Machine model reasoning capability management and control method and device

Abstract

The invention provides a method and a device for managing and controlling the reasoning ability of a machine model, which are used for managing and controlling the reasoning ability of the machine model and relate to the technical field of wireless communication. In the method, a second communication device sends first information to a first communication device, and the first information comprises AI / ML capability corresponding to AI / ML reasoning. The second communication device receives an inference result of the AI / ML inference from the first communication device. Based on the above scheme, the second communication device can indicate the AI / ML capability corresponding to AI / ML reasoning to the first communication device through the first information, thereby realizing management and control of the AI / ML capability, and enabling the first communication device to execute AI / ML reasoning of the AI / ML capability according to the first information.

Description

Technical Field

[0001] The present application relates to the field of wireless communication technology, and in particular to a method and device for managing and controlling machine model reasoning capabilities. Background Art

[0002] Currently, the Third Generation Partnership Project (3 rd The Third Generation Partnership Project (3GPP) defines the use of artificial intelligence / machine learning (AI / ML) technology to achieve radio access network (RAN) intelligence, specifically to support energy saving (ES), mobility load balancing (MLB), and mobility robustness optimization (MRO). The standard discussion also defines how to support and manage RAN intelligent reasoning type scenarios on the network management. Specific management information includes switch management and policy management of RAN intelligent reasoning types, that is, the cross-domain management node can set switches for the use case and configure execution policies. The standard also discusses how to support AI / ML management data analytics (MDA) reasoning type scenarios.

[0003] However, the management of the reasoning type of RAN intelligence in the standard only supports the management of switches and execution policies, and does not propose the management of the AI / ML capabilities of RAN intelligence. Summary of the invention

[0004] The present application provides a method and device for managing the reasoning capability of a machine model, which are used to manage the reasoning capability of a machine model.

[0005] In a first aspect, a method for managing and controlling machine model reasoning capabilities is provided. The method may be performed by a second communication device. The second communication device may be a second management device, or a chip / chip system. In the method, the second communication device sends a first message to the first communication device, and the first message includes the AI / ML capability corresponding to the AI / ML reasoning. The second communication device receives the reasoning result of the AI / ML reasoning from the first communication device.

[0006] Based on the above solution, the second communication device can, through the first information, indicate to the first communication device the AI / ML capabilities corresponding to the AI / ML inference, implement the control of the AI / ML capabilities, and enable the first communication device to perform the AI / ML inference of the AI / ML capabilities according to the first information.

[0007] In a possible implementation manner, the first information is used to request the first communication device to perform the AI / ML inference of the AI / ML capabilities. Based on this solution, the first information can request to perform the AI / ML inference of the AI / ML capabilities, that is, the second communication device can, based on the first information, control the AI / ML capabilities.

[0008] In a possible implementation manner, the second communication device receives the AI / ML capability information supported by the first communication device. The AI / ML capabilities indicated by the AI / ML capability information include the AI / ML capabilities corresponding to the AI / ML inference.

[0009] Based on the above solution, the second communication device can obtain the AI / ML capability information supported by the first communication device, thereby enabling it to control the AI / ML capabilities supported by the first communication device.

[0010] In a possible implementation manner, the second communication device sends the second information to the first communication device, and the second information is used to request the first communication device to send the AI / ML capability information.

[0011] Based on the above solution, the second communication device can request the first communication device to send the supported AI / ML capability information, thereby obtaining the AI / ML capabilities supported by the first communication device.

[0012] In a possible implementation manner, the second information instructs the first communication device to send the AI / ML capability information supported by one or more of the management data analysis function, the mobility optimization function, the network energy saving function, or the load balancing function.

[0013] Based on the above solution, the second communication device can, through the second information, instruct the first communication device to send the AI / ML capability information supported by the inference types such as the management data analysis function, the mobility optimization function, the network energy saving function, or the load balancing function, thereby realizing the control of the AI / ML capabilities indicated by the inference types such as the management data analysis function, the mobility optimization function, the network energy saving function, or the load balancing function from the AI / ML capability information.

[0014] In a possible implementation manner, the second communication device sends the inference types of the AI / ML capabilities, and the inference types include one or more of the management data analysis function, the mobility optimization function, the network energy saving function, or the load balancing function.

[0015] Based on the above solution, the second communication device can send to the first communication device which inference types the AI / ML capabilities indicated by the first information correspond to, so as to manage and control the AI / ML capabilities of specific inference types.

[0016] In a possible implementation, the AI / ML capabilities include one or more of the following: traffic analysis capability, traffic prediction capability, mobility analysis capability, mobility prediction capability, load analysis capability, load prediction capability, energy consumption analysis capability, or energy consumption prediction capability.

[0017] In a possible implementation, the second communication device receives one or more of the inference status or the inference time.

[0018] Based on the above solution, the first communication device can also send to the second communication device one or more of the inference status or the inference time, so that the second communication device can learn one or more of the inference status or the inference time, which is convenient for the second communication device to manage and control the AI / ML inference in a timely manner.

[0019] In a possible implementation, the inference result further includes the inference type of the inference result, and the inference type includes one or more of a mobility optimization function, a network energy saving function, or a load balancing function.

[0020] Based on the above solution, the first communication device can send to the second communication device the inference type of the inference result, that is, send which inference types can use the inference result, so that the second communication device can execute corresponding functions based on the inference result.

[0021] In a possible implementation, the inference result includes one or more of the following: cell personality offset, cell identifier, network device identifier, time trigger, handover trigger, information on shutting down a cell, information on shutting down a carrier, or information on shutting down a time slot.

[0022] In a second aspect, a method for managing and controlling the machine model inference capability is provided. This method can be executed by the first communication device. Wherein, the first communication device can be the first management device, or a chip / chip system. In this method, the first communication device receives first information from the second communication device, and the first information includes the AI / ML capabilities corresponding to the AI / ML inference. The first communication device performs the AI / ML inference of the AI / ML capabilities to obtain an inference result. The first communication device sends the inference result to the second communication device.

[0023] In a possible implementation, the first information is used to request the first communication device to perform the AI / ML inference of the AI / ML capabilities.

[0024] In a possible implementation, the first communication device sends the supported AI / ML capability information to the second communication device. The AI / ML capabilities indicated by the AI / ML capability information include the AI / ML capabilities corresponding to AI / ML inference.

[0025] In a possible implementation, the first communication device receives second information from the second communication device, where the second information is used to request the first communication device to send AI / ML capability information.

[0026] In a possible implementation, the second information instructs the first communication device to send the AI / ML capability information supported by one or more of the management data analysis function, mobility optimization function, network energy saving function, or load balancing function.

[0027] In a possible implementation, the first communication device receives the inference types of the AI / ML capabilities, where the inference types include one or more of the management data analysis function, mobility optimization function, network energy saving function, or load balancing function

[0028] In a possible implementation, the AI / ML capabilities include one or more of the following: traffic analysis capability, traffic prediction capability, mobility analysis capability, mobility prediction capability, load analysis capability, load prediction capability, energy consumption analysis capability, or energy consumption prediction capability.

[0029] In a possible implementation, the first communication device sends one or more of the inference status or inference time.

[0030] In a possible implementation, the inference result further includes the inference type of the inference result, where the inference type includes one or more of the mobility optimization function, network energy saving function, or load balancing function.

[0031] In a possible implementation, the inference result includes one or more of the following: cell personality offset, cell identifier, network device identifier, time trigger, handover trigger, information on shutting down a cell, information on shutting down a carrier, or information on shutting down a time slot.

[0032] In a third aspect, a method for controlling the machine model inference capability is provided. This method can be executed by the second communication device. Wherein, the second communication device can be a second management device, or a chip / chip system. In this method, the second communication device receives the supported AI / ML capability information from the first communication device. Wherein, the AI / ML capability information indicates one or more AI / ML capabilities. The AI / ML capability information is used for the control of one or more AI / ML capabilities.

[0033] Based on the above solution, the second communication device can obtain the AI / ML capabilities supported by the first communication device, and thus can control the AI / ML capabilities supported by the first communication device. In some embodiments, the AI / ML capability information is used to control one or more AI / ML capabilities. It can be understood that the control may be that the cross-domain management functional unit triggers the inference of the AI / ML capabilities of the domain management functional unit or the inference of the inference type of AI / ML of the domain management functional unit based on the received AI / ML capability information. It should also be noted that the inference of the AI / ML capabilities here can be understood as training an ML model based on the AI / ML capabilities and then applying it to the inference type of AI / ML to complete the AI / ML inference function corresponding to the inference type of AI / ML.

[0034] In a possible implementation, the second communication device receives the inference type of AI / ML to which the AI / ML capability information from the first communication device applies. Optionally, the inference type of AI / ML includes one or more of a management data analysis function, a mobility optimization function, a network energy saving function, or a load balancing function.

[0035] Based on the above solution, the second communication device can obtain the inference type of AI / ML to which the AI / ML capability information applies, and thus can control the AI / ML capabilities of one or more inference types of the first communication device.

[0036] In a possible implementation, the second communication device sends a second message to the first communication device, and the second message is used to request the first communication device to send AI / ML capability information.

[0037] Based on the above solution, the second communication device can request the first communication device to send the supported AI / ML capability information, so as to obtain the AI / ML capabilities supported by the first communication device.

[0038] In a possible implementation, the second message instructs the first communication device to send the AI / ML capability information supported by one or more of a management data analysis function, a mobility optimization function, a network energy saving function, or a load balancing function.

[0039] Based on the above solution, the second communication device can, through the second message, instruct the first communication device to send the AI / ML capability information supported by the inference types such as a management data analysis function, a mobility optimization function, a network energy saving function, or a load balancing function, so as to realize the control of the AI / ML capability information indicated by the inference types such as a management data analysis function, a mobility optimization function, a network energy saving function, or a load balancing function.

[0040] In a possible implementation, the second communication device sends first information to the first communication device, and the first information includes AI / ML capabilities corresponding to AI / ML inference. Among them, one or more AI / ML capabilities indicated by the AI / ML capability information include the AI / ML capabilities indicated by the first information. The second communication device receives the inference result of the AI / ML inference from the first communication device.

[0041] Based on the above solution, the second communication device can, through the first information, indicate to the first communication device the AI / ML capabilities corresponding to AI / ML inference, realize the control of AI / ML capabilities, and enable the first communication device to perform AI / ML inference on AI / ML capabilities according to the first information.

[0042] In a possible implementation, the first information is used to request the first communication device to perform AI / ML inference on the AI / ML capabilities indicated by the first information. Based on this solution, the first information can request to perform AI / ML inference on AI / ML capabilities, that is to say, the second communication device can control AI / ML capabilities based on the first information.

[0043] In a possible implementation, the AI / ML capabilities include one or more of the following: traffic analysis capability, traffic prediction capability, mobility analysis capability, mobility prediction capability, load analysis capability, load prediction capability, energy consumption analysis capability, or energy consumption prediction capability.

[0044] In a possible implementation, the second communication device receives one or more of the inference status or inference time. Based on the above solution, the first communication device can also send one or more of the inference status or inference time to the second communication device, so that the second communication device can understand one or more of the inference status or inference time, which is convenient for the second communication device to control the AI / ML inference in a timely manner.

[0045] In a possible implementation, the inference result further includes the inference type of the inference result, and the inference type includes one or more of a mobility optimization function, a network energy saving function, or a load balancing function.

[0046] Based on the above solution, the first communication device can send the inference type of the inference result to the second communication device, that is, send which inference types can use the inference result, so that the second communication device can execute corresponding functions through the inference result.

[0047] In a possible implementation, the inference result includes one or more of the following: cell personality offset, cell identifier, network device identifier, time trigger, handover trigger, information on shutting down a cell, information on shutting down a carrier, or information on shutting down a time slot.

[0048] Fourthly, a method for controlling the inference ability of a machine model is provided. This method can be executed by a first communication device. Among them, the first communication device can be a first management device, or a chip / chip system. In this method, the first communication device sends AI / ML capability information supported to a second communication device. Among them, the AI / ML capability information indicates one or more AI / ML capabilities. The AI / ML capability information is used for the control of one or more AI / ML capabilities.

[0049] In a possible implementation, the first communication device sends the inference type of AI / ML to which the AI / ML capability information applies to the second communication device. Optionally, the inference type of AI / ML includes one or more of a management data analysis function, a mobility optimization function, a network energy saving function, or a load balancing function.

[0050] In a possible implementation, the first communication device receives second information from the second communication device, and the second information is used to request the first communication device to send AI / ML capability information.

[0051] In a possible implementation, the second information instructs the first communication device to send AI / ML capability information supported by one or more of a management data analysis function, a mobility optimization function, a network energy saving function, or a load balancing function.

[0052] In a possible implementation, the first communication device receives first information from the second communication device, and the first information includes the AI / ML capabilities corresponding to the AI / ML inference. Among them, the AI / ML capabilities indicated by the AI / ML capability information include the AI / ML capabilities indicated by the first information. The first communication device sends the inference result of the AI / ML inference to the second communication device.

[0053] In a possible implementation, the first information is used to request the first communication device to perform AI / ML inference of the AI / ML capabilities.

[0054] In a possible implementation, the AI / ML capabilities include one or more of the following: traffic analysis capability, traffic prediction capability, mobility analysis capability, mobility prediction capability, load analysis capability, load prediction capability, energy consumption analysis capability, or energy consumption prediction capability.

[0055] In a possible implementation, the first communication device sends one or more of the inference status or the inference time to the second communication device.

[0056] In a possible implementation, the inference result further includes the inference type of the inference result, and the inference type includes one or more of a mobility optimization function, a network energy saving function, or a load balancing function.

[0057] In a possible implementation, the inference result includes one or more of the following: cell-specific offset, cell identifier, network device identifier, time trigger, handover trigger, information for shutting down the cell, information for shutting down the carrier, or information for shutting down the time slot.

[0058] In a fifth aspect, a communication device is provided, including a processing unit and a transceiver unit.

[0059] The processing unit is configured to generate first information, where the first information includes the AI / ML capabilities corresponding to the AI / ML inference. The transceiver unit is configured to send the first information to a first communication device. The transceiver unit is further configured to receive the inference result of the AI / ML inference from the first communication device.

[0060] In a possible implementation, the first information is used to request the first communication device to perform the AI / ML inference of the AI / ML capabilities.

[0061] In a possible implementation, the transceiver unit is further configured to receive the AI / ML capabilities information supported by the first communication device. The AI / ML capabilities indicated by the AI / ML capabilities information include the AI / ML capabilities corresponding to the AI / ML inference.

[0062] In a possible implementation, the transceiver unit is further configured to send second information to the first communication device, where the second information is used to request the first communication device to send the AI / ML capabilities information.

[0063] In a possible implementation, the second information instructs the first communication device to send the AI / ML capabilities information supported by one or more of the management data analysis function, mobility optimization function, network energy saving function, or load balancing function.

[0064] In a possible implementation, the transceiver unit is further configured to send the inference type of the AI / ML capabilities, where the inference type includes one or more of the management data analysis function, mobility optimization function, network energy saving function, or load balancing function.

[0065] In a possible implementation, the AI / ML capabilities include one or more of the following: traffic analysis capability, traffic prediction capability, mobility analysis capability, mobility prediction capability, load analysis capability, load prediction capability, energy consumption analysis capability, or energy consumption prediction capability.

[0066] In a possible implementation, one or more of the inference status or inference time are received.

[0067] In a possible implementation, the inference result further includes the inference type of the inference result, where the inference type includes one or more of the mobility optimization function, network energy saving function, or load balancing function.

[0068] In a possible implementation, the inference result includes one or more of the following: cell-specific offset, cell identifier, network device identifier, time trigger, handover trigger, information for shutting down the cell, information for shutting down the carrier, or information for shutting down the time slot.

[0069] In a sixth aspect, a communication device is provided, including a processing unit and a transceiver unit.

[0070] The transceiver unit is configured to receive first information from a second communication device, where the first information includes AI / ML capabilities corresponding to AI / ML inference. The processing unit is configured to perform AI / ML inference on the AI / ML capabilities to obtain an inference result. The transceiver unit is further configured to send the inference result to the second communication device.

[0071] In a possible implementation, the first information is used to request the first communication device to perform AI / ML inference on the AI / ML capabilities.

[0072] In a possible implementation, the transceiver unit is further configured to send AI / ML capabilities information it supports to the second communication device. The AI / ML capabilities indicated by the AI / ML capabilities information include the AI / ML capabilities corresponding to AI / ML inference.

[0073] In a possible implementation, the transceiver unit is further configured to receive second information from the second communication device, where the second information is used to request the first communication device to send AI / ML capabilities information.

[0074] In a possible implementation, the second information instructs the first communication device to send AI / ML capabilities information it supports for one or more of the management data analysis function, mobility optimization function, network energy saving function, or load balancing function.

[0075] In a possible implementation, the inference type for the received AI / ML capabilities includes one or more of the management data analysis function, mobility optimization function, network energy saving function, or load balancing function.

[0076] In a possible implementation, the AI / ML capabilities include one or more of the following: traffic analysis capability, traffic prediction capability, mobility analysis capability, mobility prediction capability, load analysis capability, load prediction capability, energy consumption analysis capability, or energy consumption prediction capability.

[0077] In a possible implementation, the transceiver unit is further configured to send one or more of the inference status or inference time.

[0078] In a possible implementation, the inference result further includes the inference type of the inference result, and the inference type includes one or more of a mobility optimization function, a network energy saving function, or a load balancing function.

[0079] In a possible implementation, the inference result includes one or more of the following: cell personality offset, cell identifier, network device identifier, time trigger, handover trigger, information on shutting down a cell, information on shutting down a carrier, or information on shutting down a time slot.

[0080] In a seventh aspect, a communication device is provided, including a processing unit and a transceiver unit.

[0081] The transceiver unit is configured to receive AI / ML capability information supported by a first communication device. The AI / ML capability information indicates one or more AI / ML capabilities, and is used to manage and control the one or more AI / ML capabilities. The processing unit is configured to manage and control the one or more AI / ML capabilities.

[0082] In a possible implementation, the transceiver unit is further configured to receive the AI / ML inference type applicable to the AI / ML capability information from the first communication device. Optionally, the AI / ML inference type includes one or more of a management data analysis function, a mobility optimization function, a network energy saving function, or a load balancing function.

[0083] In a possible implementation, the transceiver unit is further configured to send second information to the first communication device, where the second information is used to request the first communication device to send AI / ML capability information.

[0084] In a possible implementation, the second information instructs the first communication device to send AI / ML capability information supported by one or more of a management data analysis function, a mobility optimization function, a network energy saving function, or a load balancing function.

[0085] In a possible implementation, the transceiver unit is further configured to send first information to the first communication device, where the first information includes the AI / ML capabilities corresponding to the AI / ML inference. The one or more AI / ML capabilities indicated by the AI / ML capability information include the AI / ML capabilities indicated by the first information. The transceiver unit is further configured to receive the inference result of the AI / ML inference from the first communication device.

[0086] In a possible implementation, the first information is used to request the first communication device to perform the AI / ML inference of the AI / ML capabilities indicated by the first information.

[0087] In a possible implementation, the AI / ML capabilities include one or more of the following: traffic analysis capability, traffic prediction capability, mobility analysis capability, mobility prediction capability, load analysis capability, load prediction capability, energy consumption analysis capability, or energy consumption prediction capability.

[0088] In a possible implementation, the second communication device receives one or more of the inference state or inference time.

[0089] In a possible implementation, the inference result further includes the inference type of the inference result, and the inference type includes one or more of a mobility optimization function, a network energy saving function, or a load balancing function.

[0090] In a possible implementation, the inference result includes one or more of the following: cell personality offset, cell identifier, network device identifier, time trigger, handover trigger, information on shutting down a cell, information on shutting down a carrier, or information on shutting down a time slot.

[0091] In a 8th aspect, a communication device is provided, including a processing unit and a transceiver unit.

[0092] The processing unit is configured to determine the supported AI / ML capability information. The transceiver unit is configured to send the supported AI / ML capability information to a second communication device. Wherein, the AI / ML capability information indicates one or more AI / ML capabilities. The AI / ML capability information is used for the management and control of one or more AI / ML capabilities.

[0093] In a possible implementation, the transceiver unit is further configured to send the inference type of the AI / ML to which the AI / ML capability information applies to the second communication device. Optionally, the inference type of the AI / ML includes one or more of a management data analysis function, a mobility optimization function, a network energy saving function, or a load balancing function.

[0094] In a possible implementation, the transceiver unit is further configured to receive a second piece of information from the second communication device, where the second piece of information is used to request the first communication device to send the AI / ML capability information.

[0095] In a possible implementation, the second piece of information instructs the first communication device to send the AI / ML capability information supported by one or more of a mobility optimization function, a network energy saving function, or a load balancing function.

[0096] In a possible implementation, the transceiver unit is further configured to receive first information from a second communication device, where the first information includes AI / ML capabilities corresponding to AI / ML inference. Among them, one or more AI / ML capabilities indicated by the AI / ML capability information include the AI / ML capabilities indicated by the first information. The transceiver unit is further configured to send the inference result of the AI / ML inference to the second communication device.

[0097] In a possible implementation, the first information is used to request the first communication device to perform AI / ML inference on the AI / ML capabilities indicated by the first information.

[0098] In a possible implementation, the AI / ML capabilities include one or more of the following: traffic analysis capability, traffic prediction capability, mobility analysis capability, mobility prediction capability, load analysis capability, load prediction capability, energy consumption analysis capability, or energy consumption prediction capability.

[0099] In a possible implementation, the transceiver unit is further configured to send one or more of the inference status or inference time to the second communication device.

[0100] In a possible implementation, the inference result further includes the inference type of the inference result, and the inference type includes one or more of a mobility optimization function, a network energy saving function, or a load balancing function.

[0101] In a possible implementation, the inference result includes one or more of the following: cell personality offset, cell identifier, network device identifier, time trigger, handover trigger, information on shutting down a cell, information on shutting down a carrier, or information on shutting down a time slot.

[0102] In a ninth aspect, the present application provides a communication system, which may include a second communication device that executes the method described in the first aspect above and a first communication device that executes the method described in the second aspect above.

[0103] In a tenth aspect, the present application provides a communication system, which may include a second communication device that executes the method described in the third aspect above and a first communication device that executes the method described in the fourth aspect above.

[0104] In an eleventh aspect, the present application provides a computer-readable storage medium, where computer-readable instructions are stored in the computer storage medium. When the computer reads and executes the computer-readable instructions, the computer is caused to execute the method in any possible implementation of any one of the first aspect to the fourth aspect above.

[0105] In a twelfth aspect, the present application provides a computer program product, which, when read and executed by a computer, causes the computer to execute the method in any one of the possible implementations of any one of the first to fourth aspects described above.

[0106] In a thirteenth aspect, the present application provides a chip, which is used to read a computer program stored in a memory to execute the method in any one of the possible implementations of any one of the first to fourth aspects described above.

[0107] The technical effects that can be achieved by any one of the second to thirteenth aspects described above can be described with reference to the technical effects that can be achieved by any one of the possible implementations of any one of the first aspects described above. Repeated parts will not be elaborated. Description of the Drawings

[0108] Figure 1 It is a schematic diagram of an MnF entity provided by an embodiment of the present application;

[0109] Figure 2 It is a schematic diagram of a service-oriented management architecture provided by an embodiment of the present application;

[0110] Figure 3A It is a schematic diagram of an architecture of an AI / ML capability provided by the present application;

[0111] Figure 3B It is another schematic diagram of an architecture of an AI / ML capability provided by an embodiment of the present application;

[0112] Figure 4A It is another schematic diagram of an architecture of an AI / ML capability provided by an embodiment of the present application;

[0113] Figure 4B It is another schematic diagram of an architecture of an AI / ML capability provided by an embodiment of the present application;

[0114] Figure 4C It is another schematic diagram of an architecture of an AI / ML capability provided by an embodiment of the present application;

[0115] Figure 5A It is another schematic diagram of an architecture of an AI / ML capability provided by an embodiment of the present application;

[0116] Figure 5B It is another schematic diagram of an architecture of an AI / ML capability provided by an embodiment of the present application;

[0117] Figure 5C It is another schematic diagram of an architecture of an AI / ML capability provided by an embodiment of the present application;

[0118] Figure 6Exemplary flowchart of a method for controlling the inference ability of a machine model provided by an embodiment of the present application;

[0119] Figure 7 Exemplary flowchart of another method for controlling the inference ability of a machine model provided by an embodiment of the present application;

[0120] Figure 8 Exemplary flowchart of another method for controlling the inference ability of a machine model provided by an embodiment of the present application;

[0121] Figure 9 Exemplary flowchart of another method for controlling the inference ability of a machine model provided by an embodiment of the present application;

[0122] Figure 10 Block diagram of a communication device provided by an embodiment of the present application;

[0123] Figure 11 Block diagram of another communication device provided by an embodiment of the present application;

[0124] Figure 12 Block diagram of another communication device provided by an embodiment of the present application;

[0125] Figure 13 Block diagram of another communication device provided by an embodiment of the present application. Detailed implementation

[0126] To facilitate the understanding of the technical solutions provided by the embodiments of the present application, the following explains and describes the technical terms involved in the embodiments of the present application.

[0127] 1) A use case is a technique for obtaining requirements through usage scenarios and can also be referred to as an inference type or inference function of AI / ML, which is not specifically defined in this application. The use cases involved in the embodiments of this application may include self-organizing network (SON) use cases or the inference types of SON. Among them, the inference types of SON may include at least one of mobility robustness optimization (MRO), distributed mobility robustness optimization (DMRO), energy saving (ES), distributed energy saving (DES), mobility load balancing (MLB), or distributed mobility load balancing (DMLB). The use cases involved in the embodiments of this application may include radio access network intelligence (RAN intelligence) use cases or the inference types of radio access network intelligence. Among them, the inference types of radio access network intelligence may include mobility optimization (MRO), network energy saving (NES), and load balancing. The use cases involved in the embodiments of this application may also include management data analytics (MDA) use cases or the inference types of MDA, such as coverage problem analysis, etc., which can refer to the type MDA type in section 8.4 of 3GPP TS28.104 V18.2.0. This application does not expand the description. It may also include network data analytics function (NWDAF) use cases or the inference types of NWDAF, etc., which can refer to the use cases in 3GPP TS23.288. This application does not expand the description.

[0128] Embodiments of the present application can be applied to various mobile communication systems, such as: new radio (NR) systems, long term evolution (LTE) systems, advanced long term evolution (LTE-A) systems, future communication systems and other communication systems. Embodiments of the present application are not limited thereto. Exemplarily, each embodiment in the present application can be used in the network management architecture of NR. The network management architecture of NR may include a management function (MnF). MnF is a management entity defined by the 3rd generation partnership project (3GPP), and its externally visible behaviors and interfaces are defined as management services (MnS). In a management architecture that provides services, MnF can act as a producer or a consumer of MnS. There may be multiple MnS consumers for the MnS produced by the MnS producer of MnF. MnF can consume multiple management services from one or more management service producers.

[0129] As Figure 1 shown, the MnS provided by MnF can be used to provide services for other MnF (such as denoted as MnF#A). At this time, this MnF can be used as an MnS producer, and MnF#A can be used as an MnS consumer. In addition, this MnF can also obtain the services provided by other MnF (denoted as MnF#B, which can be the same as or different from MnF#A), that is, at this time Figure 1 shown, the MnF can be used as an MnS consumer, and MnF#B is used as an MnS producer. That is to say, the same MnF can be used as both an MnS consumer and an MnS producer.

[0130] It should be understood that Figure 1 the graphic of the circle or the graphic of the arc shown in

[0131] Figure 2 can represent a service-based interface. Figure 2Taking two domain management function units and four network elements as an example in China.

[0132] Among them, if MnS is the MnS provided for the cross-domain management function unit, then the cross-domain management function unit is the MnS producer, and the business support system is the MnS consumer. If MnS is the MnS provided for the domain management function unit, then the domain management function unit is the MnS producer, and the cross-domain management function unit is the MnS consumer. If MnS is the MnS provided for the network element, then the network element is the MnS producer, and the domain management function unit is the MnS consumer. It should be understood that the MnS producer and the MnS consumer can also be deployed in one entity, such as being deployed in the business support system, the cross-domain management function unit, the domain management function unit, or the network element.

[0133] In the embodiments of the present application, the cross-domain management function unit can be used to manage one or more domain management function units. The domain management function unit can be used to manage one or more network elements. The following will briefly introduce each unit respectively.

[0134] 1) The business support system is a system oriented to communication services, and is used to provide functions and MnS such as charging, settlement, accounting, customer service, business operation, network monitoring, communication service life cycle management, business intent translation, etc. Among them, the business support system can be the operation system of the operator or the operation system of the vertical industry (vertical OT system).

[0135] 2) The cross-domain management function unit can be a network management entity such as a network management function (NMF), a network function management service consumer (NFMS_C), an MnS producer, an MnS consumer, or a management data analytics (MDA) consumer. Among them, the cross-domain management function unit can provide one or more of the following management functions or MnS: network life cycle management, network deployment, network fault management, network performance management, network configuration management, network guarantee, network optimization function, and translation of the network intent (intent from communication service provider, intent-CSP) of the service producer, etc.

[0136] It should be understood that the "intention" in the embodiments of the present application can be understood as the expectation for the intention producer (such as a network element) and the system where the intention producer is located (such as a network or a sub-network, etc.), and can include requirements, goals, or restrictive conditions, etc. The translation of intention refers to the process of determining the strategy of the intention. For example, the strategy can be used to indicate the conditions that do not meet the intention. For example, when the intention is energy saving, strategy A can be: when the power consumption is greater than the first threshold, the power consumption is abnormal (i.e., not energy saving); strategy B can be: when the power consumption is greater than the second threshold, the power consumption is abnormal (i.e., not energy saving). It can be understood that even for the same intention, the solutions that can meet the intention determined by different strategies may be different.

[0137] Among them, the network referred to in the above management function or MnS can include one or more network elements or sub-networks, or can also be a network slice. That is to say, the network management function unit can be a network slice management function (NSMF) unit, or a cross-domain management data analytical function (MDAF) unit, or a cross-domain self-organization network function (SON Function), or a cross-domain intention management function (intent driven MnS) unit.

[0138] Optionally, in some deployment scenarios, the cross-domain management function unit can also provide life cycle management of the sub-network, deployment of the sub-network, fault management of the sub-network, performance management of the sub-network, configuration management of the sub-network, guarantee of the sub-network, optimization function of the sub-network, translation of the network intention (intent-CSP) of the service producer of the sub-network, or the network intention (intent-CSC) of the service consumer of the sub-network, etc. The sub-network here is composed of multiple small sub-networks and can be a network slice sub-network.

[0139] 3) Domain management function unit, which can be network element management entities such as NMF, element management system (EMS), network function management service provider (NFMS_P), make before break automation engine (MAE), MnS producer, MnS consumer, or MDA producer.

[0140] Among them, the domain management function unit can provide one or more of the following management functions or MnS: lifecycle management of subnets or network elements, deployment of subnets or network elements, fault management of subnets or network elements, performance management of subnets or network elements, assurance of subnets or network elements, optimization functions of subnets or network elements, and translation of intent from network operator (intent-NOP) of subnets or network elements, etc. Here, a subnet includes one or more network elements. A subnet can also include subnets, that is, one or more subnets form a larger subnet.

[0141] Optionally, the subnet here can also be a network slice subnet. The domain management system can be a network slice subnet management function (NSSMF) unit, a management data analytical function (domain MDAF) unit, a self-organization network function (SON Function), a domain intent management function unit, etc.

[0142] Among them, the domain management function unit can be classified in the following ways, including:

[0143] Classification by network type can be divided into: radio access network (RAN) domain management function unit (RAN domain management function, RAN domain MnF), core network domain management function unit (core network domain management function, CN domain MnF), transport network domain management function unit (transport network domain management function, TN domain MnF), etc. It should be noted that the domain management function unit can also be a certain domain network management system, which can manage one or more of the access network, core network, or transport network. The transport network among them is a network used to provide signal transmission and conversion, and is the basic network of the switching network, data network, and support network.

[0144] Classification by administrative region can be divided into: domain management function units in a certain region, such as the domain management function unit in City A, the domain management function unit in City B, etc.

[0145] 4) A network element is an entity that provides network services, including core network elements, radio access network elements, or transport network elements, etc. For example, Figure 2 In the shown architecture, the domain management function unit can include a radio access network domain management function unit, a core network element domain management function unit, or a transport network domain management function unit. Among them, the radio access network domain management function unit can be used to manage radio access network elements, the core network element domain management function unit can be used to manage core network elements, and the transport network domain management function unit can be used to manage transport network elements.

[0146] Exemplarily, the core network elements can include but are not limited to: access and mobility management function (AMF) elements, session management function (SMF) elements, policy control function (PCF) elements, network data analytical function (NWDAF) elements, network repository function (NRF) elements, and gateways, etc.

[0147] The radio access network network elements may include, but are not limited to: various types of base stations (such as next-generation base stations (gNBs), evolved Node Bs (eNBs), etc.), central unit control panels (CUCPs), central units (CUs), distributed units (DUs), central unit user panels (CUUPs), etc.

[0148] It should be understood that the network functions in the embodiments of this application are also referred to as network elements, or entities, etc.

[0149] Among them, the network element may provide one or more of the following management functions or MnS: life cycle management of the network element, deployment of the network element, fault management of the network element, performance management of the network element, guarantee of the network element, optimization function of the network element, and translation of network element intentions, etc.

[0150] AI / ML technologies and related applications are being increasingly adopted by a wider range of industries, and AI / ML capabilities are also used in various fields of 5GS, including intelligent optimization use cases for base stations in the RAN, such as mobility load balancing (MLB), mobility robustness optimization (MRO), and energy saving (ES) use cases, management data analysis services for network management, such as management data analytics (MDA), and network data analysis services for the core network, such as network data analytics function (NWDAF), etc.

[0151] Currently, the management workflow of AI / ML is defined, including the training phase, simulation phase, deployment phase, and inference phase.

[0152] Among them, in the training phase, one or a group of ML models are trained, including initial training and retraining. It also includes the verification of ML entities to evaluate the performance of ML entities on training data and verification data. If the verification result does not meet the expectations, such as unacceptable variance, the ML model associated with the ML entity needs to be retrained. The training phase is the initial phase of the management workflow of AI / ML.

[0153] In the simulation phase, the ML entities for inference are run in a simulation environment. The purpose is to evaluate the inference performance of the ML entities in the simulation environment before applying them to the target network or system.

[0154] Deployment phase, the process of making the trained ML entity available for the target AI / ML inference function.

[0155] Inference phase, the process of performing inference using the ML entity through the AI / ML inference function.

[0156] Currently, the realization of RAN intelligence through AI / ML technology is defined in TR 37.817 and TS 38.300, specifically for supporting NES (Network Energy Saving), LB (Load balancing), and MO (Mobility Optimization), etc. The scenarios of how to support and manage the inference types of RAN intelligence on the network management are also defined in the standard discussion. In some embodiments, the inference types of RAN intelligence may include at least one of Distributed MRO (Distributed Mobility Robustness Optimization, DMRO), Distributed ES (Distributed Energy Saving, DES), Distributed MLB (Distributed Mobility Load Balancing, DMLB), mobility optimization (mobility optimization, MO), network energy saving (network energy saving, NES), and load balancing (load balancing, LB). The specific information to be managed includes the switch management and policy management of the inference types of RAN intelligence, that is, the cross-domain management node can set the switch for the use case and configure the execution policy.

[0157] However, the management of the inference types of RAN intelligence in the standard only supports the management of switches and execution policies, and does not propose the management of the AI / ML capabilities of RAN intelligence.

[0158] In view of this, an embodiment of the present application provides a method for controlling the inference ability of a machine model. In this method, the first communication device can receive first information from the second communication device, and the first information includes the AI / ML capabilities corresponding to the AI / ML inference. The first communication device can perform the AI / ML inference of the AI / ML capabilities to obtain an inference result. The first communication device can send the inference result to the second communication device. Based on this solution, the first communication device can realize the control of the AI / ML capabilities through the first information, enabling the second communication device to perform the AI / ML inference of the AI / ML capabilities according to the first information.

[0159] It should be noted that the term "AI / ML" in this application can be replaced by "AI", "ML", "AIML", etc. AI / ML can be understood as AI and ML, or as AI or ML, or as AI and / or ML.

[0160] To facilitate the understanding of the technical solutions provided in the embodiments of this application, the AI / ML capabilities are explained and described below.

[0161] The AI / ML capabilities involved in the embodiments of this application can be understood as the inference capabilities of AI / ML and the capabilities of AI / ML inference simulation.

[0162] It should be noted that this application does not limit the specific names of the inference capabilities of AI / ML. For example, the inference capabilities of AI / ML can also be called ML inference capabilities, AI inference capabilities, or have other names. The AI / ML inference simulation capabilities can also be called ML inference simulation capabilities, AI inference simulation capabilities, ML simulation capabilities, AI simulation capabilities, or have other names.

[0163] In one possible case, the AI / ML capabilities involved in the embodiments of this application may include one or more of the following: traffic analysis capabilities, coverage analysis capabilities, mobility analysis capabilities, load analysis capabilities, fault analysis capabilities, network slice throughput analysis capabilities, slice load analysis capabilities, network slice traffic prediction analysis capabilities, service experience analysis capabilities, energy efficiency analysis capabilities, or energy consumption analysis capabilities.

[0164] It should be noted that the AI / ML capabilities here can also be capabilities defined by the manufacturer, and this application does not limit them here.

[0165] It should also be noted that the AI / ML capabilities here can be understood as the inference capabilities and / or simulation capabilities possessed in the inference types of AI / ML. That is, the second communication device can train an ML model based on the AI / ML capabilities and then apply it to the inference types of AI / ML to complete the AI / ML inference functions corresponding to the inference types of AI / ML. Taking the MRO in SON as an example of the inference type of AI / ML, as an AI / ML inference function, MRO can use the mobility analysis ML model trained with the mobility analysis capabilities to perform inferences, obtain inference results, and then use these inference results to perform the MRO function. Similarly, the AI / ML capabilities of other inference types can refer to the above description and will not be repeated here.

[0166] It should also be noted that in the embodiments of the present application, the ability of AI / ML inference simulation can be understood as performing simulation on the inference type of AI / ML or on the inference ability of AI / ML. The simulation of the inference ability of AI / ML is to run the ML model for inference in a simulation environment. The ML model is the ML model corresponding to the AI / ML ability, or it can also be understood that the ML model is the ML model obtained by training the AI / ML ability, so that before applying the ML model to the target inference type and further to the live network, its analysis or optimization performance in the simulation environment can be evaluated. The simulation of the inference type of AI / ML is to run the inference type in a simulation environment, so that before applying the inference type to the live network, its analysis or optimization performance in the simulation environment can be evaluated. Taking the inference type of AI / ML in SON as MRO as an example, MRO, as an inference function of AI / ML, can use the mobility analysis ML model trained by the mobility analysis ability to perform inference simulation and obtain the inference simulation result. Optionally, continue to use the inference simulation result to perform the simulation of the MRO function. Similarly, the ability of AI / ML inference simulation for other inference types can refer to the above description and will not be repeated here.

[0167] The following is an explanation and description of the various AI / ML capabilities shown above.

[0168] - The traffic analysis ability can be embodied as an ML model with a traffic analysis function, or it can also be understood as being able to perform traffic analysis inference, that is, using the traffic analysis ML model to perform inference. The traffic analysis function ML model can analyze the traffic metrics. Among them, the analysis of traffic metrics can include the identification of traffic problems (such as traffic congestion, etc.), the statistics of traffic-related metrics, or the prediction of traffic-related metrics, etc.

[0169] - The coverage analysis ability can be embodied as an ML model with a coverage analysis function, or it can also be understood as being able to perform coverage analysis inference, that is, using the coverage analysis ML model to perform inference. The coverage analysis function ML model can analyze the coverage metrics. Among them, the analysis of coverage metrics can include the identification of coverage problems (such as weak coverage, over-coverage, coverage holes, cross-zone coverage, etc.), the statistics of coverage-related metrics, or the prediction of coverage-related metrics, etc.

[0170] - The mobility analysis capability can be embodied as an ML model with a mobility analysis function, or it can be understood as the ability to perform mobility analysis reasoning, that is, using the ML model for mobility analysis to perform reasoning. The ML model with the mobility analysis function can analyze mobility metrics. Among them, the analysis of mobility metrics can include the identification of mobility problems (such as premature handover, late handover, handover to the wrong cell, etc.), the statistics of mobility-related metrics, or the prediction of mobility-related metrics, etc.

[0171] - The load analysis capability can be embodied as an ML model with a load analysis function, or it can be understood as the ability to perform load analysis reasoning, that is, using the ML model for load analysis to perform reasoning. The ML model with the load analysis function can analyze load metrics. Among them, the analysis of load metrics can include the identification of load problems (such as high load, low load), the statistics of load-related metrics, or the prediction of load-related metrics, etc.

[0172] - The fault analysis capability can be embodied as an ML model with a fault analysis function, or it can be understood as the ability to perform fault analysis reasoning, that is, using the ML model for fault analysis to perform reasoning. The ML model with the fault analysis function can analyze fault metrics. Among them, the analysis of fault metrics can include the identification of fault problems (such as network alarms), the statistics of fault-related metrics, or the prediction of fault-related metrics, etc.

[0173] - The network slice throughput analysis capability can be embodied as an ML model with a network slice throughput analysis function, or it can be understood as the ability to perform network slice throughput analysis reasoning, that is, using the ML model for network slice throughput analysis to perform reasoning. The ML model with the network slice throughput analysis function can analyze network slice throughput metrics. Among them, the analysis of network slice throughput metrics can include the identification of network slice throughput problems (such as throughput degradation), the statistics of network slice throughput-related metrics, or the prediction of network slice throughput-related metrics, etc.

[0174] - The slice load analysis capability can be embodied as an ML model with a slice load analysis function, or it can be understood as the ability to perform slice load analysis reasoning, that is, using the ML model for slice load analysis to perform reasoning. The ML model with the slice load analysis function can analyze slice load metrics. Among them, the analysis of slice load metrics can include the identification of slice load problems (such as the decline of key performance indicator (KPI)), the statistics of negative slice load-related metrics, or the prediction of slice load-related metrics, etc.

[0175] - The traffic prediction and analysis ability of network slices can be embodied as an ML model with the function of traffic prediction and analysis of network slices, or it can also be understood as the inference for performing traffic prediction and analysis of network slices, that is, using the ML model with the function of traffic prediction and analysis of network slices to perform inference. The ML model with the function of traffic prediction and analysis of network slices can analyze the indicators of network slice traffic prediction. Among them, the analysis of the indicators of network slice traffic prediction can include the identification of problems in network slice traffic prediction, the statistics of indicators related to network slice traffic prediction, or the prediction of indicators related to network slice traffic prediction, etc.

[0176] - The service experience analysis ability can be embodied as an ML model with the function of service experience analysis, or it can also be understood as the inference that can perform service experience analysis, that is, using the ML model with the function of service experience analysis to perform inference. The ML model with the function of service experience analysis can analyze the indicators of service experience. Among them, the analysis of the indicators of service experience can include the identification of problems in service experience, the statistics of indicators related to service experience, or the prediction ability of indicators related to service experience.

[0177] - The energy efficiency analysis ability can be embodied as an ML model with the function of energy efficiency analysis, or it can also be understood as the inference that can perform energy efficiency analysis, that is, using the ML model with the function of energy efficiency analysis to perform inference. The ML model with the function of energy efficiency analysis can analyze the indicators of energy efficiency. Among them, the analysis of energy efficiency indicators can include the identification of energy efficiency problems (such as high energy efficiency, low energy efficiency, etc.), the statistics of indicators related to energy efficiency, or the prediction of indicators related to energy efficiency, etc.

[0178] - The energy consumption analysis ability can be embodied as an ML model with the function of energy consumption analysis, or it can also be understood as the inference that can perform energy consumption analysis, that is, using the ML model with the function of energy consumption analysis to perform inference. The ML model with the function of energy consumption analysis can analyze the indicators of energy consumption. Among them, the analysis of energy consumption indicators can include the identification of energy consumption problems (such as high energy consumption, low energy consumption, etc.), the statistics of indicators related to energy consumption, or the prediction of indicators related to energy consumption, etc.

[0179] In another possible scenario, the AI / ML capabilities involved in the embodiments of this application can be defined as one or more of the inference types of AI / ML included in MDA. The inference types of AI / ML can refer to the inference types (InferenceType) defined in Section 7.5.1 of 3GPP TS 28.105 V18.2.0, such as coverage analysis, mobility analysis, etc., and can refer to the inference types of AI / ML for MDA defined in Section 8.4 of 3GPP TS 28.104 V18.2.0. This application will not elaborate. The AI / ML capabilities involved in the embodiments of this application can also be defined as one or more of the inference types of SON, such as MRO, MLB, and ES, etc., and can refer to the use cases defined in Section 7 of 3GPP TS 28.313 V17.9.0 or the use cases defined in Section 5.1 of TS 28.310 V18.4.05. This application will not elaborate. The AI / ML capabilities involved in the embodiments of this application can also be defined as one or more of the inference types of RAN intelligence, such as MO, MLB, and ES, etc., and can refer to the use cases defined in 3GPP TS 38.300. This application will not elaborate.

[0180] The AI / ML capabilities involved in the embodiments of this application can also be defined as one or more of the inference types of AI / ML included in NWDAF, such as slice load analysis and user data congestion analysis, etc., and can refer to the use cases defined in Chapter 6 of 3GPP TS 23.288 V18.4.0. This application will not elaborate.

[0181] In the embodiments of this application, the AI / ML capabilities can be general, that is, the AI / ML capabilities are available for all use cases, that is, one or more AI / ML capabilities can be shared by all inference types of AI / ML. Or, in the embodiments involved in this application, the AI / ML capabilities can be at the use case granularity, that is, each inference type of AI / ML has its own AI / ML capabilities. The following will be introduced separately.

[0182] Exemplarily, the AI / ML capabilities are general. In one example, the general AI / ML capabilities can be attached to each use case, that is, each use case can use the general AI / ML capabilities to perform inferences. Refer to Figure 3A , taking the use cases of SON as an example for illustration. The AI / ML capabilities and the use cases of SON can be coupled, and the AI / ML capabilities can be attached to the use cases of SON. Each use case of SON can use the AI / ML capabilities to perform inferences.

[0183] In another example, the general AI / ML capabilities can be embedded in each use case, and each use case with AI / ML capabilities can be used as a new use case. Refer to Figure 3B, taking the SON use case as an example. AI / ML capabilities and SON use cases are not coupled, and AI / ML capabilities can be embedded in various use cases as new use cases. Figure 3B As shown in the figure, MRO, MLB and ES are use cases of SON without AI / ML capabilities, while mobility optimization (MO), load balancing (LB) and network energy saving (NES) are use cases of SON with AI / ML capabilities. Among them, MO can be understood as MRO with AI / ML capabilities, LB can be understood as MLB with AI / ML capabilities, and NES can be understood as ES with AI / ML capabilities.

[0184] It should be noted that Figure 3B The use cases with AI / ML capabilities, such as MO, LB, and NES, are shown only as examples and do not constitute a limitation on the names of the use cases with AI / ML capabilities.

[0185] Exemplarily, general AI / ML capabilities may also be attached to MDA use cases, and MDA use cases may be deployed in domain management functional units.

[0186] In one example, AI / ML capabilities are use case granular. In one example, general AI / ML capabilities can be attached to each use case. Figure 4A , taking SON use case as an example. AI / ML capabilities and SON use cases can be coupled, and AI / ML capabilities can be attached to SON use cases. AI / ML capabilities are case specific.

[0187] Exemplarily, in the use case of SON, MRO may support AI / ML capabilities, and the AI / ML capabilities of MRO may include one or more of the following: traffic analysis capability or mobility analysis capability.

[0188] Exemplarily, in the use case of SON, MLB may support AI / ML capabilities, and the AI / ML capabilities of MLB may include load analysis capabilities.

[0189] Exemplarily, in the use case of SON, ES may support AI / ML capabilities, and the AI / ML capabilities of ES may include one or more of the following: energy efficiency analysis capability, energy consumption analysis capability, or traffic analysis capability.

[0190] The following describes how to deploy AI / ML capabilities attached to each use case.

[0191] See also Figure 4B, taking the distributed MRO (DMRO) use case as an example for illustration. Figure 4B Shows the locations of the training function, inference function, and DMRO function. Figure 4B In it, the cross-domain management function system is a service consumer of the domain management function system. The domain management function unit is illustrated taking the RAN domain management function system as an example. Among them, the RAN domain management function system can include gNB and the RAN domain management function. The AI / ML training function can be deployed in the RAN domain management function, the DRMO function and the AI / ML inference function can be deployed within the gNB, and the AI / ML capability can be deployed within the AI / ML inference function. The MnS interface between the cross-domain management function system and the domain management function system is the interface to be standardized, which is called the management service interface. Similarly, for other optimization functions supported on the gNB, such as DMLB, DES, etc., similar deployments can be made.

[0192] Refer to Figure 4C , taking the MDAF use case as an example for illustration. Figure 4C Shows the locations of the training function, inference function, and MDA function. Among them, the topmost box in the figure is the cross-domain management system (RAN / CN domain MnS consumer / Crossdomain management), that is Figure 4C In it, the cross-domain management function system is a service consumer of the domain management function system. The domain management function unit is illustrated taking the RAN domain management function system as an example. Among them, the RAN domain management function system can include gNB and the RAN domain management function. The AI / ML training function, the AI / ML inference function, and the MDA function are all deployed in the RAN domain management function, and the AI / ML capability is deployed within the AI / ML inference function. The MnS interface between the cross-domain management function system and the domain management function system is the interface to be standardized, which is called the management service interface. Similarly, for other MDA functions, NWDAF functions, etc. supported in the RAN domain management function, similar deployments can be made.

[0193] In another example, the AI / ML capability is embedded in each use case. Refer to Figure 5A , taking the SON use case as an example for illustration. The AI / ML capability and the SON use case are not coupled, and the AI / ML capability is embedded in each use case. As Figure 5AAs shown, MRO, MLB, and ES are use cases of SON without AI / ML capabilities. Mobility optimization (MO), load balancing (LB), and network energy saving (NES) have AI / ML capabilities and can be used as new use cases or functions, coexisting with MRO, MLB, and ES in related technologies. Among them, MO can be understood as MRO with AI / ML capabilities, LB can be understood as MLB with AI / ML capabilities, and NES can be understood as ES with AI / ML capabilities. Figure 5A Among them, the AI / ML capabilities are case specific.

[0194] It should be noted that Figure 5A The use cases with AI / ML capabilities in it, such as MO, LB, and NES, are only shown as examples and do not constitute a limitation on the names of the use cases with AI / ML capabilities.

[0195] Next, the deployment methods of the AI / ML capabilities embedded in each use case will be described.

[0196] Refer to Figure 5B and take MRO as an example for description. Figure 5B The difference between the deployment method in Figure 4B and the deployment method in

[0197] Refer to Figure 5C and take the use case of MDA as an example for description. Figure 5C The deployment method in Figure 4CThe difference in the deployment method lies in that the AIML capability is embedded in the MDA, that is, the AI / ML MDA capability (AIMLMDACapability) is a new and independent capability, or can be called a function, coexisting with the MDA function in related technologies. The AI / ML MDA capability (AIMLMDACapability) here can be one or more of the MDA types, and this application does not make specific limitations. For example, the AI / ML MDA capability (AIMLMDACapability) can be the AI / ML MDA coverage analysis capability (AIMLMDACoverageAnalysisCapabiltiy), and the MDA type can refer to the type definition in 3GPP TS28.104, which will not be elaborated in this application.

[0198] Next, a method for controlling the machine model inference capability provided by the embodiments of this application will be introduced. In the embodiments of this application, it can be applicable to scenarios where both the training function and the inference function are deployed on the domain management function unit, such as MDA, or it can also be applicable to scenarios where the training function is deployed on the domain management function unit and the inference function is deployed on the base station, such as SON or RAN intelligence; it can also be applicable to scenarios where both the training function and the inference function are deployed on the base station, such as SON, RAN intelligence; it can also be applicable to scenarios where the training is on the cross-domain management function unit and the inference function is on the management function unit or the base station. The function management scenarios defined in 3GPP TS28.105 V18.2.0 4a.2 can be referred to, and this application will not be elaborated. In the above scenarios, the cross-domain management function unit can act as a service consumer of the domain management function unit to manage and control the functions in the domain management function unit.

[0199] In the embodiments of this application, the first communication device can be an AI / ML MnS consumer, or a component (such as a chip or a chip system, etc.) in the AI / ML MnS consumer. The second communication device can be an AI / ML MnS producer, or a component (such as a chip or a chip system, etc.) in the AI / ML MnS producer. The first communication device and the communication device can be deployed in different entities, or they can also be deployed in the same entity, as Figure 1 shown. For the convenience of understanding the embodiments of this application, in the following, an example where the first communication device and the second communication device are deployed in different entities will be used. It can be understood that in the embodiments of this application, the role of providing the management service (MnS) of the AI / ML capability is called the producer, and the role of invoking the AI / ML capability management service is called the consumer.

[0200] Among them, if the first communication device is Figure 2 the cross-domain management function unit in Figure 2 , then the second communication device can be Figure 2 the domain management function unit in

[0201] For the descriptions of the cross-domain management function unit and the domain management function unit, please refer to Figure 6 , which is an exemplary flowchart of a method for controlling the machine model inference ability provided by an embodiment of this application, and may include the following operations.

[0202] S601: The cross-domain management function unit sends second information to the domain management function unit.

[0203] S601 is an optional step, Figure 6 and is shown by a dashed line in

[0204] Correspondingly, the domain management function unit receives the second information from the cross-domain management function unit.

[0205] Among them, the second information is used to request to obtain the AI / ML capability information of the domain management function unit. It can be understood that the AI / ML capability information of the domain management function unit can be understood as the information of the AI / ML capabilities that the domain management function unit can support. In some embodiments, the AI / ML capability information of the domain management function unit can also be understood as the types of AI / ML capabilities that the domain management function unit can support.

[0206] In a possible case, the second information can request or indicate the domain management function unit to send the AI / ML capability information. It can be understood that in this case, the second information can be used to request the AI / ML capability information supported by all AI / ML inference types.

[0207] In another possible case, the second information can request or indicate the domain management function unit to send the AI / ML capability information supported by the first AI / ML inference type. It can be understood that the first AI / ML inference type is included in the second information.

[0208] It should be noted that the first AI / ML inference type includes at least one of the following: the inference type of MDA, the inference type of SON, the inference type of NWDAF, or the inference type of radio access network intelligence (RAN intelligence). The specific definition of the AI / ML inference type here can refer to the inference type (InferenceType) defined in Section 7.5.1 of 3GPP TS28.105 V18.2.0.

[0209] In some embodiments, the inference types of MDA may include at least one of types such as energy-saving analysis, coverage problem analysis, fault analysis, network slice throughput analysis, slice load analysis, network slice traffic prediction analysis, service experience analysis, congested traffic analysis, mobile performance analysis, or handover optimization analysis, etc. The specific types can be referred to those defined in Section 8.4 of TS28.104, which will not be elaborated here.

[0210] In some embodiments, the inference types of SON may include at least one of the types of MRO, DMRO, ES, DES, MLB, DMLB. The specific types can be referred to those defined in Chapter 7 of TS28.313, which will not be elaborated here.

[0211] In some embodiments, the inference types of NWDAF may include at least one of types such as slice load analysis, network performance analysis, user data congestion analysis, etc. The specific types can be referred to the use cases defined in Sections 6.5 to 6.21 of TS23.288, which will not be elaborated here.

[0212] In some embodiments, the inference types of RAN intelligence include at least one of the types of MO, NES, LB. The specific types can be referred to those defined in the "Support of AI / ML for NG-RAN" section of TS 38.300 or those defined in Chapter 5 of TR 37.817, which will not be elaborated here.

[0213] For example, the second information may request or instruct the domain management function unit to send AI / ML capability information supported by the inference types of SON (such as MRO (MO), MLB (LB), or ES (NES)). Another example, the second information may request or instruct the domain management function unit to send AI / ML capability information supported by the inference types of MDA (such as coverage problem analysis, fault analysis). Another example, the second information may request or instruct the domain management function unit to send AI / ML capability information supported by the inference types of NWDAF.

[0214] S602: The domain management function unit sends AI / ML capability information to the cross-domain management function unit.

[0215] Correspondingly, the cross-domain management function unit receives the AI / ML capability information from the domain management function unit.

[0216] It should be noted that the AI / ML capability information is used for the management and control of the one or more AI / ML capabilities. It can be understood that the management and control can be that the cross-domain management functional unit triggers the inference or simulation of the AI / ML capabilities of the domain management functional unit based on the received AI / ML capability information, or triggers the inference or simulation of the inference type of AI / ML of the domain management functional unit. It should also be noted that the inference of the AI / ML capabilities here can be understood as training an ML model based on the AI / ML capabilities and then applying it to the inference type of AI / ML to complete the AI / ML inference function corresponding to the inference type of AI / ML.

[0217] The AI / ML capability information can be understood as the information of the AI / ML capabilities supported or possessed by the domain management functional unit. That is to say, in S602, the domain management functional unit can indicate to the cross-domain management functional unit the AI / ML capabilities supported or possessed by the domain management functional unit. Alternatively, the AI / ML capability information can also be understood as the information of the AI / ML capabilities supported or possessed by the AI / ML inference type in the domain management functional unit. It can be understood that the AI / ML capabilities can be common to all AI / ML inference types or at the granularity of each AI / ML inference type. For reference, please refer to the relevant descriptions above and will not be elaborated here. In some embodiments, the AI / ML capability information of the domain management functional unit can also be understood as the type of AI / ML capabilities that the domain management functional unit can support.

[0218] It can be understood that the AI / ML capability information includes AI / ML capabilities. In some embodiments, the AI / ML capabilities include one or more of traffic analysis capabilities, coverage analysis capabilities, mobility analysis capabilities, load analysis capabilities, fault analysis capabilities, network slice throughput analysis capabilities, slice load analysis capabilities, network slice traffic prediction analysis capabilities, service experience analysis capabilities, energy efficiency analysis capabilities, or energy consumption analysis capabilities. Optionally, the AI / ML capabilities can be represented by strings or can also be represented by custom identifiers. This application does not limit this here. It should also be noted that the AI / ML capabilities can also be defined by the manufacturer, and this application does not limit this here. It should also be noted that the AI / ML capabilities can be used as the inference capabilities of AI / ML or as the simulation capabilities of AI / ML.

[0219] Optionally, the AI / ML capability information may also include the inference type of AI / ML. Here, the inference type of AI / ML can also be understood as the applicable AI / ML inference function. In some embodiments, the inference type of AI / ML includes at least one of the following: the inference type of MDA, the inference type of SON, the inference type of NWDAF, or the inference type of radio access network intelligence (RAN intelligence). The specific definition of the inference type of AI / ML here can refer to the inference type (InferenceType) defined in Section 7.5.1 of 3GPP TS 28.105 V18.2.0

[0220] In a possible implementation, if Figure 6 S601 is executed and the second information requests the domain management function unit to send AI / ML capability information, that is, the inference type of AI / ML is not carried in the second information, then it can be considered that the cross-domain management function unit requests the domain management function unit for AI / ML capability information supported by all inference types of AI / ML. In this case, the domain management function unit may send all AI / ML capability information to the cross-domain management function unit, such as including all AI / ML capabilities. The AI / ML capabilities here can refer to the AI / ML capabilities mentioned above and will not be elaborated here. Optionally, in this possible implementation, the domain management function unit may send the inference type of AI / ML applicable to the AI / ML capability information to the cross-domain management function unit, that is, the domain management function unit may indicate to the cross-domain management function unit that the AI / ML capabilities in the sent AI / ML capability information are applicable to the inference type of AI / ML. The inference type of AI / ML here can refer to the inference type of AI / ML mentioned above and will not be elaborated here. Exemplarily, if the AI / ML capabilities sent by the domain management function unit to the cross-domain management function unit include load analysis capabilities, the domain management function unit may send the inference type of AI / ML applicable to the AI / ML capabilities, such as ES (NES), that is, the domain management function unit may indicate to the cross-domain management function unit that the load analysis capabilities are applicable to ES (NES).

[0221] In another possible implementation, if Figure 6 S601 is executed and the second information requests the domain management function unit to send AI / ML capability information of the first AI / ML inference type, that is, the first AI / ML inference type is carried in the second information, then the domain management function unit may send the AI / ML capability information of the first AI / ML inference type to the cross-domain management function unit. It can be understood that the first AI / ML inference type can refer to the implementation in S601 and will not be elaborated here.

[0222] In yet another possible implementation, ifFigure 6 If S601 is not executed, then in S602, the domain management functional unit can send all the capability information to the cross-domain management functional unit, such as including all AI / ML capabilities. The AI / ML capabilities here can refer to the AI / ML capabilities mentioned above and will not be elaborated here. Optionally, in this possible implementation, the domain management functional unit can send the AI / ML inference types applicable to the AI / ML capability information to the cross-domain management functional unit. That is to say, the domain management functional unit can indicate to the cross-domain management functional unit that the AI / ML capabilities in the sent AI / ML capability information can be used for the AI / ML inference types. The AI / ML inference types here can refer to the AI / ML inference types mentioned above and will not be elaborated here. Exemplarily, if the domain management functional unit sends AI / ML capabilities including the load analysis capability to the cross-domain management functional unit, the domain management functional unit can send the AI / ML inference types applicable to the AI / ML capability, such as ES (NES). That is to say, the domain management functional unit can indicate to the cross-domain management functional unit that the load analysis capability can be used for ES (NES).

[0223] In one example, Figure 6 The request and reporting of the above-mentioned capability information shown in can be implemented through the existing information object class (IOC), such as including: DMRO function (DMROFunction), DLBO function (DLBOFunction), DES function (DESManagementFunction), AI / ML inference capability (AiMlInferenceCapability), AI / ML capability of ES (AiMlESCapability), AI / ML capability of MO (AiMlMOCapability), AI / ML capability of LB (AiMlLBCapability), MDA function (MDAFunction), MDS request (MDARequest), AnLF function (AnLFFunction), etc., or can also be implemented through a newly defined information object class (such as SONFunction), or the capability information can be implemented through a newly defined data type (such as aIMLManagementInfomation). This application does not make a limitation here.

[0224] Exemplarily, the attributes of the above several information object classes are described below respectively:

[0225] Table 1: A kind of DMRO function (DMROFunction< <ioc>Examples of the attributes of (>)

[0226]

[0227] The relevant attributes of the DMRO function are shown in Table 1.

[0228] Exemplarily, in S602, the domain management functional unit may send the AI / ML capability information of DMRO to the cross-domain management functional unit through the attributes shown in Table 1. Exemplarily, the domain management functional unit may indicate the supported AI / ML inference capabilities of DMRO to the cross-domain management functional unit through the attribute of'supportedMLInferenceCapabilityList'. Exemplarily, the domain management functional unit may indicate the supported AI / ML emulation capabilities of DMRO to the cross-domain management functional unit through the attribute of'supportedMLEmulationCapabilityList'.

[0229] Similarly, the relevant attributes of the DLBO function (DLBOFunction) and (DESManagementFunction) can also be implemented with reference to Table 1, and will not be described in detail in this application.

[0230] Through Table 1, the domain management functional unit can indicate the AI / ML capability information of one or more inference types to the cross-domain management functional unit.

[0231] It should be noted that Table 1 shows the situation where the AI / ML inference capabilities and the AI / ML emulation capabilities coexist. In fact, the AI / ML inference capabilities and the AI / ML emulation capabilities may not coexist at the same time, that is, the DMRO function may only support the AI / ML inference capabilities or only support the AI / ML emulation capabilities. Similarly, in the following tables, the AI / ML inference capabilities and the AI / ML emulation capabilities may coexist or may not coexist at the same time, and will not be repeated.

[0232] Table 2: An AI / ML inference function of a SON (SONFunction< <ioc>Examples of the attributes of (>)

[0233]

[0234] Table 2 shows the relevant attributes of SON.

[0235] Exemplarily, in S602, the domain management functional unit can indicate the AI / ML inference capabilities supported by SON to the cross-domain management functional unit through the attribute'supportedAIMLInferenceCapabilityList'. Exemplarily, the domain management functional unit can indicate the AI / ML emulation capabilities supported by SON to the cross-domain management functional unit through the attribute'supportedAIMLEmulationCapabilityList'.

[0236] Similarly, the relevant attributes of the MDA function and the AnLF function (AnLFFunction) can be implemented with reference to Table 2, and the present application will not expand the description.

[0237] Through Table 2, the domain management functional unit can indicate the AI / ML capability information of one or more management types to the cross-domain management functional unit.

[0238] Table 3: An AI / ML inference function (AiMlInferenceCapability< <ioc>Examples of related attributes of (>)

[0239]

[0240] Table 3 shows the related attributes of the AI / ML inference function (AiMlInferenceCapability).

[0241] Exemplarily, in S602, the domain management functional unit can indicate the AI / ML inference capabilities supported by the domain management functional unit to the cross-domain management functional unit through the attribute of'supported ML Inference Capability List'. Exemplarily, the domain management functional unit can indicate the AI / ML emulation capabilities supported by the domain management functional unit to the cross-domain management functional unit through the attribute of'supported ML Emulation Capability List'.

[0242] Through Table 3, the domain management functional unit can indicate all the AI / ML capability information of the domain management functional unit to the cross-domain management functional unit.

[0243] Table 4: AI / ML Inference Function of a MO (AiMlMOCapability< <ioc>Examples of relevant attributes of (>)

[0244]

[0245] The relevant attributes of the MO function are shown in Table 4.

[0246] Exemplarily, in S602, the domain management functional unit can indicate the AI / ML inference capabilities supported by the MO to the cross-domain management functional unit through the attribute'supportedAMLInferenceCapabilityList'. Exemplarily, the domain management functional unit can indicate the AI / ML emulation capabilities supported by the MO to the cross-domain management functional unit through the attribute'supportedMLEmulationCapabilityList'.

[0247] Through Table 4, the domain management functional unit can indicate to the cross-domain management functional unit the AI / ML capability information of one or more inference types of the domain management functional unit.

[0248] The supported AI / ML inference capabilities (supportedMLInferenceCapabilityList) shown in Tables 1 to 4 above can be indicated by the relevant attributes in Table 5 below.

[0249] Table 5: Examples of relevant attributes of a supported AI / ML inference capability (SupportedMLCapabilityList<<data type>>)

[0250]

[0251] The relevant attributes of the supported AI / ML inference capabilities are shown in Table 5. That is, the supported AI / ML inference capabilities shown in Tables 1 to 4 can be indicated by the AI / ML capability type and the AI / ML inference type in Table 5.

[0252] Similarly, the supported AI / ML emulation capabilities (supportedMLEmulationCapabilityList)' shown in Tables 1 to 4 can be indicated by the relevant attributes shown in Table 6.

[0253] Table 6: A supported AI / ML emulation capability (SupportedMLEmulationCapabilityList< <datatype>Examples of related attributes of (>)

[0254]

[0255] Table 6 shows the related attributes of the supported AI / ML simulation capabilities. That is, the supported AI / ML simulation capabilities shown in Tables 1 to 4 can be indicated by the AI / ML capability types and AI / ML inference types in Table 5.

[0256] Based on Figure 6 In the shown solution, the cross-domain management functional unit can obtain the AI / ML capability information of the domain management functional unit, so as to control the AI / ML capabilities of the domain management functional unit. In a possible case, the cross-domain management functional unit can obtain the AI / ML capability information of one or more inference types, so as to control the AI / ML capabilities of one or more inference types.

[0257] Next, in combination with Figure 7 This application embodiment will introduce a method for the cross-domain management functional unit to control AI / ML capabilities. Refer to Figure 7 , which is an exemplary flowchart of a method for controlling the machine model inference capability provided by this application embodiment, and may include the following operations.

[0258] S701: The cross-domain management functional unit sends the first information to the domain management functional unit.

[0259] Correspondingly, the domain management functional unit receives the first information from the cross-domain management functional unit.

[0260] Optionally, the cross-domain management functional unit may send the first information to the domain management functional unit according to the AI / ML capability information of the domain management functional unit in S602.

[0261] Among them, the first information may indicate the AI / ML capability corresponding to the AI / ML inference. In some embodiments, the AI / ML capability corresponding to the AI / ML inference may be understood as the AI / ML capability type corresponding to the AI / ML inference. That is to say, the first information may indicate the AI / ML capability type corresponding to the AI / ML inference. The domain management functional unit may perform the AI / ML inference corresponding to the AI / ML capability type to obtain the inference result of the AI / ML capability. In a possible case, the first information may be carried in an inference request or a simulation request, or carried in an existing SON functional object class, MDArequest object class, NWDAF information object class, or AIML inference functional object class, or a newly defined object class, and this application is not limited thereto.

[0262] In some embodiments, the first information indicates AI / ML capabilities, which may include one or more of traffic analysis capabilities, coverage analysis capabilities, mobility analysis capabilities, load analysis capabilities, fault analysis capabilities, network slice throughput analysis capabilities, slice load analysis capabilities, network slice traffic prediction analysis capabilities, service experience analysis capabilities, congestion analysis capabilities, energy efficiency analysis capabilities, or energy consumption analysis capabilities.

[0263] Optionally, in Figure 7 the embodiments shown, the cross-domain management function unit may also send the AI / ML inference type of the above AI / ML capabilities to the domain management function unit. It can be understood that the AI / ML inference type may be indicated in the same message as the AI / ML capabilities, or may be indicated in different messages. This application does not make specific limitations. Exemplarily, the first information may include the AI / ML capabilities and the AI / ML inference type. In some embodiments, the AI / ML inference type may include at least one of the following: the inference type of MDA, the inference type of SON, the inference type of NWDAF, or the inference type of radio access network intelligence (RAN intelligence). The specific definition of the AI / ML inference type here may refer to the inference type (InferenceType) defined in section 7.5.1 of 3GPP TS 28.105 V18.2.0, which will not be elaborated in this application.

[0264] In some embodiments, the first information includes the AI / ML inference type and the AI / ML capabilities available for the AI / ML inference type. Then the domain management function unit may execute the function corresponding to the AI / ML inference type and may perform AI / ML inference through the available AI / ML capabilities to obtain an inference result. It can be understood that the inference result can be used to execute the function corresponding to the AI / ML inference type.

[0265] In one example, Figure 7 Indications of the AI / ML capabilities shown in the figure can be implemented through existing IOCs, such as including: DMRO function (DMROFunction), DLBO function (DLBOFunction), DES function (DESManagementFunction), AI / ML inference capability (AiMlInferenceCapability), AI / ML capability of ES (AiMlESCapability), AI / ML capability of MO (AiMlMOCapability), AI / ML capability of LB (AiMlLBCapability), MDA function (MDAFunction), MDS request (MDARequest), AnLF function (AnLFFunction), etc., or can also be implemented through newly defined object classes, which are not limited in this application. By way of example, the attributes of the above several object classes are described separately below:

[0266] Table 7: A DMRO function (DMROFunction< <ioc>Examples of the attributes of (>)

[0267]

[0268]

[0269] The relevant attributes of the DMRO function are shown in Table 7.

[0270] Exemplarily, in S701, the cross-domain management functional unit can indicate the AI / ML capabilities of the DMRO to the domain management functional unit through some of the attributes shown in Table 7. Exemplarily, the first information can carry the attribute of 'available MLinferenceCapabilityList' and send the AI / ML inference capabilities of the DMRO to the domain management functional unit, indicating that the domain management functional unit performs AI / ML inference on the AI / ML inference capabilities of the DMRO. Exemplarily, the first information can carry the attribute of 'available MLEmulationCapabilityList' and send the AI / ML emulation capabilities of the DMRO to the domain management functional unit, indicating that the domain management functional unit performs AI / ML inference emulation on the AI / ML emulation capabilities of the DMRO.

[0271] Similarly, the relevant attributes of the DLBO function (DLBOFunction) and the DES function (DESManagementFunction) can also be implemented with reference to Table 7, and the present application will not expand the description.

[0272] Through Table 7, the cross-domain management functional unit can indicate the AI / ML capabilities of one or more inference types to the domain management functional unit, indicating that the domain management functional unit performs AI / ML inference on the AI / ML capabilities of one or more inference types.

[0273] It should be noted that Table 1 and Table 7 can be implemented as one table or as different tables respectively, and the present application does not make specific limitations.

[0274] Table 8: A SON function (SONFunction< <ioc>Examples of attributes indicating AI / ML capabilities of (>)

[0275]

[0276] Table 8 shows the relevant attributes of SON.

[0277] Exemplarily, in S701, the cross-domain management functional unit can indicate the AI / ML inference capabilities of SON to the domain management functional unit through the attribute 'availableMLInferenceCapabilityList', and instruct the domain management functional unit to perform AI / ML inferences of the AI / ML inference capabilities of SON. Exemplarily, the cross-domain management functional unit can indicate the AI / ML emulation capabilities of SON to the domain management functional unit through the attribute 'availableMLEmulationCapabilityList', and instruct the domain management functional unit to perform AI / ML inference emulations of the AI / ML emulation capabilities of SON.

[0278] Similarly, the relevant attributes of the MDA function and the AnLF function (AnLFFunction) can be implemented with reference to Table 8, and are not elaborated in this application.

[0279] Through Table 8, the cross-domain management functional unit can indicate the AI / ML capabilities of one or more inference types to the domain management functional unit, and instruct the domain management functional unit to perform AI / ML inferences of the AI / ML capabilities of one or more inference types. It should be noted that Table 2 and Table 8 can be implemented as one table or as different tables respectively, and this application does not make specific limitations.

[0280] Table 9: An AI / ML inference function (AiMlInferenceCapability< <ioc>Examples of related attributes of (>)

[0281]

[0282]

[0283] Table 9 shows the related attributes of the AI / ML inference capability (AiMlInferenceCapability).

[0284] Exemplarily, in S701, the cross-domain management functional unit can indicate the AI / ML inference capability to the domain management functional unit through the attribute 'available ML Inference Capability List', and instruct the domain management functional unit to perform the AI / ML inference of the AI / ML inference capability. Exemplarily, the cross-domain management functional unit can indicate the AI / ML emulation capability to the domain management functional unit through the attribute 'available ML Emulation Capability List', and instruct the domain management functional unit to perform the AI / ML inference emulation of the AI / ML emulation capability. That is to say, the cross-domain management functional unit can, through these two attributes, instruct the domain management functional unit to perform the AI / ML inference of the AI / ML capability, rather than being limited to the AI / ML inference of the AI / ML capability of a certain inference type.

[0285] Through Table 9, the cross-domain management functional unit can indicate the AI / ML capability to the domain management functional unit, and instruct the domain management functional unit to perform the AI / ML inference of the AI / ML capability. In a possible case, through Table 9, the cross-domain management functional unit can also indicate the AI / ML capabilities of one or more inference types to the domain management functional unit, and instruct the domain management functional unit to perform the AI / ML inference of the AI / ML capabilities of one or more inference types.

[0286] It should be noted that Table 3 and Table 9 can be implemented as one table or as different tables respectively, and the present application does not make specific limitations.

[0287] Table 10: An MO function (AiMlMOCapability< <ioc>Examples of relevant attributes of the AI / ML capabilities of (>)

[0288]

[0289] Table 10 shows the relevant attributes of the MO function.

[0290] Exemplarily, in S701, the cross-domain management function unit can indicate the AI / ML inference capability of the MO to the domain management function unit through the attribute 'availableMLInferenceCapabilityList' of the MO, and instruct the domain management function unit to perform the AI / ML inference of the AI / ML inference capability of the MO. Exemplarily, the cross-domain management function unit can indicate the AI / ML emulation capability of the MO to the domain management function unit through the attribute 'availableMLEmulationCapabilityList' of the MO, and instruct the domain management function unit to perform the AI / ML inference emulation of the AI / ML emulation capability of the MO.

[0291] Through Table 10, the cross-domain management function unit can indicate the AI / ML capabilities of one or more inference types to the domain management function unit, and instruct the domain management function unit to perform the AI / ML inference of the AI / ML capabilities of one or more inference types. It should be noted that Table 4 and Table 10 can be implemented as one table or as different tables respectively, and the present application does not make specific limitations.

[0292] The available ML inference capabilities (availableMLinferenceCapabilityList) shown in Tables 7 to 10 above can be indicated by the relevant attributes in Table 11 below.

[0293] Table 11: Examples of relevant attributes of an available ML inference capability (availableMLinferenceCapabilityList<<data type>>)

[0294]

[0295]

[0296] Table 11 shows the relevant attributes of the available ML inference capabilities. That is to say, the available ML inference capabilities shown in Tables 7 to 10 can be indicated by the AI / ML capability type and the AI / ML inference type in Table 5.

[0297] Similarly, the available AI / ML emulation capabilities (availableMLEmulationCapabilityList) shown in Tables 7 to 10 can be indicated by the relevant attributes shown in Table 12.

[0298] Table 12: Examples of relevant attributes of an available AI / ML emulation capability (availableMLEmulationCapabilityList<<data type>>)

[0299]

[0300] The relevant attributes of the available AI / ML emulation capabilities are shown in Table 12. That is to say, the available AI / ML emulation capabilities shown in Tables 7 to 10 can be indicated by the AI / ML capability type and the AI / ML inference type in Table 12.

[0301] It can be understood that the AI / ML capabilities mentioned in S701 can be general or at the use case granularity, and can be implemented with reference to the previous relevant descriptions, which will not be repeated here.

[0302] S702: The domain management functional unit performs AI / ML inference on the AI / ML capabilities.

[0303] In S702, the domain management functional unit performs AI / ML inference on the AI / ML capabilities to obtain the inference results of the AI / ML capabilities. It can be understood that the domain management functional unit performing AI / ML inference on the AI / ML capabilities can be understood as the domain management functional unit performing AI / ML inference on the AI / ML capability type to obtain the inference results of the AI / ML capability type. For example, the first information can indicate one or more of the traffic analysis capability, coverage analysis capability, mobility analysis capability, load analysis capability, fault analysis capability, network slice throughput analysis capability, slice load analysis capability, network slice traffic prediction analysis capability, service experience analysis capability, energy efficiency analysis capability, or energy consumption analysis capability mentioned above, then the domain management functional unit can perform AI / ML inference on one or more of the traffic analysis capability, coverage analysis capability, mobility analysis capability, load analysis capability, fault analysis capability, network slice throughput analysis capability, slice load analysis capability, network slice traffic prediction analysis capability, service experience analysis capability, energy efficiency analysis capability, or energy consumption analysis capability indicated by the first information.

[0304] Optionally, the cross-domain management functional unit in S701 may also send the inference type of AI / ML corresponding to the AI / ML capability to the domain management functional unit. For example, if the first piece of information includes the AI / ML capability and the inference type of AI / ML, then the domain management functional unit may execute the AI / ML capability of the inference type of AI / ML. The inference type of AI / ML here includes at least one of the following: the inference type of MDA, the inference type of SON, the inference type of NWDAF, or the inference type of radio access network intelligence (RAN intelligence).

[0305] Exemplarily, the first piece of information indicates the load analysis capability. The cross-domain management functional unit sends the inference type of AI / ML corresponding to the load analysis capability to the domain management functional unit. For example, if the inference type of AI / ML is ES (NES), then the domain management functional unit may execute the AI / ML inference of the load analysis capability of ES (NES).

[0306] In some embodiments, the cross-domain management functional unit may trigger the activation of the function corresponding to the inference type of AI / ML. For example, the cross-domain management functional unit may trigger the domain management functional unit to activate or execute at least one function corresponding to the inference type of MDA, the inference type of SON, the inference type of NWDAF, or the inference type of radio access network intelligence (RAN intelligence).

[0307] It should be noted that the step of the cross-domain management functional unit triggering the activation of the function corresponding to the inference type of AI / ML may be executed before S701, may be executed after S701 and before S702, or may be executed together with S701. This application does not make a specific limitation. Then, in S702, the domain management functional unit may activate the function corresponding to the inference type of AI / ML and execute the AI / ML inference of the AI / ML capability indicated by the first piece of information, so as to obtain an inference result.

[0308] S703: The domain management functional unit sends the inference result of the AI / ML inference to the cross-domain management functional unit.

[0309] Correspondingly, the cross-domain management functional unit receives the inference result from the domain management functional unit. For example, the cross-domain management functional unit may receive the inference result from the domain management functional unit.

[0310] In a possible case, the inference result may be carried in an inference report. For example, the domain management functional unit may send an inference report of the AI / ML inference to the cross-domain management functional unit, and the inference result may be carried in the inference report.

[0311] Optionally, the domain management function unit may also send one or more of the inference status (such as one of success, in progress, or failure), and the inference time (such as one or more of the start time, end time, or duration) to the cross-domain management function unit. It can be understood that one or more of the inference status or the inference time may be sent in the same message as the inference result, such as being sent in the inference report. Alternatively, one or more of the inference status or the inference time may be sent in a different message from the inference result, and the present application does not make specific limitations.

[0312] In the embodiments of the present application, the inference result may be the inference result of the AI / ML capabilities indicated by the first information.

[0313] Exemplarily, for example, if the first information indicates the AI / ML capabilities of MRO (MO), then the inference result may include one or more of the cell identifier, base station identifier, cell individual offset (CIO), time trigger, or handover trigger.

[0314] Exemplarily, if the first information indicates the AI / ML capabilities of MLB (LB), then the inference result may include one or more of the cell identifier, base station identifier, or CIO.

[0315] Exemplarily, if the first information indicates the AI / ML capabilities of the section ES (NES), then the inference result may include one or more of the cell identifier, base station identifier, or information on the shut-down cell, or information on the shut-down carrier, or information on the shut-down time slot.

[0316] The inference result of the inference type of MDA may refer to one or more of the corresponding outputs in section 8.4 of TS28.104, and the present application does not expand the description.

[0317] In one example, the domain management function unit may also send the inference type of AI / ML applicable to the inference result to the cross-domain management function unit. The inference type of AI / ML includes at least one of the following: the inference type of MDA, the inference type of SON, the inference type of NWDAF, or the inference type of radio access network intelligence (RAN intelligence). It can be understood that the inference type of AI / ML can be understood as the inference type of AI / ML capabilities that can use the inference result. It should be noted that the inference type of AI / ML may be carried in the same message as the inference result, such as in the inference report, or may also be carried in a different message from the inference result, and the present application does not make specific limitations.

[0318] Optionally, before S703, the cross-domain management functional unit may send an inference result request to the domain management functional unit, requesting or instructing the domain management functional unit to send an inference result or an inference report.

[0319] Next, the method for controlling the machine model inference ability provided by the embodiments of the present application will be introduced in combination with specific use cases.

[0320] Exemplarily, taking the mobility optimization ability use case as an example, the AI / ML ability may be attached to MRO or distributed MRO. Refer to Figure 8 , which is an exemplary flowchart of a method for controlling the machine model inference ability provided by the embodiments of the present application, and may include the following operations. Figure 8 In the illustrated embodiment, the cross-domain management functional unit is taken as the NMS network management system (network management system, NMS), and the domain management functional unit is taken as the domain MnF for description.

[0321] S801: The NMS turns on the switch of the DMRO.

[0322] In S801, the NMS may trigger the activation of the DMRO, that is, the NMS may trigger the activation of the DMRO function.

[0323] It can be understood that the NMS may send an instruction to the domain MnF to trigger the activation of the DMRO, and the domain MnF may instruct the base station to activate the DMRO.

[0324] S802: The NMS sends the second information to the domain MnF.

[0325] Correspondingly, the domain MnF receives the second information from the NMS. This step is optional.

[0326] Among them, the second information may refer to the implementation of S601. In one possible case, the second information may request the AI / ML ability information of the DMRO of the domain MnF. For example, the second information may be an AI / ML inference ability request for the DMRO.

[0327] S803: The domain MnF sends the AI / ML ability information of the DMRO to the NMS.

[0328] Correspondingly, the NMS receives the AI / ML ability information of the DMRO from the domain MnF.

[0329] In S803, the domain MnF may respond to the second information and send the AI / ML ability information of the DMRO to the NMS.

[0330] In one possible scenario, when the AI / ML capabilities are general, the domain MnF can directly send a capability indication, such as the DMRO supports AIML inference capabilities or the DMRO supports AIML inference simulation capabilities. For example, the domain MnF can send mobility analysis capabilities, traffic analysis capabilities, coverage analysis capabilities, etc. supported by the DMRO.

[0331] In another possible scenario, when the AI / ML capabilities are at the use case granularity, the domain MnF can send the AI / ML capabilities of the DMRO to the NMS, such as coverage analysis capabilities, traffic analysis capabilities, mobile performance analysis capabilities, etc.

[0332] S804: The NMS sends the first information to the domain MnF.

[0333] Correspondingly, the domain MnF receives the first information from the NMS. Among them, the first information can be implemented with reference to S601.

[0334] In S804, the NMS can select which AI / ML capabilities to activate to perform the AI / ML inference of the DMRO based on the AI / ML capability information of the DMRO, and indicate it to the domain MnF through the first information.

[0335] In one possible scenario, if the AI / ML capabilities are deployed in the base station, the domain MnF can send the first information to the base station to instruct the base station to perform the AI / ML inference of the AI / ML capabilities.

[0336] S805: The NMS requests the inference result of the AI / ML inference from the domain MnF.

[0337] For example, the NMS can send an inference result request to the domain MnF, thereby requesting the domain MnF to send the inference result of the AI / ML inference.

[0338] S806: The domain MnF sends the inference result to the NMS.

[0339] Correspondingly, the NMS receives the inference result from the domain MnF.

[0340] In S806, the domain MnF can send the inference result to the NMS in response to the inference result request in S805, such as one or more of the coverage analysis result, traffic analysis result, mobility analysis result, etc. In one possible scenario, if the AI / ML capabilities are deployed in the base station, the domain MnF can obtain the inference result from the base station and send it to the NMS.

[0341] Exemplarily, the domain MnF sends an inference report to the NMS, and the inference report can include the inference time, the thrust state, and the inference result. Optionally, the inference report also includes the inference type of the inference result, that is, the DMRO.

[0342] In Figure 8 the illustrated embodiment, the NMS can open the DMRO and then activate the AI / ML inference. Hereinafter, in combination with Figure 9 introduce the scenario where the NMS first activates the AI / ML inference and then opens the DMRO. Refer to Figure 9 , which is an exemplary flowchart of a method for controlling the inference ability of a machine model provided by an embodiment of the present application, and may include the following operations. Figure 9 In the illustrated embodiment, the cross-domain management function unit is taken as the NMS, and the domain management function unit is taken as the domain MnF for description.

[0343] S901: The NMS sends a second piece of information to the domain MnF.

[0344] Correspondingly, the domain MnF receives the second piece of information from the NMS. Among them, the second piece of information can refer to the implementation of S601.

[0345] For example, the second piece of information can request the AI / ML ability information of the domain MnF. It can be understood that the difference between the second piece of information in S901 and the second piece of information in S801 is that the second piece of information in S901 can request the AI / ML ability information of the domain MnF rather than being limited to the AI / ML ability information of the DMRO.

[0346] This step is optional.

[0347] S902: The domain MnF sends the AI / ML ability information to the NMS.

[0348] Correspondingly, the NMS receives the AI / ML ability information from the domain MnF.

[0349] S902 can refer to the implementation of S602. In S902, the domain MnF can send the information of the supported or possessed AI / ML ability to the NMS.

[0350] S903: The NMS activates the AI / ML ability to perform AI / ML inference.

[0351] For example, the NMS can send an inference request to the domain MnF to perform AI / ML inference.

[0352] S903 can refer to the implementation of S601. In S903, the NMS can activate the AI / ML ability of the DMRO according to the AI / ML ability information obtained in S902 to perform the AI / ML inference of the AI / ML ability.

[0353] In a possible situation, if the AI / ML inference ability is deployed in the base station, the domain MnF can send a first piece of information to the base station to instruct the base station to perform the AI / ML inference of the AI / ML ability.

[0354] S904: NMS turns on the switch of DMRO and indicates the AI / ML capabilities.

[0355] In S904, NMS can, based on the obtained AI / ML capability information, indicate to the domain MnF which AI / ML capabilities of DMRO to use. That is to say, in S904, NMS can indicate to the domain MnF which AI / ML capabilities of DMRO to use.

[0356] S905 - S906 can refer to S805 and S806.

[0357] Based on the concept of the above embodiments, see Figure 10 , an embodiment of the present application provides a communication device 1000, which includes a processing unit 1001 and a transceiver unit 1002. The device 1000 can be a communication device or a device applied to a communication device that can support the communication device to execute the method for notifying quality of service parameters.

[0358] Among them, the transceiver unit can also be referred to as a transceiver module, transceiver, transceiver, transceiver device, etc. The processing unit can also be referred to as a processor, processing board, processing unit, processing device, etc. Optionally, the device for implementing the receiving function in the transceiver unit can be regarded as the receiving unit. It should be understood that the transceiver unit is used to perform the sending operation and receiving operation of the communication device in the above method embodiments. The device for implementing the sending function in the transceiver unit is regarded as the sending unit, that is, the transceiver unit includes a receiving unit and a sending unit.

[0359] In addition, it should be noted that if the device is implemented using a chip / chip circuit, the transceiver unit can be an input / output circuit and / or a communication interface, performing input operations (corresponding to the aforementioned receiving operations) and output operations (corresponding to the aforementioned sending operations); the processing unit is an integrated processor or microprocessor or integrated circuit.

[0360] The following details the implementation manners of applying the device 1000 to the first communication device and the second communication device.

[0361] Exemplarily, when the device 1000 is applied to the first communication device, the operations performed by each unit are described in detail.

[0362] In an optional implementation manner, the communication device 1000 can be applied to the first communication device and execute the method performed by the first communication device, specifically, for example, the method performed by the terminal device in the aforementioned Figures 6 - 9 illustrated embodiment.

[0363] For example, a transceiver unit 1002 is configured to receive first information from a second communication device, where the first information includes AI / ML capabilities corresponding to AI / ML inferences. A processing unit 1001 is configured to perform AI / ML inferences on the AI / ML capabilities to obtain inference results. The transceiver unit 1002 is further configured to send the inference results to the second communication device.

[0364] For another example, a processing unit 1001 is configured to determine supported AI / ML capability information. A transceiver unit 1002 is configured to send the supported AI / ML capability information to the second communication device. The AI / ML capability information indicates one or more AI / ML capabilities, and is used to manage the one or more AI / ML capabilities.

[0365] Exemplarily, when the device 1000 is applied to a second communication device, the operations performed by its respective units are described in detail.

[0366] In an alternative embodiment, the communication device 1000 can be applied to a second communication device to execute the method performed by the second communication device, specifically, for example, the method performed by the second communication device in the foregoing Figures 6 - 9 illustrated embodiment.

[0367] For example, a processing unit 1001 is configured to generate first information, where the first information includes AI / ML capabilities corresponding to AI / ML inferences. A transceiver unit 1002 is configured to send the first information to a first communication device. The transceiver unit 1002 is further configured to receive the inference results of the AI / ML inferences from the first communication device.

[0368] For another example, a transceiver unit 1002 is configured to receive supported AI / ML capability information from a first communication device. The AI / ML capability information indicates one or more AI / ML capabilities, and is used to manage the one or more AI / ML capabilities. A processing unit 1001 is configured to manage the one or more AI / ML capabilities.

[0369] Based on the concept of the embodiment, as Figure 11 illustrated, an embodiment of the present application provides a communication device 1100. The communication device 1100 includes a processor 1110. Optionally, the communication device 1100 may further include a memory 1120, which is configured to store instructions executed by the processor 1110, or input data required for the processor 1110 to run the instructions, or data generated after the processor 1110 runs the instructions. The processor 1110 can implement the method shown in the foregoing method embodiments through the instructions stored in the memory 1120.

[0370] Based on the concept of the embodiment, as Figure 12 As shown, an embodiment of the present application provides a communication device 1200, which may be a chip or a chip system. Optionally, in the embodiment of the present application, the chip system may be composed of chips, or may include chips and other discrete devices.

[0371] The communication device 1200 may include at least one processor 1210, and the processor 1210 is coupled to the memory. Optionally, the memory may be located inside or outside the device. For example, the communication device 1200 may further include at least one memory 1220. The memory 1220 stores the necessary computer programs, configuration information, computer programs or instructions, and / or data in any of the above embodiments; the processor 1210 may execute the computer programs stored in the memory 1220 to complete the methods in any of the above embodiments. Optionally, the memory may also be integrated with the processor.

[0372] The coupling in the embodiment of the present application is an indirect coupling or communication connection between devices, units or modules, which may be electrical, mechanical or other forms, and is used for information interaction between devices, units or modules. The processor 1210 may cooperate with the memory 1220. In the embodiment of the present application, the specific connection medium between the transceiver 1230, the processor 1210 and the memory 1220 is not limited.

[0373] The communication device 1200 may further include a transceiver 1230, and the communication device 1200 may perform information interaction with other devices through the transceiver 1230. The transceiver 1230 may be a circuit, a bus, a transceiver or any other device that can be used for information interaction, or is called a signal transceiver unit. As Figure 12 shown, the transceiver 1230 includes a transmitter 1231, a receiver 1232 and an antenna 1233. In addition, when the communication device 1200 is a chip-type device or circuit, the transceiver in the communication device 1200 may also be an input / output circuit and / or a communication interface, which can input data (or receive data) and output data (or send data), and the processor is an integrated processor or a microprocessor or an integrated circuit, and the processor may determine the output data according to the input data.

[0374] In a possible implementation, the communication device 1200 can be applied to a communication device. Specifically, the communication device 1200 can be a communication device or a device capable of supporting a communication device and implementing the functions of the first communication device or the second communication device in any of the above-mentioned embodiments. The memory 1220 stores the necessary computer programs, computer programs or instructions, and / or data for implementing the functions of the first communication device or the second communication device in any of the above-mentioned embodiments. The processor 1210 can execute the computer programs stored in the memory 1220 to complete the methods executed by the first communication device or the second communication device in any of the above-mentioned embodiments.

[0375] In the embodiments of the present application, the processor can be a general-purpose processor, a digital signal processor, an application-specific integrated circuit, a field-programmable gate array or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, and can implement or execute the various methods, steps and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor can be a microprocessor or any conventional processor, etc. The steps of the method disclosed in combination with the embodiments of the present application can be directly embodied as being executed by a hardware processor, or executed by a combination of hardware and software modules in the processor.

[0376] In the embodiments of the present application, the memory can be a non-volatile memory, such as a hard disk drive (HDD) or a solid-state drive (SSD), etc., or a volatile memory, such as a random-access memory (RAM). The memory can also be any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory in the embodiments of the present application can also be a circuit or any other device capable of implementing a storage function, for storing computer programs, computer programs or instructions, and / or data.

[0377] Based on the above embodiments, refer to Figure 13 , the embodiments of the present application further provide another communication device 1300, including: an input / output interface 1310 and a logic circuit 1320; the input / output interface 1310 is used to receive code instructions and transmit them to the logic circuit 1320; the logic circuit 1320 is used to run the code instructions to execute the methods executed by the first communication device or the second communication device in any of the above-mentioned embodiments.

[0378] Hereinafter, the operations performed by the device 1300 when applied to the first communication device or the second communication device will be described in detail.

[0379] In an alternative embodiment, the communication device 1300 can be applied to the first communication device and execute the method performed by the first communication device, specifically, for example, the method performed by the first communication device in the foregoing Figures 6 - 9 embodiment shown.

[0380] For example, the input / output interface 1310 is used to input the first information from the second communication device, and the first information includes the AI / ML capabilities corresponding to the AI / ML inference. The logic circuit 1320 is used to perform the AI / ML inference of the AI / ML capabilities to obtain an inference result. The input / output interface 1310 is further used to output the inference result.

[0381] For another example, the logic circuit 1320 is used to determine the supported AI / ML capability information. The input / output interface 1310 is used to output the supported AI / ML capability information. Wherein, the AI / ML capability information indicates one or more AI / ML capabilities. The AI / ML capability information is used to manage one or more AI / ML capabilities.

[0382] Since the communication device 1300 provided in this embodiment can be applied to the first communication device and execute the method performed by the first communication device, the technical effects that can be obtained can refer to the foregoing method embodiments and will not be elaborated herein.

[0383] In an alternative embodiment, the communication device 2000 can be applied to the second communication device and execute the method performed by the second communication device, specifically, for example, the method performed by the second communication device in the foregoing Figures 6 - 9 embodiment shown.

[0384] For example, the logic circuit 1320 is used to generate the first information, and the first information includes the AI / ML capabilities corresponding to the AI / ML inference. The input / output interface 1310 is used to output the first information. The input / output interface 1310 is further used to input the inference result of the AI / ML inference from the first communication device.

[0385] For another example, the input / output interface 1310 is used to input the AI / ML capability information supported by the first communication device. Wherein, the AI / ML capability information indicates one or more AI / ML capabilities. The AI / ML capability information is used to manage one or more AI / ML capabilities. The logic circuit 1320 is used to manage one or more AI / ML capabilities.

[0386] Since the communication device 1300 provided in this embodiment can be applied to the second communication device and execute the method performed by the second communication device, the technical effects that can be obtained can refer to the foregoing method embodiments and will not be elaborated herein.

[0387] Based on the above embodiments, an embodiment of the present application further provides a communication system, which includes at least one first core network and at least one second core network. Optionally, the communication system further includes at least one terminal device. The technical effects that can be obtained can refer to the above method embodiments, which will not be elaborated here.

[0388] Based on the above embodiments, an embodiment of the present application further provides a computer-readable storage medium, which stores computer programs or instructions. When the instructions are executed, the methods executed by the communication device in any of the above embodiments are implemented. The computer-readable storage medium may include: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories, random access memories, magnetic disks, or optical discs.

[0389] In order to implement the Figures 9 - 13 functions of the above communication device, an embodiment of the present application further provides a chip, including a processor, which is used to support the communication device to implement the functions involved in the first communication device or the second communication device in the above method embodiments. In a possible design, the chip is connected to a memory or the chip includes a memory, and the memory is used to store the necessary computer programs or instructions and data of the terminal device or the network device.

[0390] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program codes.

[0391] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, and the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer programs or instructions. These computer programs or instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the specified functions in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0392] These computer programs or instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to work in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction device that implements the functions specified in one or more processes and / or blocks Figure 1 in one process or a plurality of processes and / or blocks Figure 1 specified in the block or blocks.

[0393] These computer programs or instructions can also be loaded onto a computer or other programmable data processing apparatus, so as to perform a series of operation steps on the computer or other programmable apparatus to generate a computer-implemented process, thereby the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one or more processes and / or blocks Figure 1 in one process or a plurality of processes and / or blocks Figure 1 specified in the block or blocks.

[0394] Obviously, those skilled in the art can make various changes and modifications to the embodiments of the present application without departing from the scope of the embodiments of the present application. Thus, if these modifications and variations of the embodiments of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these changes and modifications.< / ioc> < / ioc> < / ioc> < / ioc> < / datatype> < / ioc> < / ioc> < / ioc> < / ioc>

Claims

1. A control method for the inference ability of a machine model, characterized in that, Including: Receiving AI / ML capability information supported by a first communication device, the AI / ML capability information indicating one or more AI / ML capabilities; Wherein, the AI / ML capability information is used for the management and control of the one or more AI / ML capabilities.

2. The method according to claim 1, characterized in that, Further including: Receiving the inference type of AI / ML to which the AI / ML capability information applies; the inference type includes one or more of a management data analysis function, a mobility optimization function, a network energy saving function, or a load balancing function.

3. The method according to claim 1 or 2, characterized in that Further including: Sending second information to the first communication device, the second information being used to request the first communication device to send the AI / ML capability information.

4. The method according to claim 3, characterized in that, The second information instructs the first communication device to send AI / ML capability information supported by one or more of a management data analysis function, a mobility optimization function, a network energy saving function, or a load balancing function.

5. The method according to any one of claims 1 to 4, characterized in that, The one or more AI / ML capabilities include one or more of the following: traffic analysis capability, traffic prediction capability, mobility analysis capability, mobility prediction capability, load analysis capability, load prediction capability, energy consumption analysis capability, or energy consumption prediction capability.

6. The method according to any one of claims 1 to 5, characterized in that Further including: Sending first information to the first communication device, the first information including the AI / ML capabilities corresponding to the AI / ML inference; wherein, the one or more AI / ML capabilities include the AI / ML capabilities corresponding to the AI / ML inference; Receiving the inference result of the AI / ML inference from the first communication device.

7. The method according to claim 6, wherein The first information is used to request the first communication device to perform the AI / ML inference of the AI / ML capabilities corresponding to the AI / ML inference.

8. The method according to claim 6 or 7, characterized in that, Further including: Sending the inference type of the AI / ML capabilities, the inference type including one or more of a management data analysis function, a mobility optimization function, a network energy saving function, or a load balancing function.

9. The method according to any one of claims 6 to 8, characterized in that Further including: Receiving one or more of an inference status or an inference time.

10. The method according to any one of claims 6 to 9, characterized in that, The inference result further includes the inference type of the inference result, the inference type including one or more of a management data analysis function, a mobility optimization function, a network energy saving function, or a load balancing function.

11. The method according to any one of claims 6 to 10, characterized in that, The inference result includes one or more of the following: Cell personality offset, cell identifier, network device identifier, time trigger, handover trigger, information on shutting down a cell, information on shutting down a carrier, or information on shutting down a time slot.

12. A control method for the inference ability of a machine model, characterized in that, Including: Sending AI / ML capability information supported by a second communication device, the AI / ML capability information indicating one or more AI / ML capabilities; Wherein, the AI / ML capability information is used for the management and control of the one or more AI / ML capabilities by the second communication device.

13. The method according to claim 12, characterized in that, Further including: Sending the inference type of AI / ML to which the AI / ML capability information applies to the second communication device; the inference type includes one or more of a management data analysis function, a mobility optimization function, a network energy saving function, or a load balancing function.

14. The method according to claim 12 or 13, characterized in that, Further including: Receiving second information from the second communication device, the second information being used to request the first communication device to send the AI / ML capability information.

15. The method according to claim 14, wherein The second information instructs the first communication device to send AI / ML capability information supported by one or more of a management data analysis function, a mobility optimization function, a network energy saving function, or a load balancing function.

16. The method according to any one of claims 12 to 15, characterized in that, The one or more AI / ML capabilities include one or more of the following: a traffic analysis capability, a traffic prediction capability, a mobility analysis capability, a mobility prediction capability, a load analysis capability, a load prediction capability, an energy consumption analysis capability, or an energy consumption prediction capability.

17. The method according to any one of claims 12 to 16, characterized in that Further included are: Receiving first information from the second communication device, where the first information includes the AI / ML capability corresponding to the AI / ML inference; wherein, the one or more AI / ML capabilities include the AI / ML capability corresponding to the AI / ML inference; Sending the inference result of the AI / ML inference to the second communication device.

18. The method according to claim 17, wherein The first information is used to request the first communication device to perform the AI / ML inference of the AI / ML capability corresponding to the AI / ML inference.

19. The method according to claim 17 or 18, characterized in that Further included are: Receiving the inference type of the AI / ML capability, where the inference type includes one or more of a management data analysis function, a mobility optimization function, a network energy saving function, or a load balancing function.

20. The method according to any one of claims 17 to 19, characterized in that Further included are: Sending one or more of an inference status or an inference time.

21. The method according to any one of claims 17 to 20, characterized in that, The inference result further includes the inference type of the inference result, where the inference type includes one or more of a management data analysis function, a mobility optimization function, a network energy saving function, or a load balancing function.

22. The method according to any one of claims 17 to 21, characterized in that, The inference result includes one or more of the following: Cell personality offset, cell identifier, network device identifier, time trigger, handover trigger, information on shutting down a cell, information on shutting down a carrier, or information on shutting down a time slot.

23. A communication device, characterized in that, Including a unit for performing the method according to any one of claims 1 to 11, or including a unit for performing the method according to any one of claims 12 to 22.

24. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which when called by an electronic device, cause the electronic device to perform the method according to any one of claims 1 to 11, or cause the electronic device to perform the method according to any one of claims 12 to 22.

25. A communication system, characterized in that, Including a device for performing the method according to any one of claims 1 to 11 and a device for performing the method according to any one of claims 12 to 22.

26. A chip system, characterized in that, The chip system includes: A communication interface; A processor for calling and running the instructions through the communication interface, so that a device installed with the chip system performs the method according to any one of claims 1 to 11, or so that a device installed with the chip system performs the method according to any one of claims 12 to 22.

27. A computer program product, characterized in that, Including computer-executable instructions, which when running on a computer, cause the computer to perform the method according to any one of claims 1 to 11, or cause the electronic device to perform the method according to any one of claims 12 to 22.

Images