Artificial intelligence service management and orchestration methods, architectures and related products

Abstract

The invention provides an artificial intelligence service management and arrangement method, architecture and related products, and relates to the technical field of communication. The method comprises the following steps: an artificial intelligence resource management function module receives a configuration updating request from a service management and open function module, wherein the configuration updating request comprises artificial intelligence service summary information subscribed by a wireless access network automation application; and the artificial intelligence resource management function module generates artificial intelligence resource configuration and artificial intelligence network function arrangement configuration according to the configuration updating request. And the artificial intelligence resource management function module issues the artificial intelligence resource configuration and the artificial intelligence network function arrangement configuration to the cloud to trigger the cloud to provide resource supply and network function arrangement for the subscribed artificial intelligence service according to the artificial intelligence resource configuration and the artificial intelligence network function arrangement configuration. According to the embodiment of the invention, full-automatic and refined closed-loop management and configuration of artificial intelligence computing tasks from demand to resource supply can be realized.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the field of communication technology, and in particular, to an artificial intelligence service management and orchestration method and architecture, an electronic device, a computer readable storage medium, and a computer program product. BACKGROUND

[0002] This section is intended to provide background information to facilitate a better understanding of embodiments of the present disclosure. Information in this section is not admitted to be prior art.

[0003] 6G (6th Generation Mobile Communication Technology) new applications (e.g., embodied robots, immersive communications, multi-agent communications, etc.) have higher requirements for network latency, bandwidth, and computing. Computing-intensive tasks (e.g., AI (Artificial Intelligence) / ML (Machine Learning) related model training or inference tasks) need to offload part of the computing tasks to the network edge (e.g., base station or edge intelligence center) for execution to reduce the resource load on the terminal side.

[0004] Therefore, a service management and orchestration (SMO) scheme for a radio access network (RAN) covering device-edge-cloud end-to-end collaboration needs to be designed to achieve unified management and intelligent orchestration of wireless AI. SUMMARY

[0005] The purpose of the present disclosure is to provide an artificial intelligence service management and orchestration method, device, electronic device, computer readable storage medium, and computer program product, which can manage and configure artificial intelligence services for a radio access network.

[0006] Other characteristics and advantages of the present disclosure will become apparent from the following detailed description, or will be learned by practice of the present disclosure.

[0007] This disclosure provides a method for managing and orchestrating artificial intelligence (AI) services, comprising: an AI resource management module receiving a configuration update request from a service management and open function module, the configuration update request containing aggregated AI service information subscribed to by the radio access network automation application; the AI ​​resource management module generating AI resource configuration and AI network function orchestration configuration based on the configuration update request; and the AI ​​resource management module distributing the AI ​​resource configuration and the AI ​​network function orchestration configuration to the cloud, triggering the cloud to provide resource provisioning and network function orchestration for the subscribed AI services based on the AI ​​resource configuration and the AI ​​network function orchestration configuration.

[0008] In some embodiments, the service management and orchestration architecture further includes an O2 interface. The artificial intelligence resource management function module includes an artificial intelligence resource management and open function unit. The method further includes: the artificial intelligence resource management and open function unit collecting operational status data of AI-related computing and storage resources from the cloud via the O2 interface, wherein the operational status data includes at least one of performance management data, configuration management data, and fault management data.

[0009] In some embodiments, the AI ​​resource management function module generates AI resource configuration based on the configuration update request, including: the AI ​​resource management and open function unit generates the AI ​​resource configuration for the subscribed AI service based on the running status data.

[0010] In some embodiments, the service management and orchestration architecture further includes an O2 interface. The artificial intelligence resource management function module includes an artificial intelligence network function orchestration unit. The method further includes: the artificial intelligence network function orchestration unit collecting at least one of performance management data, configuration management data, and fault management data related to artificial intelligence network functions from the cloud through the O2 interface.

[0011] In some embodiments, the artificial intelligence resource management function module generates an artificial intelligence network function orchestration configuration based on the configuration update request, including: the artificial intelligence network function orchestration unit generates the artificial intelligence network function orchestration configuration based on at least one of the performance management data, configuration management data, and fault management data related to the artificial intelligence network function.

[0012] In some embodiments, the method further includes: the service management and orchestration architecture selecting at least one node at the edge of the radio access network to carry the subscribed artificial intelligence service. The service management and orchestration architecture configures cloud resources and network functions for the at least one node, enabling the at least one execution node to execute the subscribed artificial intelligence service based on service data issued by the radio access network automation application, using the cloud resources and the network functions.

[0013] In some embodiments, the service management and orchestration architecture further includes an AI operations management module and an O1 interface. The method further includes: the AI ​​operations management module collecting operational status information generated by the subscribed AI service during its operation at the edge via the O1 interface, the operational status information including at least one of performance management data, configuration management data, and fault management data.

[0014] In some embodiments, the method further includes: the artificial intelligence operation management module updating the configuration of the subscribed artificial intelligence service on the edge side based on the operation status information generated during the operation of the subscribed artificial intelligence service on the edge side.

[0015] This disclosure provides an artificial intelligence service management and orchestration method, implemented by a service management and orchestration architecture for a radio access network (RAN). The service management and orchestration architecture includes an artificial intelligence operations management module. The method includes: the service management and orchestration architecture selecting at least one node on the edge side of the RAN to carry subscribed artificial intelligence services; and the artificial intelligence operations management module collecting operational status information generated by the subscribed artificial intelligence services during their operation on the edge side from the at least one node. This operational status information includes at least one of performance management data, configuration management data, and fault management data, and is used to update the configuration of the subscribed artificial intelligence services on the edge side.

[0016] In some embodiments, the method further includes: the artificial intelligence operation management module updating the configuration of the subscribed artificial intelligence service on the edge side based on the operation status information generated during the operation of the subscribed artificial intelligence service on the edge side.

[0017] This disclosure provides a service management and orchestration architecture, including an artificial intelligence resource management module, a service management and open function module, and a radio access network automation application. The artificial intelligence resource management module receives configuration update requests from the service management and open function module, the configuration update requests containing aggregated information about artificial intelligence services subscribed to by the radio access network automation application. Based on the configuration update requests, it generates artificial intelligence resource configurations and artificial intelligence network function orchestration configurations. These configurations are then distributed to the cloud to trigger the cloud to provide resource provisioning and network function orchestration for the subscribed artificial intelligence services based on these configurations.

[0018] This disclosure provides a service management and orchestration architecture, including: an artificial intelligence operation management function module, used to collect operation status information generated during the operation of subscribed artificial intelligence services on the edge side from at least one node, the operation status information including at least one of performance management data, configuration management data and fault management data, the operation status information being used to update the configuration of the subscribed artificial intelligence services on the edge side.

[0019] This disclosure provides an electronic device comprising a memory and a processor. The memory stores computer program instructions. The processor invokes the computer program instructions stored in the memory to implement the artificial intelligence service management and orchestration method described above.

[0020] This disclosure provides a computer-readable storage medium storing computer program instructions to implement the artificial intelligence service management and orchestration method as described in any of the preceding embodiments.

[0021] This disclosure provides a computer program product or computer program that includes computer program instructions stored in a computer-readable storage medium. The computer program instructions are read from the computer-readable storage medium, and the processor executes the computer program instructions to implement the aforementioned artificial intelligence service management and orchestration method.

[0022] The artificial intelligence service management and orchestration method, apparatus, electronic device, computer-readable storage medium, and computer program product provided in this disclosure achieve seamless and automated integration from service subscription to cloud execution by establishing a resource and function linkage orchestration mechanism oriented towards artificial intelligence service needs. Specifically, it can automatically convert AI service subscription requests initiated by the RAN side into cloud-executable computing resource scheduling schemes and network function deployment strategies. This fills the gap in the original SMO architecture for wireless access networks for refined and automated management of AI / ML models and their required resources, solving the problems of fragmented AI task scheduling and underlying resource supply and coarse configuration in the prior art. It provides key system-level support for achieving end-to-end AI service collaboration across devices, edge, and cloud.

[0023] It should be understood that the above general description and the following detailed description are merely exemplary and do not limit this disclosure. Attached Figure Description

[0024] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure. It is obvious that the drawings described below are merely some embodiments of this disclosure, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.

[0025] Figure 1 An exemplary system architecture diagram is shown that can be applied to the artificial intelligence service management and orchestration method in the embodiments of this disclosure.

[0026] Figure 2 This is a schematic diagram illustrating a service management and orchestration architecture based on relevant technologies.

[0027] Figure 3 This is a flowchart illustrating an artificial intelligence service management and orchestration method according to an exemplary embodiment.

[0028] Figure 4 This is a schematic diagram illustrating a service management and orchestration architecture according to an exemplary embodiment.

[0029] Figure 5 This is a flowchart illustrating an artificial intelligence resource configuration generation method according to an exemplary embodiment.

[0030] Figure 6 This is a flowchart illustrating an artificial intelligence network function orchestration configuration generation method according to an exemplary embodiment.

[0031] Figure 7This is a flowchart illustrating an artificial intelligence service execution method according to an exemplary embodiment.

[0032] Figure 8 This is a flowchart illustrating an edge-side configuration update method according to an exemplary embodiment.

[0033] Figure 9 This is a flowchart illustrating an artificial intelligence service management and orchestration method according to an exemplary embodiment.

[0034] Figure 10 This is a flowchart illustrating an artificial intelligence service management and orchestration method according to an exemplary embodiment.

[0035] Figure 11 This is an interactive schematic diagram illustrating a wireless AI end-to-end collaborative SMO management method based on smart glasses, according to an exemplary embodiment.

[0036] Figure 12 A schematic diagram of the structure of an electronic device suitable for implementing embodiments of the present disclosure is shown. Detailed Implementation

[0037] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of the exemplary embodiments to those skilled in the art. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted.

[0038] Those skilled in the art will recognize that embodiments of this disclosure can be a system, apparatus, device, method, or computer program product. Therefore, this disclosure can be implemented in the following forms: entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software.

[0039] The features, structures, or characteristics described in this disclosure can be combined in any suitable manner in one or more embodiments. Numerous specific details are provided in the following description to give a thorough understanding of embodiments of this disclosure. However, those skilled in the art will recognize that the technical solutions of this disclosure can be practiced with one or more specific details omitted, or other methods, components, apparatuses, steps, etc., can be employed. In other instances, well-known methods, apparatuses, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of this disclosure.

[0040] In this disclosure, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0041] The accompanying drawings are merely illustrative of this disclosure, and the same reference numerals in the drawings denote the same or similar parts, thus omitting repeated descriptions of them. Some block diagrams shown in the drawings do not necessarily correspond to physically or logically independent entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0042] The flowchart shown in the accompanying drawings is merely illustrative and does not necessarily include all content and steps, nor does it require execution in the described order. For example, some steps may be broken down, while others may be combined or partially combined; therefore, the actual execution order may change depending on the specific circumstances.

[0043] In the description of this disclosure, unless otherwise stated, " / " means "or," for example, A / B can mean A or B. "And / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone. Furthermore, "at least one" means one or more, and "multiple" means two or more. The terms "first," "second," etc., do not limit the quantity or order of execution, and "first," "second," etc., do not necessarily imply differences. The terms "contains," "includes," and "has" are used to indicate an open-ended meaning of inclusion and refer to the existence of additional elements / components / etc. besides those listed.

[0044] To better understand the above-mentioned objectives, features and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of the present disclosure can be combined with each other.

[0045] The following section will first explain some of the terms used in the embodiments of this disclosure so that those skilled in the art can understand them.

[0046] AI: Artificial Intelligence.

[0047] 6G: 6th Generation Mobile Communication Technology.

[0048] RAN: Radio Access Network.

[0049] O-RAN: Open RAN, Open Radio Access Network.

[0050] SMO: Service Management and Orchestration.

[0051] CU: Centralized Unit.

[0052] DU: Distributed Unit.

[0053] NWDAF: Network Data Analytics Function.

[0054] QoS: Quality of Service.

[0055] UPF: User Plane Function.

[0056] 5QI: 5G QoS Indicator.

[0057] GPU: Graphics Processing Unit.

[0058] CM: Configuration Management.

[0059] DME: Data Management and Exposure.

[0060] FOCOM: Federated O-Cloud Orchestration and Management.

[0061] IMS: Infrastructure Management Services.

[0062] Near-RT RIC: Near-Real-Time RAN Intelligent Controller.

[0063] NFO: Network Function Orchestration.

[0064] Non-RT RIC: Non-Real-Time RAN Intelligent Controller.

[0065] OAM: Operation, Administration and Maintenance.

[0066] PM: Performance Management.

[0067] SME: Service Management and Exposure.

[0068] The preceding text introduced some terms and concepts involved in the embodiments of this disclosure. The following text introduces the technical features involved in the embodiments of this disclosure.

[0069] Figure 1 An exemplary system architecture diagram is shown that can be applied to the artificial intelligence service management and orchestration method in the embodiments of this disclosure.

[0070] like Figure 1 As shown, the system architecture includes a terminal 101, a server 102, a network 103, a radio access network (RAN) 104, and a carrier network 105. Among these, Figure 1 Any of the terminals 101 shown can be used as a remote terminal or a relay terminal. Communication based on proximity service can be established between the remote terminal and the relay terminal. The remote terminal can directly access the network or access the network through the relay terminal.

[0071] Wireless access network 104 is used to establish a connection between terminal 101 and operator network 105 via wireless communication, thereby enabling the terminal to communicate with server 102.

[0072] The wireless access network can be either a traditional wireless access network or an Open Radio Access Network (O-RAN).

[0073] The Radio Access Network (RAN) is a core component of mobile communication systems. As an "air bridge" between end-user equipment (such as mobile phones and IoT terminals) and the operator's core network, it is responsible for all key functions related to radio access. Its core tasks include establishing and managing the physical and logical connections between user equipment and the network through radio electromagnetic signals. Specifically, this includes: transmitting and receiving radio signals, dynamically allocating and managing radio resources (such as spectrum, time slots, and power), channel coding and modulation / demodulation, data encryption and security protection, mobility handover management (ensuring continuous communication for users while on the move), and quality of service control. From a physical implementation perspective, traditional RAN mainly consists of base stations (such as eNodeB in 4G and gNB in ​​5G) and their transmission networks. However, in 5G and future networks, its architecture has evolved into a more open and intelligent form (such as O-RAN). By decoupling base station functions into centralized units (CU), distributed units (DU), and radio frequency units (RU / AAU), and introducing open interfaces and software-defined control, it achieves a comprehensive upgrade in network flexibility, efficiency, and multi-vendor ecosystem. In short, the wireless access network is the physical carrier and logical control entity for the "last mile" wireless access function of all mobile terminals, and is the primary wireless gateway for users to connect to the digital world.

[0074] SMO (Service Management and Orchestration) is the top-level intelligent management and coordination platform in the architecture of radio access networks (such as Open Radio Access Networks, O-RAN). As the "brain" and command center of the entire wireless network, it performs unified automated deployment, configuration, monitoring and lifecycle management of decoupled and diversified RAN network elements (such as CU, DU, RU) through standardized interfaces. Relying on its embedded data analysis and AI capabilities, it realizes intelligent perception of the global network status, policy generation and closed-loop optimization, thereby driving the evolution of wireless networks from traditional static operation and maintenance to an intelligent paradigm of automation, self-optimization and self-healing.

[0075] In the O-RAN architecture, SMO can refer to the Service Management and Orchestration Platform, which is responsible for unified automated management and intelligent coordination of the open and decoupled radio access network. It is a core component that drives network intelligence and operation and maintenance transformation.

[0076] Figure 2 This is a schematic diagram illustrating a service management and orchestration architecture based on relevant technologies.

[0077] Below, this embodiment will first provide the following... Figure 2 A comparison of Chinese and English in the text.

[0078] SMO framework: SMO framework; Other SMO framework functions: Other SMO framework functions; TE&IV related functions: Training and resource integration verification related functions; O2-related functions: O2-related functions; RAN OAM-related functions: RAN OAM-related functions; External terminations: External terminal interfaces; AL / ML workflow functions: AI / ML workflow functions; Data management and exposure functions: Data management and exposure functions; Non-RT RIC framework: Non-real-time RIC (RANIntelligent Controller) framework; Non-RT RIC: Non-real-time RIC; Service management and exposure functions: Service management and exposure functions; rApp management functions: rApp management functions; R1 termination: R1 terminal interface; Other Non-RT RIC framework functions: Other non-real-time RIC framework functions; External AI / ML services: External AI / ML services; O2termination: O2 terminal interface; Open FH M-Plane termination: Open fronthaul M-plane terminal interface; O1termination: O1 Terminal interface; O2 Cloud; M-plane; External external AI / MLoversight; Function anchored inside the Non-RT RIC framework; Functions anchored outside the Non-RT RIC framework; Non-anchored functions; O-RAN-defined interface; External interface; Production and consumption of R1 services.Non-RT RIC framework: A framework for non-real-time RIC applications; AI-related functions: AI-related functions; A1 termination: A1 terminal interface.

[0079] In the O-RAN (Open Radio Access Network) architecture defined by the O-RAN Alliance (Open Radio Access Network Alliance), the SMO framework is as follows: Figure 2 As shown, the SMO architecture primarily handles AI / ML services through the NRT RIC (Non-RT RIC: Non-Real-Time Intelligent Controller). First, the rAPP (RAN Application) discovers the corresponding AI / ML services through Service Management and Open Functions (SME). Second, the rAPP subscribes to the corresponding AI / ML services through the AI / ML Workflow Function (AIMLWF), such as performing AI / ML model training. The rAPP requests training of the AI / ML model by specifying training requirements (e.g., required data, model, validation criteria, etc.) and collects data through Data Management and Open Functions (DME). AI / ML model training can be executed by the SMO / NRT RIC framework or other rAPPs, and the results are returned to the rAPP after training is complete.

[0080] exist Figure 2 In the architecture shown, the SMO, as the service management and orchestration center in the O-RAN architecture, configures and controls the nRT RIC (Near-RT RIC) through the O1 interface, according to the current O-RAN defined SMO architecture. It collects log information related to FCAPS (Fault Management, Configuration Management, Accounting Management, Performance Management, and Security Management) from the E2 node (E2 Termination) through the O1 interface. The SMO also collects FCAPS information related to O-Cloud cloud infrastructure management and services from the O2 node (O2 Termination) interface.

[0081] Existing technical solutions have the following problems in supporting end-to-end AI collaboration.

[0082] The current SMO definition given by O-RAN only provides a schematic diagram of the RAN's OAM (Operation, Administration and Maintenance) related functions. Figure 2 The RAN OAM-related functions are defined, but the specific functions they contain are not defined.

[0083] The general definitions of the RAN OAM-related functions are as follows.

[0084] The logical functions that produce the RAN OAM-related services whichare exposed to rApps via the R1 interface are called the “RAN OAM-relatedfunctions”. They realize the following functionalities: Producing RAN OAM-related services which are exposed to rApps via the R1 interface. Performing CM conflict mitigation. Interfacing with Near-RT RICs and E2 nodes through O1 termination. -RAN OAM-related functions receive fault notifications and obtainalarm list from Near-RT RICs and E2 nodes. -RAN OAM-related functions provision configuration changes to Near-RTRICs and E2 nodes. -RAN OAM-related functions collect performance data from Near-RT RICs and E2 nodes. -RAN OAM-related functions collect trace data from E2 nodes. The following is the translation.

[0085] These logical functions that provide Radio Access Network (RAN) operation, management, and maintenance (OAM) related services to rApp through the R1 interface are called "RAN OAM related functions". They implement the following functions.

[0086] RAN OAM-related services are provided to rApp through the R1 interface.

[0087] Perform configuration management (CM) conflict mitigation.

[0088] The O1 terminal interacts with Near-RT RICs and E2 nodes.

[0089] RAN OAM-related functions receive fault notifications from Near-RT RICs and E2 nodes and obtain alarm lists.

[0090] RAN OAM-related functions distribute configuration changes to Near-RT RICs and E2 nodes.

[0091] RAN OAM-related functions collect performance data from Near-RT RICs and E2 nodes.

[0092] RAN OAM-related functions collect tracking data from the E2 node.

[0093] The OAM-related functions defined by O-RAN include: enabling capability exposure interaction with rAPP through the R1 interface; obtaining fault notifications and alarm list information from nRT RIC and E2 nodes through the O1 interface; providing configuration changes to nRT RIC and E2 nodes; collecting performance data from nRT RIC and E2 nodes; and collecting log data from E2 nodes.

[0094] The current SMO definition given by O-RAN only provides a schematic diagram of the RAN's O2 function. Figure 2 (O2-related functions), but the specific functions they contain are not defined.

[0095] The general definition of O2-related functions is as follows.

[0096] The logical functions that produce the O2-related services which are exposed to rApps via the R1 interface are called the “O2-related functions”. O2-related functions include FOCOM and NFO. They realize the following functionalities: Producing O2-related services which are exposed to rApps via the R1interface. Interfacing with O-Cloud through O2 termination. -O2-related functions receive O-Cloud infrastructure FCAPS (PM, CM, FM) data via O-Cloud IMS. -O2-related functions receive O-Cloud deployment FCAPS (PM, CM, FM)data via O-Cloud DMS. -O2-related functions provision configuration changes to O-Cloud. Performing conflict mitigation for O-Cloud configuration changes. The following is the translation.

[0097] These logical functions that provide O2-related services to rApp through the R1 interface are called "O2-related functions". O2-related functions include Federated O-Cloud Orchestration and Management (FOCOM) and Network Functions Orchestration (NFO). They implement the following functions: O2-related services are provided to rApp through the R1 interface.

[0098] Interact with O-Cloud via O2 terminal.

[0099] O2-related functions receive FCAPS (Performance Management PM, Configuration Management CM, Fault Management FM) data from the O-Cloud infrastructure through the O-Cloud Infrastructure Management Service (IMS).

[0100] O2-related functions receive FCAPS (Performance Management PM, Configuration Management CM, Fault Management FM) data deployed by O-Cloud through O-Cloud Deployment Management Service (DMS).

[0101] O2-related functions will send configuration changes to O-Cloud.

[0102] Mitigation of conflicts arising from configuration changes to O-Cloud.

[0103] Federated O-Cloud Orchestration and Management (FOCOM): The FOCOM isresponsible for accounting and asset management of the resources in the cloud. The FOCOM is the primary consumer of services provided by the IMS. The FOCOM has information about the O-Cloud resources management. Specifically, the FOCOM needs to know whether the services are within the operator domain or external.

[0104] The following is the translation.

[0105] Federal O-Cloud Orchestration and Management (FOCOM): FOCOM is responsible for the metering and asset management of resources in the cloud. It is the primary consumer of services provided by IMS. FOCOM possesses information related to O-Cloud resource management, particularly regarding whether the relevant services are located within or outside the carrier's domain.

[0106] Network Function Orchestration (NFO): The NFO is responsible fororchestrating the assembly of the network functions as a composition of NFDeployments in the O-Cloud. The NFO is the primary consumer of the DMS.

[0107] The following is the translation.

[0108] The O2-related functions defined by O-RAN include: interacting with O-Cloud through the O2 interface; obtaining performance, configuration, and fault management data of the O-Cloud infrastructure through O-Cloud IMS; obtaining performance, configuration, and fault management data of the O-Cloud deployment through O-Cloud DMS; and providing configuration changes and conflict mitigation to O-Cloud. FOCOM is responsible for cloud resource accounting and asset management, while NFO is responsible for orchestrating the network functions deployed in O-Cloud.

[0109] Network Function Orchestration (NFO): NFO is responsible for orchestrating network functions into a series of network function (NF) deployments in O-Cloud. It is the primary consumer of Deployment Management Service (DMS).

[0110] As can be seen from the above definitions, the current definitions of O1 and O2 are mainly defined around AI / ML services. O1's FCAPS (F - Fault Management) manages from the perspective of AI / ML operation monitoring, while O2 manages the needs of cloud resources (storage and computing) in the entire RAN operation. Neither of them defines related functions from the perspective of the management and control of AI / ML algorithms or models themselves.

[0111] Conclusion: The existing O1 and O2 interfaces only provide basic, passive, and rudimentary support for AI / ML operations (monitoring machine status and providing general resources). They do not proactively, meticulously, or intelligently manage the core of AI / ML—namely, algorithms and models.

[0112] To address the aforementioned issues, this application defines the relevant functions and processes for AI / ML algorithm and model management and configuration in the O1 and O2 interfaces, thereby implementing a wireless AI end-to-end collaborative SMO management solution for RAN.

[0113] For the evolution towards 6G, AI-Native Network solutions are a key technology that needs to be tackled. AI-Native Networks require end-to-end management of AI resources and AI network functions to achieve cross-domain AI collaboration between devices, edge, and cloud.

[0114] AI-native networks refer to a new type of network architecture that deeply, comprehensively, and organically integrates artificial intelligence into the entire process of designing, building, operating, and optimizing communication networks.

[0115] To address the aforementioned needs, based on the existing functionalities of O-RAN SMO and NRT RIC, this application proposes an SMO solution for end-to-end orchestration of AI resources and AI functions, and management of AI operation, so as to realize the management and configuration of AI / ML related algorithms, models and required resources by SMO.

[0116] Figure 3 This is a flowchart illustrating an artificial intelligence service management and orchestration method according to an exemplary embodiment.

[0117] In some embodiments, the AI ​​service management and orchestration method proposed in this application can be implemented by a Service Management and Orchestration Architecture (SMO) for Radio Access Networks, wherein the Service Management and Orchestration Architecture may include an AI resource management function module, a service management and open function module, and a Radio Access Network Automation Application (rAPP).

[0118] Those skilled in the art will understand that, based on the technical concepts disclosed in this application, the names of the modules can be adjusted in actual implementation, and a single functional module can be split into multiple modules or multiple functional modules can be merged into one module. Any structural changes that adopt the same or equivalent technical concepts as this invention and implement the technical solutions described in this application shall fall within the protection scope of this application.

[0119] In some embodiments, the above-mentioned artificial intelligence resource management function module may also be named an end-to-end artificial intelligence resource platform (Resource Platform for E2E AI).

[0120] Among them, the Resource Platform for E2E AI is used to: enhance the O2-related functions in the SMO architecture, add management and orchestration functions for computing and storage resources for AI algorithm or model deployment, computation, inference and storage, add orchestration functions for AI network functions running on O-Cloud, collect management data on the allocation and running status of AI-related computing and storage resources in O-Cloud through the O2 interface, and update and configure resources as needed.

[0121] Reference Figure 3 The artificial intelligence service management and orchestration method provided in this disclosure may include the following steps.

[0122] In step S302, the artificial intelligence resource management function module receives a configuration update request from the service management and open function module. The configuration update request contains summary information of artificial intelligence services subscribed by the wireless access network automation application.

[0123] In this context, a subscribed artificial intelligence service can refer to a specific artificial intelligence function for which a radio access network application (rAPP) has formally submitted a usage request to the service management and orchestration (SMO) system based on its business needs, and the SMO system has accepted the request and is preparing to allocate resources and execution capabilities.

[0124] In some embodiments, Service Management and Open Functionality (SME) can discover subscribed AI services in the Service Management and Orchestration Architecture, such as receiving AI services subscribed to by rAPP.

[0125] In some embodiments, when Service Management and Open Functions (SME) discovers an AI service, it can send a configuration update request to the AI ​​Resource Management Function Module according to certain rules, so that the AI ​​Resource Management Module can update the resource configuration and network function orchestration configuration for the subscribed AI service.

[0126] In step S304, the artificial intelligence resource management function module generates artificial intelligence resource configuration and artificial intelligence network function orchestration configuration based on the configuration update request.

[0127] A Network Function (NF) is a modular software entity with independent functionality and flexible scheduling. It can operate autonomously, be upgraded independently, and interoperate through standard interfaces. Different network functions are independent of each other, and each NF can be broken down into several network function services.

[0128] The AI ​​resource management module can dynamically generate and issue two types of precise configurations by parsing the detailed requirements in the service subscription requests of subscribed AI services and combining them with real-time cloud resource inventory and preset scheduling strategies: first, computing and storage resource allocation instructions for specific AI tasks (such as allocating a specific number of GPUs and memory to a model); and second, workflow orchestration instructions to guide the deployment of corresponding AI network function instances in the cloud (such as starting a specified version of model training or inference service, or orchestrating existing functional components into product components required by the task through function orchestration). This enables an intelligent and refined transformation from "business requirements" to "executable resource and function solutions".

[0129] In step S306, the AI ​​resource management module sends the AI ​​resource configuration and AI network function orchestration configuration to the cloud, so as to trigger the cloud to provide resource supply and network function orchestration for the subscribed AI services based on the AI ​​resource configuration and AI network function orchestration configuration.

[0130] In some embodiments, the AI ​​resource management module can package and send the generated "resource allocation plan (AI resource configuration)" (such as allocating 2 GPUs and 100GB of memory) and "function deployment instructions (AI network function orchestration configuration)" (such as starting the inference service of pipeline recognition model V2.1) to the cloud (O-Cloud). Upon receiving the package, the cloud immediately executes the plan: first, it allocates computing and storage resources according to the amount, and then starts and configures specific AI service instances on these resources according to the instructions, thereby turning a "design blueprint" into a "runnable and usable AI service environment" to provide physical support for subsequent model training or inference tasks.

[0131] In some embodiments, the service management and orchestration architecture may also include an O2 interface. The artificial intelligence resource management functional module may include artificial intelligence resource management and open functional units.

[0132] The O2 interface is the core management channel in the open wireless access network architecture that connects the service management and orchestration system with the cloud infrastructure.

[0133] In some embodiments, the Artificial Intelligence (AI) Resource Management and Open Function Unit collects operational status data of AI-related computing and storage resources from the cloud via the O2 interface. The operational status data includes at least one of performance management data, configuration management data, and fault management data.

[0134] The AI ​​resource management and open functional unit proactively collects real-time operational status data of cloud-based AI computing power and storage resources through the O2 interface, enabling global perception and dynamic monitoring of heterogeneous AI resources. This mechanism allows the resource scheduling system to make intelligent decisions based on real-time load, accurately match task requirements according to resource allocation, and avoid risky nodes by leveraging fault information. This improves the utilization rate of dedicated AI resources, ensures the reliability of resource supply, and supports efficient collaboration and elastic scaling of end-to-end AI tasks.

[0135] In some embodiments, the AI ​​resource management function module (e.g., Resource Platform for E2EAI) may include AI resource management and exposure functions for AI (RME4AI).

[0136] In some embodiments, the AI ​​resource management module, as a component of the FOCOM (Federated O-Cloud Orchestration and Management) function, provides management and access to AI-related computing and storage resources on the O-Cloud infrastructure. It collects performance management data (PM), configuration management data (CM), and fault management data (FM) related to computing and storage resources for AI algorithm or model deployment / computation, inference, and storage reported from the O2 interface. It provides functions such as computing / storage resource awareness, computing / storage resource registration, computing / storage resource access, and computing / storage resource configuration. It also generates configuration updates for AI resources and sends them to the O-Cloud.

[0137] Figure 5 This is a flowchart illustrating an artificial intelligence resource configuration generation method according to an exemplary embodiment.

[0138] In some embodiments, the service management and orchestration architecture also includes an O2 interface.

[0139] In some embodiments, the AI ​​resource management function module may include AI resource management and open function units.

[0140] refer to Figure 5 The above-mentioned method for generating artificial intelligence resource allocation may include the following steps.

[0141] In step S502, the AI ​​resource management and open function unit collects operational status data of AI-related computing and storage resources from the cloud via the O2 interface. The operational status data includes at least one of performance management data, configuration management data, and fault management data.

[0142] In step S504, the AI ​​resource management and open function unit generates AI resource configurations for subscribed AI services based on the operational status data.

[0143] In some embodiments, the service management and orchestration architecture also includes an O2 interface.

[0144] In some embodiments, the AI ​​resource management function module may also include an AI network function orchestration unit (FOAIN).

[0145] In some embodiments, the AI ​​network function orchestration unit can collect at least one of performance management data, configuration management data, and fault management data related to AI network functions from the cloud via the O2 interface.

[0146] The AI ​​network function orchestration unit collects real-time operational data of cloud-based AI network functions through the O2 interface, enabling fine-grained monitoring and closed-loop control of the entire lifecycle of AI service instances (such as model training / inference tasks). This mechanism allows the orchestration system to dynamically optimize workflows based on performance metrics, ensure version consistency according to configuration status, and quickly locate and recover from anomalies using fault data, thereby guaranteeing the service quality, operational reliability, and cross-domain collaboration efficiency of AI network functions.

[0147] FOAIN can include O-Cloud's allocation, reclamation, and orchestration services for computing and storage resources deployed for AI-related network functions. As a component of the NFO function, FOAIN manages relevant computing or storage resources through the O2 interface and collects O-Cloud AI network function-related performance management data (PM), configuration management data (CM), and fault management data (FM) from the O2 interface; it supports the deployment of O-Cloud AI model training, management, and deployment (including AI model registration, discovery, subscription / change of subscription, storage, and optional training capability registration / deregistration / query, etc.), performance monitoring, inference, and other functions; and generates configuration updates for AI network functions and distributes them to O-Cloud.

[0148] Figure 6 This is a flowchart illustrating an artificial intelligence network function orchestration configuration generation method according to an exemplary embodiment.

[0149] refer to Figure 6 The above-mentioned method for generating the orchestration and configuration of artificial intelligence network functions may include the following steps.

[0150] In step S602, the AI ​​network function orchestration unit collects at least one of the following from the cloud via the O2 interface: performance management data, configuration management data, and fault management data related to AI network functions.

[0151] In step S604, the AI ​​network function orchestration unit generates an AI network function orchestration configuration based on at least one of the performance management data, configuration management data, and fault management data related to the AI ​​network function.

[0152] Figure 7 This is a flowchart illustrating an artificial intelligence service execution method according to an exemplary embodiment.

[0153] refer to Figure 7 The above-mentioned artificial intelligence service execution method may include the following steps.

[0154] In step S702, the service management and orchestration architecture selects at least one node on the edge side of the wireless access network to carry the subscribed artificial intelligence service.

[0155] The edge of the wireless access network is an intermediate functional layer in the mobile network deployed at geographical and network locations close to user terminals. It mainly includes entities such as edge cloud servers, near real-time intelligent controllers, and base stations with computing capabilities. Its core role is to significantly reduce data transmission latency and alleviate the pressure on the core network by offloading the computing tasks of artificial intelligence services to the network edge, and to achieve efficient, real-time response and collaborative processing for 6G intelligent applications by utilizing localized computing resources.

[0156] In some embodiments, the edge side of the wireless access network can refer to the near-end network entity that can be uniformly managed by the SMO through the O1 interface and directly carry out the deployment and execution of AI models. Specifically, it can include near real-time RIC, enhanced base station / DU, edge cloud server and terminal with computing power, etc. Together, they constitute the key intermediate processing layer in end-to-end AI collaboration, receive model configurations issued by the SMO, perform localized AI inference or lightweight training, and provide real-time feedback of running data to achieve a low-latency and highly reliable edge intelligent closed loop.

[0157] Step S704: The service management and orchestration architecture configures cloud resources and network functions for at least one node, enabling at least one execution node to execute subscribed artificial intelligence services on business data issued by the wireless access network automation application based on cloud resources and network functions.

[0158] In some embodiments, the service management and orchestration architecture may also include an AI operations management function module and an O1 interface.

[0159] The O1 interface is the standard management channel in the O-RAN architecture that connects the service management and orchestration system with the radio access network element. It is used to issue configuration commands (including AI model deployment) to nodes such as base stations and intelligent controllers and to collect their performance, fault and other operational data. In this application, it constitutes the core neural link for SMO to perform closed-loop monitoring and dynamic optimization of edge-side AI services.

[0160] Among them, the artificial intelligence operation and management function module can be, for example, the Operation Platform for E2E AI (end-to-end artificial intelligence operation platform). This end-to-end artificial intelligence operation platform can be used to enhance the SMO RAN OAM-related functions, add the SMO's monitoring, configuration and management functions for the deployment and operation of corresponding AI algorithms or models on the RAN and UE sides, configure the RAN through the O1 interface, and trigger the RAN to initiate the Uu interface process to control the AI ​​on the UE side.

[0161] In some embodiments, an end-to-end AI operations platform may include the following functionalities.

[0162] The Function Orchestration for Radio AI (FO4RAI) is responsible for the unified management of the aforementioned functions related to AI algorithms or models deployed on E2 nodes, nRT RICs, and the UE side. It collects performance management data (PM), configuration management data (CM), and fault management data (FM) log data related to the configuration and operation of AI algorithms or models on E2 nodes, nRT RICs, and the UE side via the O1 interface. This includes information on the success of AI model deployment, whether the performance in radio functions (including wireless network functions and terminal wireless transmission functions) meets standards, the storage space occupied by the AI ​​model, the data required for AI model training, the deviation between the AI ​​model's operation and the expected value, and the input measurement data for modification. Based on updates to the AI ​​algorithms or models running in radio functions, it reconfigures AI-related information on E2 nodes, nRT RICs, and the terminal side.

[0163] In some embodiments, the AI ​​operation management module can collect operational status information generated during the operation of subscribed AI services at the edge via the O1 interface. The operational status information includes at least one of performance management data, configuration management data, and fault management data.

[0164] The core function of the above technical solution is to build a refined and real-time monitoring and management capability for the SMO of the operational status of AI services on the edge side: by actively collecting performance, configuration and fault data generated by AI models on edge nodes (such as base stations and terminals) during inference or training through the artificial intelligence operation management module, the SMO can accurately grasp the real-time operational health, resource utilization efficiency and task execution effect of AI services, thereby providing key data support for dynamically optimizing model deployment strategies, quickly locating and recovering anomalies, and realizing cross-device-edge-cloud collaborative optimization, ultimately ensuring the end-to-end service quality and reliability of wireless AI services.

[0165] Figure 8 This is a flowchart illustrating an edge-side configuration update method according to an exemplary embodiment.

[0166] In some embodiments, the service management and orchestration architecture may also include an AI operations management function module and an O1 interface.

[0167] refer to Figure 8 The above-mentioned edge-side configuration update method may include the following steps.

[0168] In step S802, the AI ​​operation management module collects the operation status information generated by the subscribed AI services during their operation at the edge via the O1 interface. The operation status information includes at least one of performance management data, configuration management data, and fault management data.

[0169] In step S804, the AI ​​operation management module updates the configuration of the subscribed AI services on the edge side based on the operation status information generated during the operation of the subscribed AI services on the edge side.

[0170] Figure 9 This is a flowchart illustrating an artificial intelligence service management and orchestration method according to an exemplary embodiment.

[0171] refer to Figure 9 The above-mentioned artificial intelligence service management and orchestration methods may include the following steps.

[0172] In step S902, the artificial intelligence resource management function module receives a configuration update request from the service management and open function module. The configuration update request contains summary information of artificial intelligence services subscribed by the radio access network automation application.

[0173] In step S904, the AI ​​resource management and open function unit collects operational status data of AI-related computing and storage resources from the cloud via the O2 interface. The operational status data includes at least one of performance management data, configuration management data, and fault management data.

[0174] In step S906, the AI ​​resource management and open function unit generates AI resource configurations for subscribed AI services based on the operational status data.

[0175] In step S908, the AI ​​network function orchestration unit collects at least one of the following from the cloud via the O2 interface: performance management data, configuration management data, and fault management data related to AI network functions.

[0176] In step S910, the AI ​​network function orchestration unit generates an AI network function orchestration configuration based on at least one of performance management data, configuration management data, and fault management data related to the AI ​​network function.

[0177] In step S912, the AI ​​resource management module distributes the AI ​​resource configuration and AI network function orchestration configuration to the cloud, thereby triggering the cloud to provide resource supply and network function orchestration for subscribed AI services based on the AI ​​resource configuration and AI network function orchestration configuration.

[0178] In step S914, the service management and orchestration architecture selects at least one node on the edge side of the wireless access network to carry the subscribed artificial intelligence service.

[0179] Step S916: The service management and orchestration architecture configures cloud resources and network functions for at least one node, enabling at least one execution node to execute subscribed artificial intelligence services on business data issued by the wireless access network automation application based on cloud resources and network functions.

[0180] In step S918, the AI ​​operation management module collects the operation status information generated by the subscribed AI services during their operation on the edge through the O1 interface. The operation status information includes at least one of performance management data, configuration management data, and fault management data.

[0181] In step S920, the AI ​​operation management module updates the configuration of the subscribed AI services on the edge side based on the operation status information generated during the operation of the subscribed AI services on the edge side.

[0182] Figure 10 This is a flowchart illustrating an artificial intelligence service management and orchestration method according to an exemplary embodiment.

[0183] In some embodiments, the AI ​​service management and orchestration method can be implemented by a service management and orchestration architecture for wireless access networks, wherein the service management and orchestration architecture includes an AI operation management function module.

[0184] refer to Figure 10The above-mentioned artificial intelligence service management and orchestration methods may include the following steps.

[0185] In step S1002, the service management and orchestration architecture selects at least one node on the edge side of the wireless access network to carry the subscribed artificial intelligence service.

[0186] Step S1004: The AI ​​operation management module collects the operation status information generated during the operation of the subscribed AI service on the edge from at least one node. The operation status information includes at least one of performance management data, configuration management data, and fault management data. The operation status information is used to update the configuration of the subscribed AI service on the edge.

[0187] In some embodiments, the aforementioned AI operation management module can update the configuration of the subscribed AI service at the edge based on the operational status information generated during the operation of the subscribed AI service at the edge.

[0188] In some embodiments, this application also provides a service management and orchestration architecture.

[0189] like Figure 4 As shown, the service management and orchestration architecture may include an artificial intelligence resource management module (such as Resource Platform for E2E AI), a service management and exposure function module, and a radio access network automation application (rAPP).

[0190] In some embodiments, the above service management and orchestration architecture may not include an AI operations management module, such as... Figure 4 Operation Platform for E2E AI (in the context of AI).

[0191] The AI ​​resource management module can receive configuration update requests from the service management and open function module. The configuration update requests include summary information of AI services subscribed by the wireless access network automation application. Based on the configuration update requests, it generates AI resource configuration and AI network function orchestration configuration. It then distributes the AI ​​resource configuration and AI network function orchestration configuration to the cloud to trigger the cloud to provide resource supply and network function orchestration for the subscribed AI services based on the AI ​​resource configuration and AI network function orchestration configuration.

[0192] In some embodiments, this application also provides a service management and orchestration architecture.

[0193] like Figure 4As shown, the service management and orchestration architecture may include AI operations management functional modules (such as...). Figure 4 The Operation Platform for E2E AI is used to collect operational status information generated during the operation of subscribed AI services at the edge from at least one node. The operational status information includes at least one of performance management data, configuration management data, and fault management data. The operational status information is used to update the configuration of subscribed AI services at the edge.

[0194] In some embodiments, the above service management and orchestration architecture may not include an artificial intelligence resource management module (such as...). Figure 4 The Resource Platform for E2E AI (in the context of AI).

[0195] In some embodiments, to address the issues existing in the SMO framework's support for AI / ML services, this application also enhances and extends the SMO architecture to support unified management and orchestration of end-to-end AI / ML services, specifically as follows: Figure 4 The Resource Platform for E2E AI and Operation Platform for E2E AI are included, and the functions related to the O1 and O2 interfaces are expanded or added based on the corresponding new functions.

[0196] In some embodiments, to enable orchestration of end-to-end AI resources and AI functions, a Resource Platform for E2E AI module and an Operation Platform for E2E AI module can be added to the SMO, such as... Figure 4 As shown.

[0197] Based on the O2 interface, the Resource Platform for E2E AI module provides management and orchestration functions for computing and storage resources for AI algorithm or model deployment, computation, inference and storage, as well as unified orchestration services for AI network functions deployed in O-Cloud.

[0198] Based on the O1 interface, the Operation Platform for E2E AI module provides monitoring, configuration and management of the deployment and operation of corresponding AI algorithms or models on the RAN and UE sides, as well as unified orchestration services for AI network functions deployed on the RAN side.

[0199] The Resource Platform for E2E AI enhances SMO O2-related functions, adding management and orchestration capabilities for computing and storage resources used in AI algorithm or model deployment, computation, inference, and storage. It also enhances the orchestration of AI network functions running on the O-Cloud, collects management data on the allocation and operational status of AI-related computing and storage resources from the O-Cloud via the O2 interface, and updates resource configurations as needed. It includes the following functions: (1) AI Resource Management and Exposure Functions (RME4AI): As a component of the FOCOM function, it provides management and exposure of AI-related computing and storage resources on the O-Cloud infrastructure. It collects performance management data (PM), configuration management data (CM), and fault management data (FM) related to computing and storage resources for AI algorithm or model deployment / computation, inference, and storage reported from the O2 interface. It provides functions such as computing / storage resource awareness, computing / storage resource registration, computing / storage resource exposure, and computing / storage resource configuration. It generates configuration updates for AI resources and sends them to O-Cloud. (2) Function orchestration for AI in Network (FOAIN): FOAIN includes O-Cloud's allocation, reclamation, and orchestration services for computing and storage resources for deployed AI-related network functions. As a component of the NFO function, FOAIN manages the relevant computing or storage resources through the O2 interface, and collects O-Cloud's AI network function-related performance management data (PM), configuration management data (CM), and fault management data (FM) from the O2 interface; supports the deployment of O-Cloud AI model training, management, and opening (including AI model registration, discovery, subscription / change of subscription, storage, and optional training capability registration / deregistration / query, etc.), performance monitoring, inference, and other functions; and generates configuration updates for AI network functions and sends them to O-Cloud.

[0200] Among them, the Operation Platform for E2E AI enhances the SMO RAN OAM-related functions, adding monitoring, configuration, and management capabilities for the deployment and operation of corresponding AI algorithms or models on the RAN and UE sides. It configures the RAN through the O1 interface and triggers the RAN to initiate Uu interface procedures to control the AI ​​on the UE side. It includes the following functions: (1) Function orchestration for Radio AI (FO4RAI): Unified management of the above-mentioned functions related to AI algorithms or models deployed on E2 nodes, nRT RICs and UEs, and collection of performance management data (PM), configuration management data (CM) and fault management data (FM) of AI algorithms or models configured and running on E2 nodes, nRT RICs and UEs through the O1 interface. This includes information on whether the AI ​​model deployment is successful, whether the performance in the radio function (including the wireless network function and the terminal wireless transmission function) meets the standard, the storage space occupied by the AI ​​model, the data required for AI model training, the deviation between the AI ​​model and the expected value during operation, and the input measurement data for modification, etc.; and reconfigure the AI-related information of E2 nodes, nRT RICs and terminals based on the updates of the AI ​​algorithm or model during operation in the radio function.

[0201] Below, this application will be approved. Figure 11 This specific embodiment explains and illustrates the artificial intelligence service management and function orchestration method proposed in this application.

[0202] Figure 11 This is an interactive schematic diagram illustrating a wireless AI end-to-end collaborative SMO management method based on smart glasses, according to an exemplary embodiment.

[0203] It should be noted that those skilled in the art can extend the technical solutions provided in this embodiment to other smart devices.

[0204] refer to Figure 11 As shown, the above-mentioned wireless AI end-to-end collaborative SMO management method based on smart glasses.

[0205] Step 1: The user terminal (UE) selects an AI service based on the smart glasses and associates it with the rAPP. The rAPP sends an rAPP registration request to the rAPP management function through the R1 interface, and the rAPP management function notifies the rAPP that registration is complete.

[0206] Step 2: rAPP sends a service discovery request to Service Management and Open Functionality (SME). SME searches for AI services (AI algorithms or models) that meet rAPP's selection criteria, generates a list of AI services, sends a service notification to rAPP, and opens up the list of currently available AI services.

[0207] Step 3: rAPP sends an AI service subscription request to SME based on business needs and the AI ​​service list. SME then compiles the AI ​​service subscription information.

[0208] Step 4: The SME sends a configuration update request to the Resource Platform for E2E AI and sends summary information about the AI ​​service subscription via rAPP. On the Resource Platform for E2E AI, AI Resource Management and Open Functionality (RME4AI) generates AI resource configuration updates, and AI Network Function Orchestration (FOAIN) generates AI network function configuration updates. These configuration updates are then distributed to O-Cloud via the O2 interface. The Resource Platform for E2E AI sends a configuration update completion notification to the SME. The SMO determines the AI ​​service execution nodes / node groups (E2 nodes, nRT RIC, etc.), and O-Cloud configures the corresponding AI resources and AI network functions for the AI ​​service execution nodes / node groups.

[0209] Step 5: rAPP sends a data registration request to the Data Management and Open Functionality (DME), and the DME notifies rAPP that data registration is complete. rAPP then sends its business data to the DME, which forwards the rAPP's business data to the corresponding AI service execution node / node group.

[0210] Step 6: The AI ​​service execution node / node group runs the AI ​​algorithm or model based on the rAPP service data, collecting performance management data (PM), configuration management data (CM), and fault management data (FM) of the AI ​​algorithm or model during configuration and operation in the E2 node, nRT RIC, and UE side. This AI runtime data is then reported to the Operation Platform for E2E AI via the O1 interface. The Operation Platform for E2E AI's wireless AI network function configuration (FO4RAI) generates a configuration update for the wireless AI network function based on the AI ​​runtime data and distributes it to the AI ​​service execution node / node group and / or the UE side via the O1 interface. The AI ​​service execution node / node group completes the reconfiguration of the wireless AI network function and repeats the AI ​​service execution. Based on the AI ​​service execution results, the UE side can choose to resubscribe to the AI ​​service or perform other operations via rAPP, executing steps 1-6.

[0211] Step 7: SME periodically or non-periodically updates the list of AI services for rAPP subscription.

[0212] This embodiment proposes a wireless AI end-to-end collaborative SMO management scheme for RAN and provides a specific architecture design to support unified management and orchestration of end-to-end AI / ML services.

[0213] This embodiment proposes adding a Resource Platform for E2E AI module to the SMO and provides a specific architecture design method. Based on the O2 interface, the Resource Platform for E2E AI module provides management and orchestration functions for computing and storage resources for AI algorithm or model deployment, computation, inference, and storage, as well as unified orchestration services for AI network functions deployed in O-Cloud.

[0214] This embodiment proposes the design of AI resource management and open functions (RME4AI) and AI network function orchestration (FOAIN) based on the Resource Platform for E2E AI module, and provides specific functional designs.

[0215] This embodiment proposes adding an Operation Platform for E2E AI to the SMO and provides a specific architecture design. Based on the O1 interface, the Operation Platform for E2E AI module provides monitoring, configuration, and management of the deployment and operation of corresponding AI algorithms or models on the RAN and terminal (UE) sides, as well as unified orchestration services for AI network functions deployed on the RAN side.

[0216] Based on the Operation Platform for E2E AI module, this embodiment designs a wireless AI network function configuration (FO4RAI) and provides specific functional design.

[0217] It should be particularly noted that the steps in each embodiment of the above-described artificial intelligence service management and orchestration method can be overlapped, substituted, added, or deleted from each other. Therefore, these reasonable permutations and combinations of the artificial intelligence service management and orchestration method should also fall within the protection scope of this disclosure, and the protection scope of this disclosure should not be limited to the embodiments.

[0218] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a portion of a module or program segment containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer program instructions.

[0219] Furthermore, the above figures are merely illustrative of the processes included in the method according to exemplary embodiments of this disclosure and are not intended to be limiting. It is readily understood that the processes shown in the above figures do not indicate or limit the temporal order of these processes. Additionally, it is readily understood that these processes may be executed synchronously or asynchronously, for example, in multiple modules.

[0220] Figure 12 A schematic diagram of an electronic device suitable for implementing embodiments of the present disclosure is shown. It should be noted that... Figure 12 The illustrated electronic device 1200 is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments disclosed herein.

[0221] like Figure 12 As shown, the electronic device 1200 includes a central processing unit (CPU) 1201, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1202 or a program loaded from a storage section 1208 into a random access memory (RAM) 1203. The RAM 1203 also stores various programs and data required for the operation of the electronic device 1200. The CPU 1201, ROM 1202, and RAM 1203 are interconnected via a bus 1204. An input / output (I / O) interface 1205 is also connected to the bus 1204.

[0222] The following components are connected to I / O interface 1205: an input section 1206 including a keyboard, mouse, etc.; an output section 1207 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 1208 including a hard disk, etc.; and a communication section 1209 including a network interface card such as a LAN card, modem, etc. The communication section 1209 performs communication processing via a network such as the Internet. A drive 1210 is also connected to I / O interface 1205 as needed. A removable medium 1211, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 1210 as needed so that computer programs read from it can be installed into storage section 1208 as needed.

[0223] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable storage medium, the computer program containing computer program instructions for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 1209, and / or installed from removable medium 1211. When the computer program is executed by central processing unit (CPU) 1201, it performs the functions defined above in the system of this disclosure.

[0224] It should be noted that the computer-readable storage medium disclosed herein may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this disclosure, a computer-readable storage medium may 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. In this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable computer program instructions. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media can also be any computer-readable storage medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. Computer program instructions contained on a computer-readable storage medium can be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0225] In another aspect, this disclosure also provides a computer-readable storage medium, which may be included in the device described in the above embodiments, or it may exist independently and not assembled into the device. The computer-readable storage medium carries one or more programs that, when executed by the device, enable the device to perform the following functions: receiving a configuration update request from a service management and open function module via an artificial intelligence resource management function module, the configuration update request containing aggregated information of artificial intelligence services subscribed by the wireless access network automation application; generating an artificial intelligence resource configuration and an artificial intelligence network function orchestration configuration based on the configuration update request via the artificial intelligence resource management function module; and distributing the artificial intelligence resource configuration and the artificial intelligence network function orchestration configuration to the cloud via the artificial intelligence resource management function module, thereby triggering the cloud to provide resource provisioning and network function orchestration for the subscribed artificial intelligence services based on the artificial intelligence resource configuration and the artificial intelligence network function orchestration configuration.

[0226] According to one aspect of this disclosure, a computer program product or computer program is provided, comprising computer program instructions stored in a computer-readable storage medium. The computer program instructions are read from the computer-readable storage medium, and a processor executes the computer program instructions to implement the methods provided in various optional implementations of the above embodiments.

[0227] From the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solutions of the embodiments of this disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, or portable hard drive) and includes several computer program instructions to cause an electronic device (such as a server or terminal device) to execute the method according to the embodiments of this disclosure.

[0228] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the claims.

[0229] It should be understood that this disclosure is not limited to the detailed structures, drawing arrangements or implementations shown herein; rather, this disclosure is intended to cover various modifications and equivalent arrangements contained within the spirit and scope of the appended claims.

Claims

1. A method for managing and orchestrating artificial intelligence services, characterized in that, The method is implemented by a service management and orchestration architecture for radio access networks, wherein the service management and orchestration architecture includes an artificial intelligence resource management module, a service management and open function module, and radio access network automation applications. The method includes: The AI ​​resource management module receives a configuration update request from the service management and open function module. The configuration update request includes aggregated information of AI services subscribed to by the wireless access network automation application. The AI ​​resource management module generates AI resource configuration and AI network function orchestration configuration based on the configuration update request. The AI ​​resource management module distributes the AI ​​resource configuration and the AI ​​network function orchestration configuration to the cloud, thereby triggering the cloud to provide resource supply and network function orchestration for subscribed AI services based on the AI ​​resource configuration and the AI ​​network function orchestration configuration.

2. The method according to claim 1, characterized in that, The service management and orchestration architecture also includes an O2 interface; the artificial intelligence resource management function module includes artificial intelligence resource management and open function units; wherein, the method further includes: The AI ​​resource management and open function unit collects operational status data of AI-related computing and storage resources from the cloud through the O2 interface. The operational status data includes at least one of performance management data, configuration management data, and fault management data.

3. The method according to claim 2, characterized in that, The AI ​​resource management module generates AI resource configuration based on the configuration update request, including: The AI ​​resource management and open function unit generates the AI ​​resource configuration for the subscribed AI services based on the operational status data.

4. The method according to claim 1, characterized in that, The service management and orchestration architecture also includes an O2 interface; the artificial intelligence resource management function module includes an artificial intelligence network function orchestration unit; the method further includes: The artificial intelligence network function orchestration unit collects at least one of the following from the cloud through the O2 interface: performance management data, configuration management data, and fault management data related to the artificial intelligence network function.

5. The method according to claim 4, characterized in that, The AI ​​resource management module generates an AI network function orchestration configuration based on the configuration update request, including: The artificial intelligence network function orchestration unit generates the artificial intelligence network function orchestration configuration based on at least one of the performance management data, configuration management data, and fault management data related to the artificial intelligence network function.

6. The method according to claim 1, characterized in that, The method further includes: The service management and orchestration architecture selects at least one node at the edge of the wireless access network to carry the subscribed artificial intelligence service; The service management and orchestration architecture configures cloud resources and network functions for the at least one node, enabling the at least one execution node to execute the subscribed artificial intelligence service based on the cloud resources and network functions and the service data issued by the wireless access network automation application.

7. The method according to claim 1, characterized in that, The service management and orchestration architecture also includes an AI operations management module and an O1 interface; the method further includes: The AI ​​operation management module collects the operational status information generated by the subscribed AI service during its operation at the edge through the O1 interface. The operational status information includes at least one of performance management data, configuration management data, and fault management data.

8. The method according to claim 7, characterized in that, The method further includes: The AI ​​operation management module updates the configuration of the subscribed AI service on the edge side based on the operation status information generated during the operation of the subscribed AI service on the edge side.

9. A method for managing and orchestrating artificial intelligence services, characterized in that, The method is implemented by a service management and orchestration architecture for wireless access networks, wherein the service management and orchestration architecture includes an artificial intelligence operations management function module, and the method includes: The service management and orchestration architecture selects at least one node at the edge of the wireless access network to carry subscribed artificial intelligence services. The AI ​​operation management module collects operational status information generated by the subscribed AI service during its operation at the edge from the at least one node. The operational status information includes at least one of performance management data, configuration management data, and fault management data. The operational status information is used to update the configuration of the subscribed AI service at the edge.

10. The method according to claim 9, characterized in that, The method further includes: The AI ​​operation management module updates the configuration of the subscribed AI service on the edge side based on the operation status information generated during the operation of the subscribed AI service on the edge side.

11. A service management and orchestration architecture, characterized in that, This includes an AI resource management module, a service management and open functionality module, and wireless access network automation applications; The AI ​​resource management module is configured to receive configuration update requests from the service management and open function module, the configuration update requests containing aggregated information of AI services subscribed by the radio access network automation application; and generate AI resource configuration and AI network function orchestration configuration based on the configuration update requests. The AI ​​resource configuration and the AI ​​network function orchestration configuration are sent to the cloud to trigger the cloud to provide resource supply and network function orchestration for subscribed AI services based on the AI ​​resource configuration and the AI ​​network function orchestration configuration.

12. A service management and orchestration architecture, characterized in that, include: The AI ​​operation management function module is used to collect operational status information generated during the operation of subscribed AI services at the edge from at least one node. The operational status information includes at least one of performance management data, configuration management data, and fault management data. The operational status information is used to update the configuration of the subscribed AI services at the edge.

13. An electronic device, characterized in that, include: Memory and processor; The memory is used to store computer program instructions; the processor calls the computer program instructions stored in the memory to implement the artificial intelligence service management and orchestration method as described in any one of claims 1-8, 9-10.

14. A computer-readable storage medium storing computer program instructions thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the artificial intelligence service management and orchestration method as described in any one of claims 1-8 or as described in any one of claims 9-10.

15. A computer program product comprising computer program instructions stored in a computer-readable storage medium, characterized in that, When the computer program instructions are executed by the processor, they implement the artificial intelligence service management and orchestration method according to any one of claims 1-8 or the artificial intelligence service management and orchestration method according to any one of claims 9-10.

Images