service update processes are provided for different events in the embodiments of the present disclosure.

[0127] On the basis of the above-mentioned embodiment 1:

[0128] In an optional embodiment of the present disclosure, in response to a change in the artificial intelligence service and / or the computing service provided by the initial intelligent computing node, a first update request is sent to the AICF network element and / or the CPCF network element through the SNP, so that the AICF network element updates the first service capability parameter and / or the CPCF network element updates the second service capability parameter.

[0129] For example, for infrequent events (for example: AI service type, AI inference capability, AI training capability, address information, etc.), in order to ensure real-time and accuracy, when it is detected that at least one of the artificial intelligence service and the computing service provided by the initial intelligent computing node changes, a first update request is immediately sent to the AICF network element and / or the CPCF network element for updating the service capability.

[0130] Suppose the artificial intelligence service provided by the initial intelligent computing node changes (for example, version upgrade, parameter update, etc.), a first update request is sent to the AICF network element, and the AICF network element performs rapid update according to the first node information of the intelligent computing node contained in the first update request.

[0131] Suppose the computing service provided by the initial intelligent computing node changes (for example, resource load is too high, link interruption, etc.), a first update request is sent to the CPCF network element, and the CPCF network element performs rapid update according to the second node information of the intelligent computing node contained in the first update request.

[0132] It should be explained that the AICF network element and / or the CPCF network element can only update the changed content, or can update all the content, and the embodiments of the present disclosure do not make any special limitation on this.

[0133] Since there are also frequently updated parameters or events in the communication network, for example: AI service state, in order to improve efficiency and save channel transmission resources, periodic update can be performed.

[0134] In an optional embodiment of the present disclosure, in response to the communication duration being greater than the communication duration threshold, a second update request is sent to the AICF network element and / or the CPCF network element through the SNP, so that the AICF network element updates the first service capability parameter and / or the CPCF network element updates the second service capability parameter.

[0135] Among them, the second update request is an update request triggered by periodic update.

[0136] Exemplarily, the second update request can be periodically sent to the AICF network element and / or the CPCF network element through the SNP by detecting whether the communication duration is greater than a communication duration threshold, so as to enable the AICF network element and / or the CPCF network element to periodically update the frequently-updated service parameter. Meanwhile, the AI service state and other infrequently-updated information can also be updated together with the message.

[0137] As can be easily understood by those skilled in the art, a combination of the above two update methods can also be used. For example, when a frequently-updated parameter or event is detected, the AICF network element and / or the CPCF network element can be triggered to periodically update the frequently-updated service parameter upon detection of the update. Since the AI service state and other infrequently-updated information can also be updated together with the message, the system detects that the update has been performed, and skips the current period of update.

[0138] Based on the above embodiment 2:

[0139] In an optional embodiment, in response to a change in the computing service provided by the initial computing power node (such as excessive resource load, link interruption, etc.), a third update request is sent to the CPCF network element through the SNP to enable the CPCF network element to update the third service capability parameter. Alternatively, in response to the communication duration being greater than the communication duration threshold, a third update request is sent to the CPCF network element through the SNP to enable the CPCF network element to update the third service capability parameter.

[0140] It can be understood that the communication duration threshold can be pre-configured by the developer, or can be determined in other ways, and the embodiments of the present disclosure do not make any special limitation thereto.

[0141] In step S203, the to-be-coordinated network element in the communication network receives the service request and the service quality requirement parameter corresponding to the target service.

[0142] In the case where the communication network supports the service type, the to-be-coordinated network element can obtain the service quality requirement parameter of the target service and receive the service request from the SMF network element.

[0143] Exemplarily, the service quality requirement parameter of the target service can be transmitted to the to-be-coordinated network element synchronously with the service request after being configured and determined locally in the SMF network element, or can be transmitted to the to-be-coordinated network element in advance before the service request is sent, and the embodiments of the present disclosure do not make any limitation thereto.

[0144] In addition, the quality of service requirement parameter of the target service can be retrieved from a unified data management (UDM) or obtained by interacting with a policy control function (PCF) network element in the communication network.

[0145] For example, the quality of service requirement parameter can be represented by quality of service (QoS) information. The QoS information can include parameters such as latency, bandwidth, and packet loss rate.

[0146] For example, the AI service type can be an AI inference service or an AI training service, and the AI service requirement can be a latency requirement or an accuracy requirement.

[0147] For example, the latency requirement in the AI service requirement can be represented by the 3rd, 4th, and 5th bits of octet x+6, and the accuracy requirement in the AI service requirement can be represented by the 6th, 7th, and 8th bits of octet x+6. Figure 3 For example, if the values of the 3rd, 4th, and 5th bits of octet x+6 are 000, it represents that the latency requirement of AI training is greater than or equal to 15 minutes or the latency requirement of AI inference is greater than or equal to 4 seconds. If the values of the 6th, 7th, and 8th bits of octet x+6 are 000, it represents that the accuracy requirement of the AI service is within 45%.

[0148] It should be noted that the values in this embodiment are only exemplary, and the specific values can be standardized according to the service requirements.

[0149] For example, the SMF network element can forward the service request and the QoS information to the to-be-coordinated network element when it is determined that the communication network supports the AI service. For example, the SMF can forward the service request and the QoS information to the AI CF network element through Naicf_AIPolicyControl_Create Request, and the SMF can forward the service request and the QoS information to the CPCF network element through Ncpcf_ComputingPolicyControl_Create Request.

[0150] The following will be described taking the AI service corresponding to the intelligent computing fusion service type as an example.

[0151] In some example embodiments of the present disclosure, the AICF network element, in response to the service type containing an artificial intelligence service, also needs to determine an artificial intelligence service type of the artificial intelligence service.

[0152] The artificial intelligence service type is, for example, an AI training task, an AI inference task, and the like mentioned in the above embodiments. Continuing to refer to Figure 3 The artificial intelligence service type (AI service type) can be represented by the 1st and 2nd bits of octet x+6.

[0153] Since the capabilities and QoS parameters of the artificial intelligence nodes required by different AI service types are different, in order to further improve the accuracy, the AI services can be distinguished according to the AI service types.

[0154] In step S204, the target computing node is determined from the initial computing nodes by the to-be-cooperated network element based on the service request, the service quality requirement parameter, and the node state parameter of the initial computing node matched with the to-be-cooperated network element, so as to provide the service for the target service based on the target computing node.

[0155] For example, in addition to receiving the service request and the service quality requirement parameter corresponding to the target service, the to-be-cooperated network element also acquires the node state parameter of the initial computing node matched with the to-be-cooperated network element, so that in this step, the to-be-cooperated network element determines the target computing node from the initial computing nodes based on the service request, the service quality requirement parameter, and the node state parameter of the initial computing node matched with the to-be-cooperated network element.

[0156] The initial computing node can be, for example, all the artificial intelligence nodes, computing power nodes, and the like registered in the to-be-cooperated network element in the above embodiments.

[0157] It needs to be explained that the to-be-cooperated network element can acquire the node state parameter of the initial computing node matched therewith at any time. For example, the node state parameter can be acquired after receiving the service request and the service quality requirement parameter corresponding to the target service, or the node state parameter of the initial computing node can be acquired after the initial computing node is registered in the to-be-cooperated network element, and the present disclosure does not make any specific limitation thereto.

[0158] For example, the to-be-cooperated network element can determine the target computing node from the initial computing nodes according to the received service request and service quality requirement parameter, and in combination with the acquired node state parameter of the initial computing node matched therewith, so as to provide the service for the target service based on the target computing node.

[0159] The following also takes the AICF network element and the CPCF network element as examples for the to-be-cooperated network element determined by the AI service.

[0160] In an optional embodiment of the present disclosure, the AICF network element receives a service request and a quality of service requirement parameter via the SMF network element; if the service type in the service request contains an artificial intelligence service, an artificial intelligence service type and an artificial intelligence service requirement of the artificial intelligence service are determined; the AICF network element determines an intermediate intelligent calculation node from an initial intelligent calculation node based on a node state parameter of the initial intelligent calculation node, the quality of service requirement parameter, the artificial intelligence service requirement, and the artificial intelligence service type; and the CPCF network element determines a target intelligent calculation node from the intermediate intelligent calculation node based on a delay requirement of a target service contained in the service request received via the AICF network element, the quality of service requirement parameter, node information of the intermediate intelligent calculation node, and a node state parameter of the intermediate intelligent calculation node.

[0161] In the node information of the intermediate intelligent calculation node, at least one of artificial intelligence service delay information and artificial intelligence service accuracy information provided by the intermediate intelligent calculation node is contained; and the artificial intelligence service requirement contains at least one of a delay requirement and an accuracy requirement corresponding to the target service.

[0162] For example, after determining the AI service type, the AICF network element can determine the intermediate intelligent calculation node from the initial intelligent calculation node based on the service request, the quality of service requirement parameter, and a node state parameter of the initial calculation node to be matched with the network element.

[0163] In the AI service type, the required initial calculation node is specifically the initial intelligent calculation node, that is, the initial intelligent calculation node represents all intelligent calculation nodes registered in the AICF network element, and the intelligent calculation node is a calculation node providing artificial intelligence service and computing service capability.

[0164] For example, the AICF network element can perform preliminary screening of the intelligent calculation node according to the AI service type and the quality of service requirement parameter (QoS information) carried in the service request received from the SMF, and in combination with the node state parameter of the initial intelligent calculation node obtained from the SNP, to obtain the intermediate intelligent calculation node screened initially.

[0165] Specifically, the AICF network element can match the AI service type in the service request with the AI service type of the initial intelligent computing node, and the AI service state needs to be in an enabled or ended state. At the same time, the AICF network element matches the delay requirement and accuracy requirement, QoS information corresponding to the target service in the service request with the AI inference / training capability that the initial intelligent computing node can provide (for example, the sum of the uplink and downlink packet delay budget in the QoS information and the artificial intelligence service delay information provided by the initial intelligent computing node must be less than or equal to the delay requirement in the service request), and generates the AI service delay information of each initial intelligent computing node according to the "inference / training delay under different parameters or software and hardware configurations" of the matched initial intelligent computing node, and includes it in the intelligent computing node preliminary selection information (i.e., the node information of the intermediate intelligent computing node).

[0166] After determining the intermediate intelligent computing node, the AICF network element can send the service request, QoS information and node information of the intermediate intelligent computing node to the CPCF network element through Ncpcf_ComputingPolicyControl_Create Request.

[0167] Finally, the following steps are performed: the CPCF network element determines the target intelligent computing node based on the service request, service quality requirement parameters, node information of the intermediate intelligent computing node, delay requirement parameters of the target service, and node state parameters of the intermediate intelligent computing node received via the AICF network element.

[0168] The node information of the intermediate intelligent computing node includes at least one of the artificial intelligence service delay information and the artificial intelligence service accuracy information provided by the intermediate intelligent computing node.

[0169] It needs to be explained that the AICF network element obtains the delay requirement parameters of the target service via the SMF network element and forwards them to the CPCF network element; and the CPCF network element can obtain the node information of the intermediate intelligent computing node from the AICF network element; and the node state parameters of the intermediate intelligent computing node are obtained by the CPCF network element from the SNP.

[0170] In addition to including AI service delay information and / or accuracy information, the node information of the intermediate intelligent computing node can also be, for example, the node identifier, address information, model / algorithm information of the intermediate intelligent computing node obtained through preliminary selection by the AICF network element, and the embodiments of the present disclosure do not make any special limitation on this.

[0171] Exemplarily, the CPCF network element can obtain the latency requirement parameter in the service request received from the AICF network element, the AI service latency information in the node information of the intermediate computing node, and the QoS information, and combine and compare the node state parameter of the intermediate computing node obtained from the SNP, so as to select the target computing node from the initially selected intermediate computing node.

[0172] Specifically, the CPCF network element matches the latency requirement in the service request, the AI service latency information in the node information of the intermediate computing node, and the QoS information with the computing capability of the intermediate computing node (for example, the sum of the uplink and downlink data packet delay budget in the AI service latency and QoS information and the communication latency in the computing capability of the intermediate computing node is less than or equal to the latency requirement in the service request, and meanwhile, the computing resource type, size and load condition of the intermediate computing node meet the corresponding software and hardware configuration requirements in the AI service latency information). Meanwhile, the computing service latency information of the intermediate computing node is generated according to the matched communication latency and AI service latency information of the intermediate computing node, and is included in the finally obtained node information of the target computing node (for example, including the node identification, address and computing service latency information of the target computing node), and is sent to the AICF network element through Ncpcf_ComputingPolicyControl_Create Response.

[0173] In an optional embodiment of the present disclosure, when the target computing node determined by the CPCF network element from the intermediate computing node is multiple, the target computing node can be determined based on the following embodiments:

[0174] In response to the CPCF network element determining multiple first target computing nodes from the intermediate computing node, the CPCF network element is controlled to determine the target computing node from the multiple first target computing nodes according to a preset selection strategy.

[0175] The preset selection strategy includes at least one of the following:

[0176] The computing node with the lowest current load among the multiple first target computing nodes is determined as the target computing node.

[0177] The computing node with the lowest communication latency among the multiple first target computing nodes is determined as the target computing node.

[0178] The computing node with the most computing resource types among the multiple first target computing nodes is determined as the target computing node.

[0179] For example, when the CPCF network element matches multiple intermediate intelligent computing nodes that meet the requirements, the CPCF network element can select a target intelligent computing node according to a preset selection policy preconfigured locally. For example, the CPCF network element can preferentially select an intelligent computing node with low load, preferentially select an intelligent computing node with low communication delay, preferentially select an intelligent computing node with multiple types of computing resources, and the like.

[0180] It should be noted that the above preset selection policy is only exemplary and not limited, and the embodiments of the present disclosure do not make any exhaustive enumeration.

[0181] Further, after determining the target intelligent computing node / target computing node, the AICF network element sends the service identifier, the QoS information, and the node information of the target intelligent computing node to the SMF network element through Naicf_AIPolicyControl_Create Response.

[0182] The SMF network element can select a user plane function (UPF) and allocate an Internet Protocol (IP) of a UE device according to the node information of the target intelligent computing node. For example, the SMF network element can interact with a network repository function (NRF) or be configured locally (for example, an association relationship between the UPF and the intelligent computing node is registered to the NRF by the UPF or is configured to the SMF by an operation administration and maintenance (OAM)). In addition, the SMF network element notifies a policy control function (PCF) of the IP of the UE device and the node information of the target intelligent computing node, so that the PCF updates a session management policy (SM policy) and a QoS policy, for example, adjusts an uplink and downlink packet delay budget according to an intelligent computing service delay.

[0183] The SM policy is a rule set formulated by the PCF in the core network and sent to the SMF network element, and is used to dynamically control behaviors such as QoS, charging policy, and traffic routing of a user session.

[0184] Continue to perform the following steps: the SMF network element sends the service identifier, QoS information, node information of the target AI node, and IP of the UE device to the AICF network element through the Naicf_AIPolicyControl_Update Request. The AICF network element sends the service identifier, QoS information, node information of the target AI node, and UE IP to the CPCF network element through the Ncpcf_ComputingPolicyControl_Update Request.

[0185] The CPCF network element requests the SNP corresponding to the selected target AI node to establish AI resources through the Nsnp_AIandComputing_ServiceResource_Setup Request, and the request includes the service identifier, QoS information, node information of the target AI node, IP of the UE device, computing resource type, computing resource size, AI service type, AI service demand (such as delay demand, accuracy demand, etc.), model / algorithm information, and the like. Subsequently, the target AI node will perform AI resource scheduling and data processing on the service data stream corresponding to the service identifier and / or UE IP according to the information in the Nsnp_AIandComputing_ServiceResource_Setup Request.

[0186] After the SNP successfully enables the corresponding AI service and successfully schedules the corresponding computing resources for the target AI node, the SNP sends a service resource establishment response to the CPCF network element through the Nsnp_AIandComputing_ServiceResource_Setup Response. The CPCF network element notifies the AICF network element that the communication network has successfully established AI service resources through the Ncpcf_ComputingPolicyControl_Update Response, including the service identifier, QoS information, node information of the target AI node, and IP of the UE device.

[0187] The AICF network element notifies the SMF network element that the intelligent computing service resource of the communication network has been successfully established through the Naicf_AIPolicyControl_Update Response, including the service identifier, QoS information, node information of the target intelligent computing node, and IP of the UE device. After the SMF network element receives the response, the SMF network element will perform N4 session establishment / modification with the corresponding user plane function (UPF). This includes the service identifier, QoS information, node information of the target intelligent computing node, and IP of the UE device. The UPF will perform packet detection, rate matching, etc. according to the QoS information and the communication delay in the intelligent computing node selection information.

[0188] Finally, the SMF network element sends a service response to the UE device, including the service identifier, QoS information, target intelligent computing node IP, and UE IP. For example, when the network notifies the UE device of the successful establishment of a PDU session, the service request is carried in the "Protocol Configuration Options" IE of the PDU session establishment acceptance message. (Network to MS direction: Container identifier = FF00H-FFFFH).

[0189] On the basis of the above embodiments, an architecture diagram of a data processing system is also provided, which can realize control separation and execution separation. For details, refer to Figure 10 The following embodiments can be implemented, as shown in the figure:

[0190] For the registration of computing nodes, the AI service node (i.e., intelligent computing node) can be registered to the AICF network element, and the computing power node can be registered to the CPCF network element. The specific registration steps can refer to the above embodiments, and will not be described here.

[0191] For on-demand service orchestration, the SMF network element identifies multi-element requirements and forwards the service request to the AICF network element first. The AICF network element arranges the service logic, i.e., the sequence of AI services and computing services.

[0192] In terms of node selection and service channel establishment, the service request first reaches the AICF network element. The AICF network element selects an AI service node that meets the conditions according to the identified AI service requirements while arranging the service, and then sends the decomposed computing power requirements and the information (such as IP address) of the selected AI service node to the CPCF network element. The CPCF network element selects a computing power node according to the computing power requirements and the computing power node capabilities, and returns the information (such as IP address) of the selected computing power node to the AICF network element.

[0193] In the aspect of service coordination execution, for example, the AI service node provides a model for the computing node, and then the computing node performs inference calculation. The AI service node and the computing node need to be able to identify each other's IP addresses.

[0194] Further, one embodiment of the present disclosure provides a data processing apparatus.

[0195] Figure 11 An illustrative architecture diagram of the data processing apparatus is schematically shown. Referring to Figure 11 As shown, the data processing apparatus 1100 includes a demand analysis module 1101, a network element determination module 1102, a parameter acquisition module 1103, and a computing node determination module 1104.

[0196] The demand analysis module 1101 is configured to perform demand analysis on a service request based on a session management function (SMF) network element in response to receiving the service request sent by a user equipment for a target service, to obtain a service type required by the target service; the network element determination module 1102 is configured to perform determination of a to-be-coordinated network element based on the service type by the SMF network element; the parameter acquisition module 1103 is configured to perform reception of a service quality demand parameter corresponding to the target service by the to-be-coordinated network element based on the service request; and the computing node determination module 1104 is configured to perform determination of a target computing node from an initial computing node based on the service request, the service quality demand parameter, and a node state parameter of the initial computing node matched by the to-be-coordinated network element, to provide the target service for the target service based on the target computing node.

[0197] In an optional embodiment, the service type includes one or more of the following converged services:

[0198] Artificial intelligence service;

[0199] Computing service;

[0200] Data service;

[0201] Perception service;

[0202] Security service.

[0203] In an optional embodiment, the computing nodes at least include an intelligent computing node and a computing node, the network element determining module 1102 is configured to perform the following: in response to the service type including an artificial intelligence service, the SMF network element determines that the to-be-coordinated network element includes an artificial intelligence control function (AICF) network element and a computing power policy control function (CPCF) network element, and the initial computing node matched by the to-be-coordinated network element is an initial intelligent computing node, which represents a node for providing an artificial intelligence service and a computing service; in response to the service type including a computing service, the SMF network element determines that the to-be-coordinated network element includes a computing power policy control function (CPCF) network element, and the initial computing node matched by the to-be-coordinated network element is an initial computing node, which represents a node for providing a computing service.

[0204] In an optional embodiment, the computing node determining module 1104 is configured to perform the following: in response to the service type including an artificial intelligence service, the artificial intelligence service type of the artificial intelligence service is determined; the AICF network element receives a service request and a quality of service requirement parameter via the SMF network element; if the service type in the service request includes an artificial intelligence service, the artificial intelligence service type of the artificial intelligence service and the artificial intelligence service requirement are determined; the artificial intelligence service requirement includes at least one of a delay requirement and an accuracy requirement corresponding to the target service; the AICF network element determines an intermediate intelligent computing node from the initial intelligent computing node based on the node state parameter of the initial intelligent computing node, the quality of service requirement parameter, the artificial intelligence service requirement, and the artificial intelligence service type; the CPCF network element determines a target intelligent computing node from the intermediate intelligent computing node based on the delay requirement of the target service included in the service request received via the AICF network element, the quality of service requirement parameter, the node information of the intermediate intelligent computing node, and the node state parameter of the intermediate intelligent computing node; the node information of the intermediate intelligent computing node includes at least one of artificial intelligence service delay information and artificial intelligence service accuracy information provided by the intermediate intelligent computing node.

[0205] In an optional embodiment, the computing node determining module 1104 is configured to perform the following: if the CPCF network element determines a plurality of first target intelligent computing nodes from the intermediate intelligent computing node, the CPCF network element determines the target intelligent computing node from the plurality of first target intelligent computing nodes according to a preset selection strategy.

[0206] The preset selection strategy includes at least one of the following:

[0207] The intelligent computing node with the lowest current load among the plurality of first target intelligent computing nodes is determined as the target intelligent computing node.

[0208] The intelligent computing node with the lowest communication delay among the plurality of first target intelligent computing nodes is determined as the target intelligent computing node.

[0209] The computing resource type of the target intelligent computing node is determined from the plurality of first target intelligent computing nodes.

[0210] In an optional embodiment, the communication network comprises the intelligent computing node, and the apparatus further comprises a sending module configured to send, through a service node proxy (SNP), a first registration request for the initial intelligent computing node to an AICF network element, wherein the first registration request comprises first node information of the initial intelligent computing node, and the first node information at least comprises first service capability parameters of the provided artificial intelligence service; the sending module is configured to send, by the AICF network element, first notification information to a session management function (SMF) network element, and forward the first notification information to an access and mobility management function (AMF) network element via the SMF network element, so that the AMF network element determines that the communication network provides the artificial intelligence service based on the first notification information, and the first notification information is forwarded to a user equipment (UE) through the AMF network element, so that the UE determines that the communication network provides the artificial intelligence service; and the sending module is configured to send, through the SNP, a second registration request for the initial intelligent computing node to a common policy control function (CPCF) network element, wherein the second registration request comprises second node information of the initial intelligent computing node, and the second node information at least comprises second service capability parameters of the provided computing service; the sending module is configured to send, by the CPCF network element, second notification information to the SMF network element, and forward the second notification information to the AMF network element via the SMF network element, so that the AMF network element determines that the communication network provides the computing service based on the second notification information, and the second notification information is forwarded to the UE through the AMF network element, so that the UE determines that the communication network provides the computing service.

[0211] In an optional embodiment, the apparatus further comprises an updating module configured to send, through the SNP, a first update request to the AICF network element and / or the CPCF network element in response to a change in the artificial intelligence service and / or the computing service provided by the initial intelligent computing node, so that the AICF network element updates the first service capability parameters and / or the CPCF network element updates the second service capability parameters; and the updating module is configured to send, through the SNP, a second update request to the AICF network element and / or the CPCF network element in response to the communication duration being greater than a communication duration threshold, so that the AICF network element updates the first service capability parameters and / or the CPCF network element updates the second service capability parameters.

[0212] The data processing apparatus 1100 provided by the embodiments of the present disclosure can execute the technical solutions of the data processing method in any of the above embodiments, and the implementation principles and beneficial effects thereof are similar to those of the data processing method. For details, refer to the implementation principles and beneficial effects of the data processing method, which will not be described here in detail.

[0213] In the exemplary embodiments of the present disclosure, a computer readable storage medium is also provided, on which a program product capable of implementing the above-mentioned method of the present specification is stored. In some possible implementations, various aspects of the present application can also be implemented in the form of a program product, which includes program codes for causing a terminal device to perform the steps described in the above-mentioned "Exemplary Methods" section of the present specification according to various exemplary embodiments of the present application when the program product is run on the terminal device.

[0214] The program product for implementing the above-mentioned method according to the embodiments of the present application can take a portable compact disc read-only memory (CD-ROM) and include program codes, and can be run on a terminal device, such as a personal computer. However, the program product of the present application is not limited thereto, and in the present document, the readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus or device.

[0215] The program product can take any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium may, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or any combination thereof. More specific examples (non-exhaustive list) of readable storage media include an electrical connection having one or more wires, a portable disc, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0216] The computer readable signal medium can include a data signal propagated in a baseband or as a carrier wave in a propagated data signal, in which readable program codes are borne. Such a propagated data signal can take various forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination thereof. The readable signal medium can also be any readable medium other than the readable storage medium, which can send, propagate or transmit the program for use by or in conjunction with an instruction execution system, apparatus or device.

[0217] The program codes contained on the readable medium can be transmitted by any suitable medium, including but not limited to wireless, wired, optical cable, radio frequency (RF), etc., or any suitable combination thereof.

[0218] The program code, which can also be called an "application program", can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C++, or the like, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computing device, partly on the user's computing device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device or entirely on the remote computing device or server. In the latter scenario, the remote computing device can be connected to the user's computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computing device, such as through the Internet using an Internet Service Provider (ISP). The application program can be embodied in a computer readable medium, which can be any device or medium that can store the program for use by or in connection with the computing device.

[0219] In an exemplary embodiment of the present disclosure, an electronic device capable of implementing the above-described method is also provided.

[0220] Those skilled in the art can understand that each aspect of the present application can be implemented as a system, a method or a program product. Therefore, each aspect of the present application can be embodied in a form of entirely hardware, entirely software (including firmware, microcode, etc.), or a combination of hardware and software, which can be generally referred to as "circuitry", "module" or "system".

[0221] The electronic device 1200 according to this embodiment of the present application will be described below with reference to Figure 12 Figure 12 The electronic device 1200 shown is merely an example and should not be taken as limiting the functionality or use of embodiments of the present application.

[0222] As shown in Figure 12 The components of electronic device 1200 can include, but are not limited to, at least one processing unit 1210, at least one storage unit 1220, a bus 1230 connecting the different system components, including the storage unit 1220 and the processing unit 1210, and a display unit 1240.

[0223] The storage unit stores program code which can be executed by the processing unit 1210, so that the processing unit 1210 performs the steps described above in the "Exemplary Methods" section of the present specification according to various exemplary embodiments of the present application. For example, the processing unit 1210 can perform the steps described above in the "Exemplary Methods" section of the present specification according to various exemplary embodiments of the present application. Figure 3 ​Steps S201 to S204 shown in FIG. 2.

[0224] The storage unit 1220 can include a readable medium in the form of volatile storage such as a random access memory (RAM) 12201 and / or cache memory 12202, and also possibly non-volatile storage such as a read-only memory (ROM) 12203.

[0225] The storage unit 1220 can also include a program / utility 12204 having a set (at least one) of program modules 12205 such as an operating system, one or more application programs, other program modules, and program data, each of which

[0226] The bus 1230 can represent one or more of several types of bus structures, including a storage bus or bus controller, a peripheral bus, a graphics acceleration bus, a processor or local bus using any of a variety of bus architectures.

[0227] The electronic device 1200 can also communicate with one or more external devices 2000 such as a keyboard or pointing device, a Bluetooth device, etc.; one or more devices that enable a user to interact with the electronic device 1200; and / or one or more devices (e.g., a router, a modem, a server, etc.) that enable the electronic device 1200 to communicate with one or more other computing devices. Such communication can occur via an input / output (I / O) interface 1250. Still yet, the electronic device 1200 can communicate with one or more networks, such as one or more local area networks (LANs), one or more wide area networks (WANs), and / or one or more public networks, such as the Internet, via the network adapter 1260. As depicted, the network adapter 1260 can communicate with the other components of the electronic device 1200 via the bus 1230. It should be appreciated that although the network adapter 1260 is depicted as a single component, the network adapter 1260 can include any number of components, such as a plurality of network adapters, a plurality of network interface cards, etc.

[0228] Through the above description of the embodiments, those skilled in the art can easily understand that the example embodiments described herein can be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solutions according to the embodiments of the present disclosure can be embodied in the form of a software product. The software product can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash disk, a mobile hard disk, or the like) or a network, and includes a number of instructions to make a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) execute the methods according to the embodiments of the present disclosure.

[0229] In addition, the above-described figures are only schematic illustrations of the processes included in the method according to the example embodiments of the present application, and are not intended for limiting purposes. It is easy to understand that the processes shown in the above-described figures do not indicate or limit the time sequence of these processes. In addition, it is also easy to understand that these processes can be executed synchronously or asynchronously, for example, in multiple modules.

[0230] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, such a division is not mandatory. Indeed, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided into embodied by multiple modules or units.

[0231] Other embodiments of the present disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure following, in general, the principles of the present disclosure and including such departures from the present disclosure that come within known or customary practice in the art to which the present disclosure pertains. The specification and examples are to be regarded as exemplary only, and the true scope and spirit of the present disclosure are indicated by the appended claims.

[0232] It should be understood that the present disclosure is not limited to the precise structures described above and illustrated in the drawings and that various modifications and changes can be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A data processing method, characterized by, The method comprises: In response to receiving a service request sent by a user equipment for a target service, performing demand analysis on the service request based on a session management function (SMF) network element to obtain a service type required by the target service; The SMF network element determines a to-be-coordinated network element based on the service type; The to-be-coordinated network element receives the service request and a service quality requirement parameter corresponding to the target service; The to-be-coordinated network element determines a target computing node from an initial computing node based on the service request, the service quality requirement parameter, and a node state parameter of the initial computing node matched by the to-be-coordinated network element, to provide service for the target service based on the target computing node.

2. The method of claim 1, wherein, The service type includes one of the following services or a plurality of fusion services: An artificial intelligence service; A computing service; A data service; A perception service; A security service.

3. The method according to claim 1 or 2, characterized in that, The computing node at least includes an intelligent computing node and a computing power node, and the SMF network element determines the to-be-coordinated network element based on the service type, including: In response to the service type including an artificial intelligence service, the SMF network element determines that the to-be-coordinated network element includes an artificial intelligence control function (AICF) network element and a computing power policy control function (CPCF) network element, and the initial computing node matched by the to-be-coordinated network element is an initial intelligent computing node, which represents a node for providing an artificial intelligence service and a computing service in a communication network; In response to the service type including a computing service, the SMF network element determines that the to-be-coordinated network element includes a computing power policy control function (CPCF) network element, and the initial computing node matched by the to-be-coordinated network element is an initial computing power node, which represents a node for providing a computing service in a communication network.

4. The method of claim 3, wherein, The to-be-coordinated network element determines a target computing node from the initial computing node based on the service request, the node state parameter, and the service quality requirement parameter, including: The AICF network element receives the service request and the service quality requirement parameter via the SMF network element; If the service type in the service request includes the artificial intelligence service, the artificial intelligence service type and the artificial intelligence service requirement of the artificial intelligence service are determined; wherein the artificial intelligence service requirement includes at least one of a latency requirement and an accuracy requirement corresponding to the target service; The AICF network element determines an intermediate intelligent computing node from the initial intelligent computing node based on the node state parameter of the initial intelligent computing node, the service quality requirement parameter, the artificial intelligence service requirement, and the artificial intelligence service type; The CPCF network element determines a target intelligent computing node from the intermediate intelligent computing node based on the latency requirement of the target service included in the service request received via the AICF network element, the service quality requirement parameter, node information of the intermediate intelligent computing node, and a node state parameter of the intermediate intelligent computing node; wherein the node information of the intermediate intelligent computing node includes at least one of artificial intelligence service latency information and artificial intelligence service accuracy information provided by the intermediate intelligent computing node.

5. The method of claim 4, wherein, Determining a target intelligent computing node from the intermediate intelligent computing node includes: If the CPCF network element determines multiple first target intelligent computing nodes from the intermediate intelligent computing node, the CPCF network element determines the target intelligent computing node from the multiple first target intelligent computing nodes based on a preset selection strategy; The preset selection strategy includes at least one of the following: Determine the intelligent computing node with the lowest current load among the multiple first target intelligent computing nodes as the target intelligent computing node; Determine the intelligent computing node with the lowest communication delay among the multiple first target intelligent computing nodes as the target intelligent computing node; Determine the intelligent computing node with the most types of computing resources among the multiple first target intelligent computing nodes as the target intelligent computing node.

6. The method of claim 3, wherein, The communication network includes an intelligent computing node, and the method further includes: sending, by a service node proxy (SNP), a first registration request for the initial intelligent computing node to the AICF network element, wherein the first registration request includes first node information of the initial intelligent computing node, and the first node information at least includes first service capability parameters of a provided artificial intelligence service; sending, by the AICF network element, first notification information to the SMF network element, and forwarding the first notification information to an access and mobility management function (AMF) network element via the SMF network element, wherein the AMF network element determines that the communication network provides the artificial intelligence service based on the first notification information, and forwards the first notification information to the user equipment via the AMF network element, so that the user equipment determines that the communication network provides the artificial intelligence service; and sending, by the SNP, a second registration request for the initial intelligent computing node to the CPCF network element, wherein the second registration request includes second node information of the initial intelligent computing node, and the second node information at least includes second service capability parameters of a provided computing service; sending, by the CPCF network element, second notification information to the SMF network element, and forwarding the second notification information to the AMF network element via the SMF network element, wherein the AMF network element determines that the communication network provides the computing service based on the second notification information, and forwards the second notification information to the user equipment via the AMF network element, so that the user equipment determines that the communication network provides the computing service.

7. The method of claim 6, wherein, The method further includes: in response to a change in the artificial intelligence service and / or the computing service provided by the initial intelligent computing node, sending, by the SNP, a first update request to the AICF network element and / or the CPCF network element, so that the AICF network element updates the first service capability parameters, and / or the CPCF network element updates the second service capability parameters; in response to a communication duration being greater than a communication duration threshold, sending, by the SNP, a second update request to the AICF network element and / or the CPCF network element, so that the AICF network element updates the first service capability parameters, and / or the CPCF network element updates the second service capability parameters.

8. A data processing apparatus, characterized by, includes: The demand analysis module is configured to perform demand analysis on the service request based on a session management function (SMF) network element, to obtain a service type required by the target service in response to receiving a service request sent by a user equipment for a target service; The network element determination module is configured to perform determination of a to-be-coordinated network element by the SMF network element based on the service type; The parameter acquisition module is configured to perform reception of the service request and service quality demand parameters corresponding to the target service by the to-be-coordinated network element; The computing node determination module is configured to perform determination of a target computing node from an initial computing node matched by the to-be-coordinated network element based on the service request, the service quality demand parameters, and node state parameters of the initial computing node, to provide service for the target service based on the target computing node.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by a processor to implement the data processing method of any one of claims 1 to 7.

10. An electronic device, comprising: Comprise: A processor; And A memory for storing executable instructions of the processor; Wherein the processor is configured to execute the data processing method of any one of claims 1 to 7 by executing the executable instructions.

Images