Model construction method and device

Abstract

A model configuration method and apparatus are provided. The method includes an operation, administration and maintenance OAM receiving a first model request message from a first access network device. The OAM determines a first model based on the first model request message and a first mapping relationship, where the first mapping relationship includes a mapping relationship between the model, a model application scenario and a model function, or the first mapping relationship includes a mapping relationship between the model, a model performance level and a model function. The OAM sends information about the first model to the first access network device. According to the method and apparatus of the present application, a first model corresponding to an application scenario or a performance indicator is configured for the first access network device, which can improve the performance and speed of model inference.

Description

[Technical field]

[0001] The present application relates to the field of communications technology, and in particular to a model construction method and apparatus. [Background technology]

[0002] In wireless communication networks, e.g., in mobile communication networks, the services supported by the network are becoming more and more diverse, and therefore the requirements that need to be satisfied are also becoming more and more diverse. For example, the network must be capable of supporting very high speed, very low latency, and / or very large connections. This characteristic makes network planning, network configuration, and / or resource scheduling more and more complex. These new requirements, scenarios, and functions create unprecedented challenges in network planning, operation, and maintenance, as well as efficient operation. To satisfy this challenge, artificial intelligence techniques may be introduced into wireless communication technologies to implement network intelligence. In light of this, how to configure an artificial intelligence model for an access network device in a network is a problem worth considering. Summary of the Invention

[0003] The present application provides a model configuration method and apparatus for configuring a first model corresponding to an application scenario or performance level for a first access network device, so as to improve the performance and speed of model prediction.

[0004] According to a first aspect, there is provided a model configuration method, which may be executed by an Operation, Administration and Maintenance OAM or may be a component (processor, chip, etc.) configured in the OAM, or may be a software module, etc. The method comprises: The method includes receiving a first model request message from the first access network device, determining a first model based on the first model request message and a first mapping relationship, the first mapping relationship including a mapping relationship between the model, the model application scenario, and the model function, or the first mapping relationship including a mapping relationship between the model, the model performance level, and the model function, and sending information about the first model to the first access network device. Optionally, the mapping relationship between the model, the model performance level, and the model function may be replaced with a mapping relationship between the model, the model performance indicator, and the model function.

[0005] According to the above method, the requirements for the performance indicators of models implementing the same function are different in different application scenarios. In the present application, the OAM stores a first mapping relationship, and the first mapping relationship is a mapping relationship between the model, the model application scenario, and the model function. Based on the mapping relationship, for the same model function, in different application scenarios, a model corresponding to the scenario can be assigned, so that the model satisfies the requirements for the performance indicators in the scenario corresponding to the model. Compared with using the same model for model inference in any scenario, the solution of the present application can improve prediction performance, prediction speed, etc. Alternatively, in the present application, the OAM establishes the first mapping relationship, and the first mapping relationship is a mapping relationship between the model, the model performance level (or the model performance indicator), and the model function. Since the requirements for the model performance indicators of the model are different in different applications, and there is a correspondence between the model performance indicators and the performance level, the OAM can configure different models for different model performance indicators in different applications based on the first mapping relationship. Compared with using the same model for model inference under any scenario of model performance indicators, the solution of the present application can improve prediction performance, prediction speed, etc.

[0006] In a possible design, the first model request message indicates a model application scenario and a model function of the first model.

[0007] In a possible design, the first model request message indicates a model capability and a model performance level of the first model, or indicates a model capability and a model performance indicator of the first model.

[0008] In a possible design, the information about the first model is the following for the first model: model index, model structure information, model parameters, model input format, model output format, model performance indicators, model application scenarios, model functions, or training parameters At least one of the following is shown.

[0009] According to the above method, the OAM sends information about the first model to the first access network device. The first access network device can determine or restore the first model based on the information about the first model. Optionally, the information about the first model can include training parameters, and the first access network device can subsequently train the first model based on the training parameters to satisfy various requirements for the model.

[0010] In one possible design, the method further includes receiving first feedback information from the first access network device and updating the first model based on the first feedback information.

[0011] In a possible design, the first feedback information indicates an actual accuracy of the prediction information output by the first model, and updating the first model based on the feedback information includes retraining the first model and updating the first model when the actual accuracy of the prediction information output by the first model does not meet a threshold.

[0012] According to the above method, when determining the first model, the first access network device may perform model inference based on the first model, and the inference result may be referred to as prediction information. The first feedback information is determined based on the prediction information, and the first feedback information includes the actual accuracy of the prediction information output by the first model, and the first feedback information is fed back to the OAM. The OAM updates the first model based on the first feedback information, so as to improve the accuracy of the inference result of the first model.

[0013] In a possible design, the first feedback information indicates a scenario change instruction, a model performance level change instruction, or a model performance indicator change instruction. Updating the first model based on the first feedback information includes selecting a second model for the terminal device based on the scenario change instruction, the model performance level change instruction, or the model performance indicator change instruction, and sending information about the second model to the first access network device, where the second model is for updating the first model.

[0014] Similar to the above method, the difference is that in this design, when the application scenario, model performance level, or model performance indicator changes, the first access network device sends instruction information to the OAM indicating the application scenario change, model performance level change, or model performance indicator change, and the OAM reassigns the model to the first access network device based on the changed application scenario, model performance level, or model performance indicator for model inference, and it can be guaranteed as far as possible that the model assigned to the first access network device meets the requirements for the application scenario, model performance level, and model performance indicator.

[0015] In a possible design, the method further includes receiving a second model request message from the second access network device indicating an index of the first model, determining the first model based on the index of the first model, and sending information regarding the first model to the second access network device.

[0016] According to a second aspect, a model configuration method is provided. For advantageous effects of the method on the first access network device side corresponding to the first aspect, please refer to the first aspect. The method may be executed by the first access network device, or may be a component (processor, chip, etc.) configured in the first access network device, or may be a software module, etc. The method includes: sending a first model request message to an operations, administration and maintenance OAM, where the first model request message indicates a model application scenario and model capabilities of the first model, or the first model request message indicates model capabilities and model performance levels of the first model, or indicates model capabilities and model performance indicators of the first model; and receiving information regarding the first model from the OAM.

[0017] In a possible design, before sending the first model request message to the OAM, the method further includes determining whether an artificial intelligence (AI) based enhanced service needs to be provided.

[0018] In a possible design, the information about the first model is the following for the first model: model index, model structure information, model parameters, model input format, model output format, model performance indicators, model application scenarios, model functions, or training parameters At least one of the following is shown.

[0019] In a possible design, the method further includes performing model inference based on the first model to obtain predictive information.

[0020] In a possible design, the method further includes sending prediction information including the prediction result to a terminal device.

[0021] In a possible design, the prediction information further includes an accuracy for predicting the predicted outcome.

[0022] In a possible design, the method further includes receiving second feedback information from the terminal device, the second feedback information indicating an actual accuracy of the prediction information.

[0023] In a possible design, the method further includes sending first feedback information to the OAM indicating an actual accuracy of the forecast information, an application scenario change instruction, a model performance level change instruction, or a model performance indicator change instruction.

[0024] It should be noted that in the present application, the first model may provide an artificial intelligence (AI) or machine learning (ML) inference service for the terminal device, or may provide an AI or ML inference service for the first access network device. When the first model is configured to provide an AI or ML inference service for the terminal device, the method further includes the first RAN node sending inference information (or called prediction information) of the first model to the terminal device. The terminal device sends second feedback information to the first access network device based on the inference information, where the second feedback information can indicate the actual accuracy of the inference information, and the first access network device can directly forward the second feedback information to the OAM. Alternatively, the first access network device obtains the actual accuracy of the inference information based on the second feedback information, and sends first feedback information indicating the actual accuracy of the inference information, etc., to the OAM. The OAM can update the first model based on the actual accuracy of the inference information to improve the inference performance, inference speed, etc. of the first model.

[0025] In a possible design, the method further includes determining a target cell for the terminal device and sending an index of the first model to a second access network device corresponding to the target cell.

[0026] In a possible design, the method further includes sending first indication information to a second access network device corresponding to the target cell, where the first indication information indicates type information of the input data of the first model.

[0027] In a possible design, the method further includes sending historical data information necessary for model inference to the second access network device.

[0028] In a possible design, the method further includes receiving second indication information from the second access network device, the second indication information indicating one of the following: indicating that the terminal is to be handed over to the target cell, instructing the first access network device to delete the first model, or instructing the first access network device to release corresponding computational resources, and deleting the first model to release the corresponding computational resources.

[0029] According to a third aspect, a model configuration method is provided. For advantageous effects of the method on the first access network device side corresponding to the first aspect, please refer to the first aspect. The method may be executed by the second access network device, or may be a component (processor, chip, etc.) configured in the second access network device, or may be a software module, etc. The method includes: The method includes receiving an index of a first model from the first access network device, sending a second model request message to an Operations, Administration and Maintenance OAM indicating the index of the first model, and receiving information regarding the first model from the OAM.

[0030] In a possible design, the information about the first model is the following for the first model: model index, model structure information, model parameters, model input format, model output format, model performance indicators, model application scenarios, model functions, or training parameters At least one of the following is shown.

[0031] In a possible design, the method further includes receiving first indication information from the first access network device, the first indication information indicating type information of the input data of the first model.

[0032] In a possible design, the method further includes receiving historical data information necessary for model inference from the first access network device, where in the model inference, the historical data information is for determining input data of the first model.

[0033] In a possible design, the method further includes sending second instruction information to the first access network device, where the second instruction information indicates one of the following: indicating that the terminal is to be handed over to the target cell, instructing the first access network device to delete the first model, or instructing the first access network device to release corresponding computational resources.

[0034] According to a fourth aspect, there is provided an apparatus, advantageous effects of which refer to the description of the first aspect, the apparatus may be an operation, administration and maintenance (OAM) apparatus, or an apparatus configured in or adapted for use with an OAM.

[0035] In design, the apparatus may include units that correspond one-to-one to the methods / operations / steps / acts described in the first aspect. The units may be implemented using hardware circuits, software, or a combination of hardware circuits and software.

[0036] For example, an apparatus may include a processing unit and a communication unit, and the processing unit and the communication unit may perform corresponding functions in any design example of the first aspect.

[0037] The communication unit is configured to receive a first model request message from the first access network device. The processing unit is configured to determine a first model based on the first model request message and a first mapping relationship, where the first mapping relationship includes a mapping relationship between the model, the model application scenario, and the model function, or the first mapping relationship includes a mapping relationship between the model, the model performance level, and the model function. The communication unit is further configured to send information about the first model to the first access network device. For specific execution processes of the processing unit and the communication unit, please refer to the first aspect. Details will not be described again here.

[0038] For example, the apparatus includes a memory configured to implement the method described in the first aspect. The apparatus may further include a memory configured to store instructions and / or data. The memory is coupled to the processor. When executing the program instructions stored in the memory, the processor can perform the method described in the first aspect. The apparatus may further include a communication interface, the communication interface used by the apparatus to communicate with other devices. For example, the communication interface may be a transceiver, a circuit, a bus, a module, a pin, or other type of communication interface, and the other device may be a network device, etc. In a possible design, the apparatus may include: a memory configured to store program instructions; a communications interface configured to receive a first model request message from a first access network device; a processor configured to determine a first model based on the first model request message and the first mapping relationship; The first mapping relationship includes a mapping relationship between the model, the model application scenario, and the model function, or the first mapping relationship includes a mapping relationship between the model, the model performance level, and the model function; The communication unit is further configured to send information regarding the first model to the first access network device.

[0039] For the specific execution process of the communication interface and the processor, please refer to the first aspect, and the details will not be described again here.

[0040] According to the fifth aspect, an apparatus is provided. For advantageous effects, please refer to the description of the second aspect. The apparatus may be a first access network device, or an apparatus configured in the first access network device, or an apparatus adapted for use with the first access network device. In design, the apparatus may include units that correspond one-to-one to the methods / operations / steps / acts described in the second aspect. The units may be implemented by using hardware circuits, software, or a combination of hardware circuits and software.

[0041] For example, an apparatus may include a processing unit and a communication unit, and the processing unit and the communication unit may perform corresponding functions in any design example of the second aspect.

[0042] The communication unit is configured to send a first model request message to the operation, administration and maintenance OAM, where the first model request message indicates a model application scenario and a model function of the first model, or the first model request message indicates a model function and a model performance level of the first model, and receive information about the first model from the OAM. The processing unit is configured to determine the first model based on the information about the first model. For specific execution processes of the processing unit and the communication unit, please refer to the second aspect. Details will not be described again here.

[0043] For example, the apparatus includes a memory configured to implement the method described in the second aspect. The apparatus may further include a memory configured to store instructions and / or data. The memory is coupled to the processor. When executing the program instructions stored in the memory, the processor can perform the method described in the second aspect. The apparatus may further include a communication interface, the communication interface used by the apparatus to communicate with other devices. For example, the communication interface may be a transceiver, a circuit, a bus, a module, a pin, or other type of communication interface, and the other device may be a network device, etc. In a possible design, the apparatus may include: a memory configured to store program instructions; a communication interface configured to send a first model request message to an operations, administration, and maintenance OAM, the first model request message indicating a model application scenario and a model capability of the first model, or the first model request message indicating a model capability and a model performance level of the first model, and receive information about the first model from the OAM; and a processor configured to determine the first model based on information about the first model.

[0044] For the specific execution process of the communication interface and the processor, please refer to the second aspect, and the details will not be described again here.

[0045] According to the sixth aspect, an apparatus is provided. For advantageous effects, see the description of the third aspect. The apparatus may be a second access network device, or an apparatus configured in the second access network device, or an apparatus adapted for use with the second access network device. In design, the apparatus may include units that correspond one-to-one to the methods / operations / steps / acts described in the third aspect. The units may be implemented by using hardware circuits, software, or a combination of hardware circuits and software.

[0046] For example, an apparatus may include a processing unit and a communication unit, and the processing unit and the communication unit may perform corresponding functions in any design example of the second aspect.

[0047] The communication unit is configured to receive an index of the first model from the first access network device, send a second model request message indicating the index of the first model to the operation, administration and maintenance OAM, and receive information about the first model from the OAM. The processing unit is configured to determine the first model based on the information about the first model. For specific execution processes of the processing unit and the communication unit, please refer to the third aspect. Details will not be described again here.

[0048] For example, the apparatus includes a memory configured to implement the method described in the third aspect. The apparatus may further include a memory configured to store instructions and / or data. The memory is coupled to the processor. When executing the program instructions stored in the memory, the processor can perform the method described in the third aspect. The apparatus may further include a communication interface, the communication interface used by the apparatus to communicate with other devices. For example, the communication interface may be a transceiver, a circuit, a bus, a module, a pin, or other type of communication interface, and the other device may be a network device, etc. In a possible design, the apparatus may include: a memory configured to store program instructions; a communications interface configured to receive an index of the first model from the first access network device, send a second model request message indicating the index of the first model to an operations, administration, and maintenance OAM, and receive information regarding the first model from the OAM; and a processor configured to determine the first model based on information about the first model.

[0049] For the specific execution process of the communication interface and the processor, please refer to the third aspect, and the details will not be described again here.

[0050] According to a seventh aspect, an embodiment of the present application further provides a computer-readable storage medium comprising instructions which, when executed by a computer, can cause the computer to perform a method according to any one of the first aspect, the second aspect, or the third aspect.

[0051] According to an eighth aspect, the present application further provides a chip system. The chip system includes a processor and may further include a memory configured to implement the method of any one of the first, second, or third aspects. The chip system may include a chip, or may include a chip and other discrete components.

[0052] According to a ninth aspect, the present application further provides a computer program product comprising instructions which, when executed on a computer, enable the computer to carry out a method according to any one of the first aspect, the second aspect or the third aspect.

[0053] According to a tenth aspect, the present application further provides a system comprising an apparatus according to the fourth aspect and an apparatus according to the fifth aspect, or an apparatus according to the fourth aspect, an apparatus according to the fifth aspect and an apparatus according to the sixth aspect. [Brief description of the drawings]

[0054] [Figure 1] FIG. 1 is a diagram of a communication system according to the present application. [Diagram 2] FIG. 1 is a diagram of model deployment according to the present application. [Diagram 3] FIG. 1 is a diagram of model deployment according to the present application. [Figure 4a] FIG. 1 is a diagram of the architecture of a communication system according to the present application. [Figure 4b] FIG. 1 is a diagram of the architecture of a communication system according to the present application. [Figure 4c] FIG. 1 is a diagram of the application of the AI ​​model according to the present application. [Diagram 5] FIG. 1 is a diagram of a neuron according to the present application. [Figure 6] FIG. 2 is a diagram of a neural network according to the present application. [Figure 7] 1 is a schematic flow chart according to the present application. [Figure 8] 1 is a schematic flow chart according to the present application. [Figure 9] FIG. 1 is a diagram of the structure of the device according to the present application. [Figure 10] FIG. 1 is a diagram of the structure of the device according to the present application. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0055] FIG. 1 is an architecture diagram of a communication system 1000 to which the present application is applied. As shown in FIG. 1, the communication system includes a radio access network 100 and a core network 200. Optionally, the communication system 1000 may further include the Internet 300. The radio access network 100 may include at least one access network device (e.g., 110a and 110b in FIG. 1), and may further include at least one terminal device (e.g., 120a-120j in FIG. 1). The terminal device is connected to the access network device in a wireless manner, and the access network device is connected to the core network in a wireless or wired manner. The core network device and the access network device may be different physical devices independent of each other, or the functions of the core network device and the logical functions of the access network device may be integrated into the same physical device, or some of the functions of the core network device and some of the functions of the access network device may be integrated into one physical device. The terminal devices may be connected to each other in a wired or wireless manner, and the access network devices may be connected to each other in a wired or wireless manner. FIG. 1 is merely a diagram. The communication system may further include other network devices, such as wireless relay devices, wireless backhaul devices, etc., which are not shown in FIG.

[0056] The access network device may be a base station, an evolved NodeB (eNodeB), a transmission reception point (TRP), a next generation NodeB (gNB) in a 5th generation (5G) mobile communication system, an access network device in an open radio access network (O-RAN), a next generation base station in a 6th generation (6G) mobile communication system, a base station in a future mobile communication system, an access node in a wireless fidelity (Wi-Fi) system, etc., or may be a module or unit, such as a central unit (CU), a distributed unit (DU), a central unit control plane (CU-CP) module, or a central unit user plane (CU-UP) module that realizes part of the functions of a base station. The access network device may be a macro base station (e.g., 110a in FIG. 1), or may be a micro base station or an indoor base station (e.g., 110b in FIG. 1), or may be a relay node, a donor node, etc. The specific technology used by the access network device and the specific device form are not limited by this application.

[0057] In the present application, an apparatus configured to implement a function of an access network device may be an access network device, or may be an apparatus capable of supporting an access network device to implement a function, such as a chip system, a hardware circuit, a software module, or a hardware circuit and a software module. The apparatus may be installed in an access network device, or may be used in conjunction with an access network device. In the present application, a chip system may include a chip, or may include a chip and other discrete components. For ease of description, the following describes the technical solution provided in the present application by using an example in which an apparatus configured to implement a function of an access network device is an access network device, and the access network device is a RAN node.

[0058] (1) Protocol layer structure

[0059] The communication between the access network device and the terminal device follows a specific protocol layer structure. The protocol layer structure may include a control plane protocol layer structure and a user plane protocol layer structure. For example, the control plane protocol layer structure may include protocol layer functions such as a radio resource control (RRC) layer, a packet data convergence protocol (PDCP) layer, a radio link control (RLC) layer, a media access control (MAC) layer, and a physical layer. For example, the user plane protocol layer structure may include protocol layer functions such as a PDCP layer, an RLC layer, a MAC layer, and a physical layer. In a possible implementation, a service layer adaptation protocol (SDAP) layer may be further included above the PDCP layer.

[0060] Optionally, the protocol layer structure between the access network device and the terminal device may further include an artificial intelligence (AI) layer configured to transmit data related to AI functions.

[0061] (2) Central unit (CU) and distributed unit (DU)

[0062] An access device may include a CU and a DU. A plurality of DUs may be controlled by one CU in a centralized manner. For example, an interface between a CU and a DU may be called an F1 interface. A control plane (CP) interface may be F1-C, and a user plane (UP) interface may be F1-U. The specific names of each interface are not limited in this application. The CU and the DU may be divided based on the protocol layers of a wireless network. For example, the functions of the PDCP layer and the protocol layers above the PDCP layer are set in the CU, and the functions of the protocol layers below the PDCP layer (e.g., the RLC layer and the MAC layer) are set in the DU. As another example, the functions of the protocol layers above the PDCP layer are set in the CU, and the functions of the PDCP layer and the protocol layers below the PDCP layer are set in the DU. This is not limited.

[0063] The above division of the processing functions of the CU and the DU based on the protocol layer is only an example, and other divisions may alternatively be used. For example, the CU or the DU may be defined to have more protocol layer functions, and as another example, the CU or the DU may alternatively be defined to have a part of the processing functions of the protocol layers. In a design, a part of the functions of the RLC layer and the functions of the protocol layers above the RLC layer are set in the CU, and the remaining functions of the RLC layer and the functions of the protocol layers below the RLC layer are set in the DU. In other designs, the division of the functions of the CU or the DU may alternatively be performed based on the service type or other system requirements. For example, the division may be performed based on delay. Functions whose processing time needs to satisfy the delay requirement are set in the DU, and functions whose processing time does not need to satisfy the delay requirement are set in the CU. In other designs, the CU may alternatively have one or more functions of the core network. For example, the CU may be deployed on the network side to facilitate centralized management. In other designs, the radio unit (RU) of the DU is deployed remotely. Optionally, the RU may have radio frequency functions.

[0064] Optionally, the DU and the RU may be differentiated by physical layer (PHY). For example, the DU may implement the upper layer functions of the PHY layer, and the RU may implement the lower layer functions of the PHY layer. When the PHY layer is for transmission, the functions of the PHY layer may include at least one of the following functions: adding a cyclic redundancy check (CRC) code, channel coding, rate matching, scrambling, modulation, layer mapping, precoding, resource mapping, physical antenna mapping, or radio frequency transmission. When the PHY layer is for reception, the functions of the PHY layer may include at least one of the following functions: CRC check, channel decoding, de-rate matching, descrambling, demodulation, layer demapping, channel detection, resource demapping, physical antenna demapping, or radio frequency reception. The upper layer functions of the PHY layer may include some of the functions of the PHY layer. For example, some of the functions are closer to the MAC layer. The lower layer functions of the PHY layer may include other parts of the functions of the PHY layer. For example, the parts of the functions are closer to the radio frequency functions. For example, the upper layer functions of the PHY layer may include CRC code addition, channel coding, rate matching, scrambling, modulation, and layer mapping, and the lower layer functions of the PHY layer may include precoding, resource mapping, physical antenna mapping, and radio frequency transmission functions. Alternatively, the upper layer functions of the PHY layer may include CRC code addition, channel coding, rate matching, scrambling, modulation, layer mapping, and precoding. The lower layer functions of the PHY layer may include resource mapping, physical antenna mapping, and radio frequency transmission functions. For example, the upper layer functions of the PHY layer may include CRC check, channel decoding, de-rate matching, decoding, demodulation, and layer demapping, and the lower layer functions of the PHY layer may include channel detection, resource demapping, physical antenna demapping, and radio frequency reception functions.Alternatively, the upper layer functions of the PHY layer may include CRC checking, channel decoding, de-rate matching, decoding, demodulation, layer demapping, and channel detection, and the lower layer functions of the PHY layer may include resource demapping, physical antenna demapping, and radio frequency reception functions.

[0065] For example, the functionality of the CU may be implemented by one entity, or may be implemented by different entities. For example, the functionality of the CU may be further divided. Specifically, the control plane and user plane of the CU are separated and implemented by different entities, which are a control plane CU entity (i.e., a CU-CP entity) and a user plane CU entity (i.e., a CU-UP entity). The CU-CP entity and the CU-UP entity may be coupled to the DU to jointly realize the functionality of an access network device.

[0066] Optionally, any one of the DU, CU, CU-CP, CU-UP, and RU may be a software module, a hardware structure, or a combination of a software module and a hardware structure. This is not limited. Different entities may exist in different forms, this is not limited. For example, the DU, CU, CU-CP, and CU-UP are software modules, and the RU is a hardware structure. These modules and the methods performed by these modules also fall within the scope of protection of the present disclosure.

[0067] In a possible implementation, the access network device includes a CU-CP, a CU-UP, a DU, and a RU. For example, the present application is performed by a DU, or a DU and a RU, or a CU-CP, a DU, and a RU, or a CU-UP, a DU, and a RU. This is not limited. The method performed by the module also falls within the scope of protection of the present application.

[0068] The terminal device may also be referred to as a terminal, user equipment (UE), mobile station, mobile terminal device, etc. The terminal device is widely used for communication in various scenarios, including, but not limited to, at least one of the following scenarios: device-to-device (D2D), vehicle-to-everything (V2X), machine-type communication (MTC), internet of things (IOT), virtual reality, augmented reality, industrial control, autonomous driving, telemedicine, smart grid, smart furniture, smart office, smart wearable, smart transportation, smart city, etc. The terminal device may be a mobile phone, a tablet computer, a computer with a wireless transceiver function, a wearable device, a vehicle, an unmanned aerial vehicle, a helicopter, an airplane, a ship, a robot, a robot arm, a smart home device, etc. The specific technology used by the terminal device and the specific device form are not limited by this application.

[0069] In the present application, the apparatus configured to implement the functions of the terminal device may be a terminal device, or may be an apparatus capable of supporting the terminal device to implement the functions, such as a chip system, a hardware circuit, a software module, or a hardware circuit and a software module. The apparatus may be installed in the terminal device, or may be compatible with and used in the terminal device. For ease of description, the following describes the technical solution provided in the present application by using an example in which the apparatus configured to implement the functions of the terminal device is a terminal device, and the terminal device is a UE.

[0070] The base station and the terminal device may be fixed or mobile. The base station and / or the terminal device may be deployed on the ground, including indoor or outdoor scenarios, and handheld or vehicle-mounted scenarios, or on water, or in the air, on airplanes, balloons, and satellites. The application scenarios of the base station and the terminal device are not limited in this application. The base station and the terminal device may be deployed in the same scenario or in different scenarios. For example, the base station and the terminal device are both deployed on the ground. Alternatively, the base station is deployed on the ground and the terminal is deployed on the water. The examples are not described one by one.

[0071] The roles of the base station and the terminal device may be relative. For example, the helicopter or unmanned aerial vehicle 120i in FIG. 1 may be configured as a mobile base station. In the case of the terminal device 120j accessing the wireless access network 100 via 120i, the terminal device 120i is a base station, while in the case of the base station 110a, 120i is a terminal device, in other words, 110a and 120i communicate with each other based on a wireless air interface protocol. 110a and 120i may alternatively communicate with each other based on an interface protocol between base stations. In this case, 120i is also a base station for 110a. Thus, both the base station and the terminal device may be collectively referred to as a communication device, and 110a and 110b in FIG. 1 may be referred to as a communication device having the function of a base station, and 120a-120j in FIG. 1 may be referred to as a communication device having the function of a terminal device.

[0072] In this application, an independent network element (e.g., referred to as an AI network element or AI node) may be introduced into the communication system shown in FIG. 1 to implement AI-related operations. The AI ​​network element may be directly connected to the access network device in the communication system, or may be indirectly connected to the access network device via a third-party network element. The third-party network element may be a core network element, such as an authentication management function (AMF) or a user plane function (UPF). Alternatively, the AI ​​function, AI module, or AI entity may be configured in another network element in the communication system to implement AI-related operations. For example, the other network element may be an access network device (e.g., gNB), a core network device, or an operation, administration and maintenance (OAM). In this case, the network element that performs AI-related operations is a network element with built-in AI functions. The OAM is configured to perform operation, administration, maintenance, etc. on the access network device and / or the core network device.

[0073] In the present application, as shown in FIG. 2 or FIG. 3, the AI ​​model may be deployed in at least one of a core network device, an access network device, a terminal device, an OAM, etc., and the corresponding function is implemented by using the AI ​​model. In the present application, the AI ​​models deployed in different nodes may be the same or different. The models differ in at least one of the following: different structural parameters of the model, for example, different numbers of layers and / or weights of the model, different input parameters of the model, or different output parameters of the model. Different input parameters of the model and / or different output parameters of the model may be described as different functions of the model. Different from FIG. 2, in FIG. 3, the access network device is divided into a CU and a DU. Optionally, the CU and the DU may be the CU and the DU in the O-RAN architecture. One or more AI models may be deployed in the CU, and / or one or more AI models may be deployed in the DU. Optionally, the CU in FIG. 3 may be further divided into a CU-CP and a CU-UP. Optionally, one or more AI models may be deployed in the CU-CP and / or one or more AI models may be deployed in the CU-UP. Optionally, in FIG. 2 or FIG. 3, the OAM of the access network device and the OAM of the core network device may be deployed separately.

[0074] Optionally, FIG. 4a is an architecture of a communication system according to the present application. As shown in FIG. 4a, in a first design, an access network device includes a near-real-time access network intelligent controller (RAN intelligent controller, RIC) module configured to train a model and perform inference. For example, the near-real-time RIC may be configured to train an AI model and perform inference by using the AI ​​model. For example, the near-real-time RIC may obtain network-side and / or terminal-side information from at least one of a CU, a DU, or a RU, and the information may be used as training data or inference data. Optionally, the near-real-time RIC may present an inference result to at least one of a CU, a DU, a RU, or a terminal device. Optionally, the CU and the DU may exchange inference effects. Optionally, the DU and the RU may exchange inference results. For example, the near-real-time RIC presents the inference result to the DU, and the DU forwards the inference result to the RU.

[0075] Alternatively, in a second design, as shown in FIG. 4a, the access network device may include a non-real-time RIC (optionally, the non-real-time RIC may be located in an OAM or core network device) configured to train the model and perform inference. For example, the non-real-time RIC is configured to train an AI model and perform inference by using the model. For example, the non-real-time RIC may obtain network-side and / or terminal-side information from at least one of the CU, DU, or RU, and the information may be used as training data or inference data. The inference result may be presented to at least one of the CU, DU, RU, or terminal device. Optionally, the CU and DU may exchange the inference result. Optionally, the DU and RU may exchange the inference result. For example, the non-real-time RIC presents the inference result to the DU, and the DU forwards the inference result to the RU.

[0076] Alternatively, in a third design, as shown in FIG. 4a, the access network device includes a near real-time RIC, and the non-real-time RIC is located outside the access network device (optionally, the non-real-time RIC may be located in an OAM or core network device). As in the second design above, the non-real-time RIC may be configured to train and infer the model, and / or as in the first design, the non-real-time RIC may train the model, and the near real-time RIC may obtain AI model information from the non-real-time RIC, obtain network-side and / or terminal-side information from at least one of the CU, DU, or RU, and obtain an inference result by using the information and the AI ​​model information. Optionally, the near real-time RIC may present the inference result to at least one of the CU, DU, RU, or terminal device. Optionally, the CU and DU may exchange the inference result. Optionally, the DU and RU may exchange the inference result. For example, the near real-time RIC presents the inference result to the DU, and the DU forwards the inference result to the RU. For example, the near real-time RIC is configured to train model A and perform inference by using model A. For example, the non-real-time RIC is configured to train model B and perform inference by using model B. For example, the non-real-time RIC is configured to train model C and send information about model C to the near real-time RIC, which uses model C for inference.

[0077] Fig. 4b is another architecture of a communication system according to the present application. Compared with Fig. 4a, in Fig. 4b, the CU is separated into a CU-CP and a CU-UP.

[0078] An AI model is a specific implementation of an AI function. An AI model represents a mapping relationship between the input and output of a model. An AI model may be a neural network, a linear regression model, a decision tree model, a support vector machine (SVM), a Bayesian network, a Q-learning model, other machine learning models, etc. In this application, an AI function may include at least one of the following: data collection (collection of training data and / or inference data), data preprocessing, model training (also called model learning), model information release (model information setting), model validation, model inference, or inference result release. Inference may also be called prediction. In this application, an AI model may be called model for short.

[0079] FIG. 4c is a diagram of the application architecture of the AI ​​model. A data source is configured to store training data and inference data. A model training host analyzes or trains the training data provided by the data source to obtain an AI model, and deploys the AI ​​model to a model inference node (model inference host). Optionally, the model training node may further update the AI ​​model deployed to the model inference node. The model inference node may further feed back related information of the deployed model to the model training node, so that the model training node optimizes or updates the deployed AI model, etc.

[0080] Obtaining an AI model through learning by the model training node is equivalent to the model training node obtaining a mapping relationship between the input and output of the model through learning by using the training data. The model inference node uses the AI ​​model to make inferences based on the inference data provided by the data source, and obtains an inference result. The method may also be described as follows: the model inference node inputs the inference data to the AI ​​model, and obtains an output by using the AI ​​model. The output is the inference result. The inference result may indicate the configuration parameters used (performed) by the subject of the action, and / or the action performed by the subject of the action. The inference result may be centrally planned by an actor entity, and sent to one or more subjects of the action (e.g., network entities) for action. Optionally, the actor entity or the subject of the action may further feed back parameters or measurements collected by the actor entity or actor object to the data source. This process may be called performance feedback, and the fed back parameters may be used as training data or inference data. Optionally, the actor entity or actor object may further determine feedback information regarding the model performance based on the inference result output by the model inference node, and feed back the feedback information to the model inference node. The model inference node may feed back the model performance information to the model training node based on the feedback information, so that the model training node optimizes or updates the deployed AI model. This process may be referred to as model feedback.

[0081] The AI ​​model may be a neural network or other machine learning model. A neural network is used as an example. A neural network is a specific embodiment of machine learning technology. According to the universal approximation theorem, a neural network can theoretically approximate any continuous function, so that the neural network has the ability to learn any mapping. Therefore, the neural network can accurately perform abstract modeling for complex high-order problems.

[0082] The idea of ​​neural networks comes from the neuron structure of the brain tissue. Each neuron performs a weighted sum operation on the neuron's input values ​​and outputs the weighted sum result through an activation function. Figure 5 shows the structure of a neuron. The input values ​​of a neuron are x=[x0,x1,...,x n ], and the weights corresponding to the input values ​​are w=[w,w1, ,w n ] and the bias of the weighted sum is assumed to be b. The form of the activation function can be varied. The activation function of a neuron is assumed to be y = f(z) = max(0,z). In this case, the output of the neuron is

number

number

[0083] A neural network usually includes a multi-layer structure, and each layer may include one or more neurons. Increasing the depth and / or width of a neural network can improve the representation capability of the neural network, providing a stronger information extraction and abstract modeling capability for complex systems. The depth of a neural network can refer to the number of layers included in the neural network, and the number of neurons included in each layer can be called the layer width. FIG. 6 is a diagram of the layer relationship of a neural network. In an implementation, a neural network includes an input layer and an output layer. After performing neuron processing on the received input, the input layer of the neural network transfers the result to the output layer, and the output layer obtains the output result of the neural network. In another implementation, a neural network includes an input layer, a hidden layer, and an output layer. After performing neuron processing on the received input, the input layer of the neural network transfers the result to an intermediate hidden layer, and then the hidden layer transfers the calculation result to the output layer or an adjacent hidden layer, and finally, the output layer obtains the output result of the neural network. A neural network may include one hidden layer or multiple hidden layers connected in series. This is not limited. In the training process of a neural network, a loss function may be defined. The loss function represents the gap or difference between the output value of the neural network and the ideal target value of the neural network. The specific form of the loss function is not limited in this application. The training process of a neural network is a process of adjusting neural network parameters such as the number and width of layers of the neural network, the weights of neurons, and parameters in the activation functions of neurons, so that the value of the loss function is smaller than a threshold or meets the target requirement.

[0084] In one solution, the RAN node supports multiple AI-assisted cases, such as mobility enhancement, load balancing, and network energy saving. The OAM is configured to train or hold the AI ​​model. The RAN node requests the AI ​​model from the OAM and performs inference by using the AI ​​model. How the OAM configures the corresponding model for the RAN node is a problem to be solved in this application. In this specification, the OAM may be replaced by other nodes that can train or hold the AI ​​model, such as a cloud server, a core network device, or other possible devices.

[0085] As shown in FIG. 7, a procedure of a model construction method is provided, which includes at least the following steps:

[0086] Step 701: A first RAN node sends a first model request message to an OAM, and correspondingly, the OAM receives a first model request message from the first RAN node.

[0087] For ease of distinction, a model request message sent by a first RAN node to an OAM is called a first model request message. A model request message sent by a second RAN node to an OAM is called a second model request message. A model request message is used by a RAN node to request a model from an OAM, and the name of the model request message is not limited. For example, a first model request message may be called a first message, and a second model request message may be called a second message. For distinction between different RAN nodes, these RAN nodes are called a first RAN node, a second RAN node, etc. The first RAN node and the second RAN node are different RAN nodes. The relationship between the first RAN node and the second RAN node is not limited.

[0088] In the design, the first RAN node may determine a service scenario and functional requirements, and determine the first model request message based on the service scenario and functional requirements. The service scenario and functional requirements are service scenarios and functional requirements of the RAN node. For example, the first model requested by using the first model request message is for model inference related to the RAN node, for example, for cell energy saving. The first RAN node may determine the first model request message, etc., based on the service scenario and functional requirements of the RAN node. Alternatively, the service scenario and functional requirements are service scenarios and functional requirements of the UE. For example, the first model requested by using the first model request message is for model inference related to the UE, for example, for prediction of the UE's movement trajectory. The first RAN node may determine the first model request message, etc., based on the service scenario and functional requirements of the UE. Alternatively, the first RAN node may determine the first model request message, etc., based on the service scenario and functional requirements of other nodes other than the RAN node and the UE. This is not limited. The first RAN node may obtain the service scenario and functional requirements of the UE based on information reported by the UE. For example, the UE is in an autonomous driving scenario and sends a quality of service (QoS) prediction service request to the RAN node. The RAN node may obtain the service scenario and functional requirements of the UE as the autonomous driving and QoS prediction. Alternatively, the first RAN node may obtain the service scenario and functional requirements of the UE through analysis. For example, the RAN node finds that the UE moves to a cell edge and requires a mobility enhancement service, and the mobility enhancement service requires a prediction of the trajectory of the UE, so the service scenario and functional requirements of the UE are mobility enhancement and prediction of the trajectory information of the UE. Alternatively, the first RAN node may obtain the service scenario and functional requirements of the UE from another device, including but not limited to a minimization of drive tests (MDT) device, etc. Alternatively, the first RAN node stores the service scenario and functional requirements of the UE.For example, when accessing the first RAN node, the UE proactively reports the UE's service scenario and function requirements to the first RAN node.

[0089] In this application, the service scenario includes, but is not limited to, at least one of the following: mobility enhancement, load balancing, virtual reality (VR), augmented reality (AR), vehicle-to-everything (autonomous driving), wireless robot cloud control, remote diagnosis with force feedback, high definition video (e.g., 8K video), etc. For example, mobility enhancement refers to enhancing services to provide cell handover to a moving UE. For example, a RAN node predicts a target cell for a UE, a handover time, etc., based on at least one of a UE movement trajectory, a measurement report, a cell load status, etc. The target cell for a UE may refer to a serving cell to which the UE is about to be handed over, and the movement manner of a UE includes movement on a fixed route, movement on a non-fixed route, etc. Mobility enhancement may include mobility enhancement for vehicular movement on a fixed route, mobility enhancement for vehicular movement on a non-fixed route, mobility enhancement for pedestrian scenarios, etc. In remote diagnosis, also called remote consultation, a doctor diagnoses or operates on a patient at another location through a network. The above-mentioned force feedback is a haptic feedback of force, which is a service with very high latency requirements, and a latency prediction service needs to be provided. For example, if the predicted latency exceeds a threshold, the information transmission solution needs to be adjusted, or an alarm needs to be provided to respond to an emergency situation in advance.

[0090] In this application, the functional requirements include, but are not limited to, at least one of the following: UE trajectory prediction, load prediction, signal to interference plus noise ratio (SINR) prediction, reference signal received power (RSRP) prediction, reference signal received quality (RSRQ) prediction, latency prediction, QoS prediction, throughput prediction, cell energy savings, etc.

[0091] In design, after the first RAN node determines the first model request message based on the service scenario and the functional requirement, the first model request message indicates the service scenario and the functional requirement, in other words, the first model request message indicates the model application scenario, the model function, etc. of the first model being requested. The model application scenario corresponds to the above service scenario, and the model function corresponds to the above functional requirement. The instruction in this application may include an implicit instruction, an explicit instruction, etc. This is not limited. For example, the first model request message may include the service scenario, the functional requirement, etc. Alternatively, the first RAN node may directly send the service scenario and the functional requirement to the OAM, etc. For example, the combination of the service scenario and the functional requirement sent by the first RAN node to the OAM may include mobility enhancement scenario and trajectory prediction, load balancing and trajectory prediction, autonomous driving and QoS prediction, VR / AR and QoS prediction, wireless robot cloud control and latency prediction, remote diagnosis with force feedback and latency prediction, etc.

[0092] In another design, the process of determining the first model request message by the first RAN node based on the service scenario and the functional requirements includes: the first RAN node determines model capabilities and model performance indicators of the requested first model based on the service scenario and the functional requirements. The model capabilities correspond to the functional requirements above. The model performance indicators include at least one of the following: accuracy, computational complexity, memory occupancy, inference duration, etc. The accuracy is a percentage of the number of samples currently predicted to the total number of samples. The computational complexity is the number of calculations required by the model. The total computational complexity of the model is equal to the sum of the computational complexity of the operators in the model, and the computational complexity feedbacks the model's requirements to the hardware computation unit. Alternatively, the computational complexity may be the size of the computational resources occupied when the model is running. The memory occupancy refers to the size of memory and video RAM occupied when the model is running, or the memory occupancy may be referred to as the size of the model, etc. The inference time refers to the duration required by the model to infer the prediction result. The first model request message sent by the first RAN node to the OAM indicates the model capabilities and model performance indicators of the first model. For example, the first model request message includes the model capabilities and the model performance indicators. Alternatively, the first RAN node may directly send the model capabilities and the model performance indicators to the OAM. The model capabilities may also be called model categories. For example, the first RAN node learns through analysis that a QoS prediction service in an autonomous driving scenario needs to be provided to the UE, and the first RAN node can continue to analyze the performance indicators required for the QoS prediction service in an autonomous driving scenario, and the first RAN node sends the QoS prediction service and the corresponding performance indicators to the OAM. Alternatively, the first RAN node determines a model performance level based on the model performance indicators. There is a correspondence between the model performance level and the model performance indicators. For details, please refer to the following description. The first model request message sent by the first RAN node to the OA indicates the model capabilities and the model performance level of the first model being requested.

[0093] Optionally, before step 701, the method may further include: the first RAN node determining whether an AI-based enhanced service needs to be provided. If an AI-based enhanced service needs to be provided, step 701 is needed to send a first model request message to the OAM. If an AI-based enhanced service does not need to be provided, the procedure shown in FIG. 7 is not executed. For example, in an AI-based mobility enhancement solution, the source RAN node may predict the target cell of the UE in an AI manner. However, if the UE is not located at the cell edge, the source RAN node does not need to predict the target cell of the UE. In other words, only when the UE is at the cell edge, the source RAN node needs to send a first model request message to the OAM to request a model for predicting the target cell. Alternatively, if the source RAN node predicts the target cell of the UE in an AI manner, the input of the model includes the trajectory information of the UE. However, if the trajectory of the UE is very random, the accuracy of the target cell predicted by the AI ​​manner may be greatly reduced. In this case, the source RAN node may consider not using the AI ​​manner to predict the target cell of the UE. In this case, the source RAN node no longer sends a first model request message to the OAM, but predicts the target cell for the UE in a conventional manner.

[0094] Step 702: The OAM determines a first model based on the first model request message and the first mapping relationship.

[0095] In this application, in order to distinguish different models, the different models are respectively referred to as a first model, a second model, etc. A model represents a mapping relationship between an input and an output of a model, and is for implementing a specific function. The name of a model is not limited. For example, a model may be called an AI model. An AI model may be a neural network, a linear regression model, a decision tree model, an SVM, a Bayesian network, a Q-learning model, or other machine learning model. The first mapping relationship represents a mapping, a correspondence, etc. The first mapping relationship may also be called a correspondence, etc. Similar cases are not described one by one thereafter.

[0096] In the present application, the OAM may obtain the first mapping relationship. The first mapping relationship may be established by the OAM or another node. The OAM obtains the first mapping relationship via the other node.

[0097] In an example, the first mapping relationship includes a mapping relationship between a model, a model performance level, and a model function, where the model performance level may be referred to as a performance level for short.

[0098] For example, there is a one-to-many correspondence between model performance levels and model performance indicators. For example, the model performance indicator for model level classification includes accuracy, and the levels may be classified based on accuracy. The higher the accuracy, the higher the model performance level. Alternatively, the model performance indicator for model level classification includes inference time, and the levels may be classified based on inference time. The shorter the inference time, the higher the model performance level. Alternatively, the model performance indicator for model level classification includes at least two indicators of accuracy, computational complexity, inference time, and memory occupancy. For example, the multiple indicators are weighted in a specific weighting manner to determine a weighted value, and the levels are classified based on the weighted value. For models with different functions, the methods of model performance level classification are the same or different. For example, the model performance levels are classified based on the model accuracy of the model for trajectory prediction and the model for SINR prediction. It is assumed that the model for trajectory prediction and the model for SINR prediction both include three model performance levels. As shown in Table 1, for a model for trajectory prediction, the accuracy corresponding to model performance level 1 is [n1, n2], the accuracy corresponding to model performance level 2 is (n2, n3], and the accuracy corresponding to model performance level 3 is (n3, n4], where n4 is equal to or greater than n3, n3 is equal to or greater than n2, and n2 is equal to or greater than n1. As shown in Table 2, for a model for SINR prediction, the accuracy corresponding to model performance level 1 is [m1, m2], and the accuracy corresponding to model performance level 2 is (m2, m3], where m3 is equal to or greater than m2, and m2 is equal to or greater than m1. As another example, as shown in Table 3, for a model for SINR prediction, the accuracy corresponding to model performance level 1 is [t1, t2], the accuracy corresponding to model performance level 2 is (t2, t3], and the accuracy corresponding to model performance level 3 is (t3, t4], where t4 is equal to or greater than t3, t3 is equal to or greater than t2, and t2 is equal to or greater than t1. [Table 1] [Table 2] [Table 3]

[0099] In design, the first RAN node may obtain a model application scenario, determine corresponding model performance indicators based on the model application scenario, and determine a model performance level based on the model performance indicators. The first model request message sent to the OAM indicates the model performance level and the model capability. The OAM determines a model, referred to as the first model, from a first mapping relationship based on the model performance level and the model capability indicated by the first model request message. The first mapping relationship includes a mapping relationship between the model, the model performance level, and the model capability. See the following description of Table 6.

[0100] In an example, the first mapping relationship includes a mapping relationship between a model, a model application scenario, and a model function. For example, the first mapping relationship includes an AI model, an autonomous driving scenario, and a QoS prediction.

[0101] For example, the first RAN node may obtain a model application scenario and a model function of the first model. See the description of step 701. The first model request message sent by the first RAN node to the OAM indicates the model application scenario and the model function of the first model. The OAM may determine an AI model from the first mapping relationship based on the application scenario and the model function indicated by the first model request message. The AI ​​model is called a first model. For example, as shown in Table 4, the first mapping relationship includes four application scenarios (having indexes 1 to 4) and three model functions (having indexes 1 to 3). The four application scenarios and three model functions can form 12 mapping relationships, and the indexes of the 12 mapping relationships are 1 to 12. For example, the mapping relationship 1 includes an application scenario 1, a model function 1, and a model 1, and the mapping relationship 2 includes an application scenario 1, a model function 2, and a model 2. [Table 4]

[0102] In the first mapping relationship shown in Table 4, for the same model function, the corresponding models are different in different scenarios. For example, model function 1 corresponds to model 1, model 4, model 7, and model 10 in application scenarios 1 to 4, respectively. It is used as an example that the OAM establishes the first mapping relationship shown in Table 4. For the same model function, the OAM may separately collect corresponding training data in different scenarios and train the corresponding AI model. For example, in application scenario 1, the OAM may collect training data corresponding to model function 1 and train model 1 by using the training data, so that model 1 can satisfy the performance indicators of application scenario 1. The training process may be machine learning, non-machine learning, reinforcement learning, etc. This is not limited.

[0103] Alternatively, in this example, there is a one-to-one correspondence between model performance levels and model performance indicators, where each model performance level corresponds to a unique model performance indicator, and each model performance indicator corresponds to a unique model performance level. For example, as shown in Table 5, model performance levels 1-3 correspond to model performance indicators 1-3, respectively. It is used as an example that the model performance indicators include accuracy, computational complexity, inference time, and memory occupancy. Model performance level 1 corresponds to model performance indicator 1, and correspondingly, the accuracy is at least 99%, the computational complexity is up to 3%, the inference time is up to 1 second, and the memory occupancy is up to 500KB. [Table 5]

[0104] In this example, the first RAN node may obtain a model capability indicator for the requested model, and the first model request message sent to the OAM indicates the model capability and the model capability indicator for the requested model. Upon receiving the first model request message, the OAM may determine a model capability level based on the model capability indicator indicated by the first model request message. Alternatively, the first RAN node may determine a model capability level based on the model capability indicator, and the first model request message sent to the OAM indicates the model capability and the model capability level. The OAM determines a model from a first mapping relationship based on the model capability and the model capability level, and the model is referred to as the first model, etc. For example, as shown in Table 6, the first mapping relationship includes 12 mapping relationships with indexes of 1 to 12. For example, mapping relationship 1 includes model 1, model performance level 1, and model capability 1, and mapping relationship 2 includes model 2, model performance level 1, and model capability 2. [Table 6]

[0105] It should be noted that in the above example where there is a one-to-one correspondence between the model performance level and the model performance indicator, the model performance level in the above first mapping relationship may be replaced with the model performance indicator. Alternatively, in the present application, the first mapping relationship may be established based on the model performance indicator. For example, the first mapping relationship includes a mapping relationship between a model, a model performance indicator, and a model function. In this design, the first model request message includes a model function and a model performance indicator of the first model. The OAM determines a model from the first mapping relationship based on the model function and the model performance indicator indicated by the first model request message. The model may be referred to as a first model. For example, as shown in Table 7, the first mapping relationship includes 12 mapping relationships with indexes of 1 to 12. For example, mapping relationship 1 includes model 1, model performance indicator 1, and model function 1, and mapping relationship 2 includes model 2, model performance indicator 1, and model function 2. [Table 7]

[0106] In the first mapping relationship shown in Table 6 or Table 7, the models corresponding to the same model function are different under different model performance levels and model performance indicators. For example, refer to Table 6 or Table 7. In the case of model function 1, the models corresponding to model performance indicators 1 to 4 or model performance indicators 1 to 4 are model 1, model 4, model 7, and model 10, respectively. It is used as an example that OAM establishes the first mapping relationship shown in Table 6 or Table 7. OAM may separately collect corresponding training data under different model performance levels or model performance indicators and train corresponding models. For example, OAM may collect training data corresponding to model function 1 under model performance level 1 or model performance indicator 1 and train model 1.

[0107] As mentioned above, the model performance indicators include at least one of the following: accuracy, computational complexity, memory occupancy, inference time, etc. In different scenarios, the same model function may have different requirements for the model performance indicators. Optionally, the computational complexity in the model performance indicators is also called computational resource occupancy, and the memory occupancy is also called model size. In a possible case, as shown in Table 8, in the VR / AR scenario and the V2X scenario, the requirements for the model's performance indicators in QoS prediction are different. In another possible case, as shown in Table 9, in scenarios such as a vehicle on a fixed route, a vehicle on a non-fixed route, and a pedestrian scenario, the requirements for the model's performance indicators in trajectory prediction are different. [Table 8] [Table 9]

[0108] In different scenarios, the requirements for performance indicators of the model of the same function may be different. If the same model for a function is used for inference in any scenario, it is difficult to satisfy the requirements for the performance indicators of the corresponding scenario. For example, in Table 8, in the case of the AR / VR scenario and the V2X scenario, if the same model in QoS prediction is used for model inference, it is difficult to satisfy the requirements for the performance indicators of each scenario. In the solution of the present application, the first mapping relationship shown in Table 4 above can be referred to. For the same model function, different models can be configured for different scenarios. For example, in the VR / AR scenario, a model is configured for QoS prediction. In the V2X scenario, another model is configured for QoS prediction. Alternatively, the mapping relationship shown in Table 6 above can be referred to. Models with corresponding performance levels can be configured for the requirements for performance indicators in different scenarios. Alternatively, the mapping relationship shown in Table 7 above can be referred to, and different models can be configured for different performance indicators, so that each model satisfies the requirements for the performance indicators of the corresponding application scenario.

[0109] Step 703: The OAM sends information about the first model to the first RAN node.

[0110] In this application, the information about the first model indicates at least one of the following of the first model: model index, model structure information, model parameters, model input format, model output format, input data processing method, output data processing method, model performance indicators, model application scenario, model functions, training parameters, etc.

[0111] The model index is for distinguishing different models. The model structure information, the model parameters, the model input format, the model output format, etc. are used by the first RAN node to determine the first model. In other words, the first RAN node determines the first model based on the model structure information, the model parameters, the model input format, the model output format, etc. The model structure information includes at least one of the following: the number of layers included in the model (also called the depth of the model), the network type of each layer (e.g., a fully connected layer network, a convolutional layer network, a long short-term memory (LSTM) layer), the number of neurons included in each layer (also called the width of the model), the connection relationship between layers, etc. The model parameters may include at least one of the following: the weights of neurons, the activation functions of neurons, the offsets in the activation functions, etc. The model input format and the model output format may be formats that need to be satisfied by the input data and the output data when model training or inference is performed on the model. The above input data processing method includes a method of preprocessing the collected raw data before the raw data is input to the model. For example, the imaginary part is separated from the real part of the raw data, a normalization process is performed on the raw data, or the phase and amplitude of the raw data are separated. The above output data processing method includes a method of processing the output data when the model outputs data, for example, intercepting the output data. For the model performance indicators, model application scenarios, and model functions, please refer to the above description. In this application, when determining the first model based on information such as the model structure, model parameters, model input format, and model output format, the first RAN node may further continue to train the first model based on the above training parameters. The training parameters may include instruction information of the loss function, etc. For example, the first RAN node may obtain training data. The training data may be from the first RAN node, or from the UE, or from both the first RAN node and the UE, or from another node.The first RAN node trains the first model based on the training data. For example, the training data is input to the first model. A value of a loss function is determined based on the output of the first model. If the value of the loss function is less than a threshold or satisfies a target requirement, the training for the first model ends. If the value of the loss function is equal to or greater than the threshold or does not satisfy the target requirement, the parameters of the first model are updated, and the training for the first model continues. Updating the parameters of the first model includes updating at least one of the following information of the first model: the number of layers of the neural network, the width of the neural network, the connection relationship between the layers, the weights of the neurons, the activation functions of the neurons, the offsets in the activation functions, etc. The training process may be label learning, unlabel learning, reinforcement learning, etc. This is not limited. When the first model is trained, the initial model of the first model may be instructed by information about the first model, or may be randomly generated, or may be agreed upon in a protocol. This is not limited.

[0112] Alternatively, the information about the first model indicates at least one of the following of the first model: model, model usage instructions, or model index. The model includes network architecture, parameter weights, training hyperparameters, optimizer, batch size, learning rate, momentum, etc. Optionally, the first RAN node may perform online learning, etc. based on the training hyperparameters. The model usage instructions include at least one of model input / output format, model performance indicators, model application scenarios, and model functions. The model index is a model number, etc., used to distinguish different models.

[0113] During practical application, the requirements for performance indicators of models implementing the same function are different in different application scenarios. For example, as shown in the above Table 8 or Table 9, in the case of QoS prediction and trajectory prediction, the requirements for performance indicators of models are different in different scenarios. In the present application, the OAM establishes a first mapping relationship, and the first mapping relationship is a mapping relationship between a model, a model application scenario, and a model function. Based on the mapping relationship, in the case of the same model function, in different application scenarios, a model corresponding to the scenario can be constructed, so that the model satisfies the requirements for performance indicators in the scenario corresponding to the model. Compared with the case where the same model is used for model inference in any scenario, the solution of the present application can improve prediction performance, prediction accuracy, etc. Alternatively, in the present application, the OAM establishes a first mapping relationship, and the first mapping relationship is a mapping relationship between a model, a model performance level, and a model function, or the first mapping relationship is a mapping relationship between a model, a model performance indicator, and a model function. Since the requirements for model performance indicators are different in different application scenarios, and there is a correspondence between the model performance indicators and the model performance levels, the OAM can configure different models for different requirements for the model performance indicators in different application scenarios based on the first mapping relationship.Compared with the case where the same model is used for model inference under the model performance indicators in any scenario, the solution of the present application can improve prediction performance, prediction speed, etc.

[0114] Optionally, in addition to steps 701 to 703, the procedure shown in FIG. 7 may further include the following steps:

[0115] Step 704: The first RAN node performs model inference based on the first model to obtain prediction information.

[0116] For example, the first RAN node may collect input data and allocate computational resources, memory resources, etc. for model inference. The input data may be from the first RAN node and / or UE, or from other nodes, etc. This is not limited. Optionally, the first RAN node may collect input data based on the model usage instructions. The input data is input to the first model and output as prediction information. Alternatively, the input data is processed, the processed input data is input to the first model, and / or the output data of the first model is processed, and the processed output data is prediction information. It may be understood that the input data and the output data satisfy the requirements for the input format and the output format of the first model. The prediction information includes a prediction result. For example, when a model training node such as an OAM trains a model, the model training node performs model training by using a training set to obtain a model, and this process may be referred to as model training. The model is validated by using a validation set, and this process may be referred to as model validation. Model validation is usually performed during model training. For example, each time the model is trained for one or more epochs, the validation set may be used to validate the current model to monitor the model training status, for example, to validate whether the current model is underfitted, overfitted, converged, and to determine whether training should be terminated. Optionally, in the model validation process, the hyperparameters of the model may be further adjusted. The hyperparameters may refer to at least one of the following parameters of the model: the number of layers of the neural network, the number of neurons, the activation function, the loss function, etc. The model is tested by using the test set, and this process may be called model testing. For example, the generalization ability of the model is evaluated, it is determined whether the model satisfies the requirements, it is determined whether the modem is available, or the prediction accuracy of the model is obtained. In this application, the information about the first model transmitted by the OAM to the first RAN node may further include the prediction accuracy, etc.Alternatively, when determining the first model, the first RAN node may test the first model by using locally collected test data to obtain prediction accuracy, etc. In addition to the prediction result inferred by the first model, optionally, the prediction information transmitted by the first RAN node to the UE may further include the prediction accuracy. For example, in trajectory prediction, the prediction information transmitted by the first RAN node to the UE includes predicted trajectory information of the UE and the accuracy (e.g., 98%) of the predicted trajectory information.

[0117] In this application, when the prediction information is prediction information related to the first RAN node, the first RAN node determines the first feedback information based on the prediction information. For example, the first RAN performs a corresponding operation based on the prediction information. For example, in a network energy saving scenario, the prediction information is an energy saving policy, such as carrier shutdown, symbol shutdown, cell shutdown, or deep sleep. After the first RAN node infers the energy saving policy by using the first model, the first RAN node may execute the energy saving policy. The first RAN node determines the first feedback information based on a change in performance (e.g., a change in cell throughput or QoS of the UE) before and after the prediction information is executed. Alternatively, the prediction information is prediction information related to the UE, and the first RAN node needs to transmit the prediction information to the UE. The UE determines the second feedback information based on the prediction information. For example, the UE may determine the second feedback information based on a change in performance after the prediction information is executed. For example, in a mobility enhancement scenario, the prediction information is a target cell to which the UE should be handed over, a handover time, etc. The UE may perform corresponding operations based on the prediction information, and determine the second feedback information based on a change in performance after the prediction information is performed. Alternatively, the UE compares the prediction information with the actual information to determine the second feedback information. For example, in a trajectory prediction scenario, the UE may compare the predicted movement trajectory of the UE with the actual movement trajectory of the UE to determine the second feedback information. The UE transmits the second feedback information to the first RAN node. The first RAN node transmits the first feedback information to the OAM based on the received second feedback information, and the OAM updates the first model based on the first feedback information, etc. For details, please refer to the following description of the first design in steps 705 and 706.Alternatively, when determining that the application scenario changes or the performance level (or performance indicator) of the required model changes, the first RAN node may alternatively send to the OAM first feedback information indicating that the scenario changes or the performance level changes. The OAM may reconfigure the model for the first RAN node based on the scenario change instruction or the performance level change instruction. The model may be called a second model, etc. For example, when the UE changes from a walking scenario to a scenario in which the vehicle drives on a fixed route, the first RAN node predicts the movement trajectory in the scenario in which the vehicle drives on a fixed route by using the model in the trajectory prediction required in the walking scenario, which may result in a decrease in prediction accuracy. In the present application, upon finding that the UE's scenario has changed, the first RAN node may instruct the scenario change to the OAM, etc., and the OAM may configure a model for the first RAN node that fits the current scenario, etc. The above scenario change instruction or performance level change instruction may be discovered by the first RAN node by analyzing the UE, or may be reported by the UE to the first RAN node. For details, see the description of the second design in step 706. Alternatively, the first model is a model related to a first RAN node, and the first RAN may determine that application scenarios change, performance levels change, etc. by analyzing the first RAN.

[0118] Step 705: The first RAN node sends the prediction information to the UE. Correspondingly, the UE receives the prediction information from the first RAN node.

[0119] Step 706: The UE sends the second feedback information to the first RAN node. Correspondingly, the first RAN node receives the second feedback information from the UE.

[0120] In the first design, upon receiving the prediction information, the UE may compare the prediction information transmitted by the first RAN node with the actual information to determine an accuracy of the prediction information. The accuracy is referred to as an actual accuracy of the prediction information. For example, the prediction information is trajectory information. The UE may compare the predicted trajectory information with the actual trajectory information of the UE to determine an actual accuracy of the prediction information, and transmit second feedback information to the first RAN node, where the second feedback information indicates the actual accuracy of the prediction information. The indication may be an implicit indication, an explicit indication, etc. In other words, the UE transmits the actual accuracy of the prediction information to the first RAN node. For example, the UE may feedback the actual accuracy of the prediction information within a certain period to the first RAN node, where the feedback may be a periodic feedback, an event-triggered feedback, etc. For example, the UE may use 10 seconds as a period, and calculate the actual accuracy of the prediction information every 10 seconds, and feed back the actual accuracy to the first RAN node. Alternatively, the UE also calculates the accuracy of the prediction information every 10 seconds. If the accuracy is less than a threshold (eg, 80%), the UE transmits the actual accuracy of the prediction information to the first RAN node.

[0121] In the second design, upon finding that the application scenario of the UE changes or the performance level of the UE changes, the UE may alternatively send second feedback information to the first RAN node, where the second feedback information indicates that the application scenario has changed or the performance level has transformed. In other words, the UE sends an application scenario change indication, a performance level change indication, etc. to the first RAN node.

[0122] Step 707: The first RAN node sends the first feedback information to the OAM. Correspondingly, the OAM receives the first feedback information from the first RAN node.

[0123] In design, the first RAN node may send the first feedback information to the OAM periodically, based on an event trigger, etc. The first feedback information may indicate the actual accuracy of the prediction information implicitly or explicitly. In other words, the first RAN node sends the actual accuracy of the prediction information to the OAM. Similar to step 706 above, the first RAN node may feedback the actual accuracy of the prediction information within a certain period of time to the OAM. For example, the first RAN node may feedback the actual accuracy of the prediction information within a certain period of time to the OAM periodically, based on an event trigger, etc.

[0124] In another design, when an application scenario or a performance level related to the first model changes, the first RAN node sends first feedback information to the OAM, where the first feedback information indicates that the application scenario has changed or that the performance level has changed. In other words, the first RAN node sends an application scenario change instruction, a performance level change instruction, etc. to the OAM.

[0125] Step 708: The OAM updates the first model based on the first feedback information.

[0126] In design, when the first feedback information indicates an actual accuracy of the prediction information output by the first model, the OAM updating the first model based on the first feedback information includes: if the actual accuracy of the prediction information output by the first model is below a threshold, the first model is retrained and the first model is updated. For example, the OAM may collect training data, where the training data is from the first RAN node and / or the UE, other nodes, etc. The OAM may retrain the first model by using the collected training data. For the model training process, please refer to the above description.

[0127] In another design, the first feedback information indicates a scenario change instruction, a model performance level change instruction, or a model performance indicator change instruction. The OAM updating the first model based on the first feedback information includes: a second model is determined based on the scenario change instruction, the model performance level change instruction, or the model performance indicator change instruction. For example, the first mapping relationship includes a mapping relationship between a model, a model application scenario, and a model function, and the OAM may determine a changed scenario based on the indicated scenario change instruction. A model corresponding to the changed scenario is determined based on the above first mapping relationship and is referred to as a second model. Alternatively, the first mapping relationship includes a mapping relationship between a model, a model performance level, and a model function, and the OAM may determine a changed model performance level based on the indicated model performance level change instruction. Alternatively, the OAM may determine a changed model performance level based on the indicated model performance indicator change instruction. A model corresponding to the changed model performance level is determined based on the above mapping relationship and is referred to as a second model. Alternatively, the first mapping relationship includes a mapping relationship between a model, a model performance indicator, and a model function. The OAM determines a model that corresponds to the indicated changed model performance indicator based on the first mapping relationship, where the model is referred to as the second model.

[0128] Step 709: The OAM sends information about the updated model to the first RAN node. Correspondingly, the first RAN node receives information about the updated model from the OAM.

[0129] For the contents contained in the information about the updated model, please refer to the contents contained in the information about the first model in the above step 703. The details will not be described again here.

[0130] In this application, the first RAN node may send the first feedback information to the OAM periodically, based on an event trigger, etc. The OAM evaluates whether the first model needs to be updated based on the first feedback information, and updates the first model accordingly, thereby improving the prediction performance and prediction speed of the first model.

[0131] After the first model is configured for the first RAN node by using the method of the procedure in Fig. 7, the present application provides a cell handover flow chart, as shown in Fig. 8. In this procedure, the first RAN node is called a source RAN node, the second RAN node is called a target RAN node, and the procedure includes at least the following steps:

[0132] Step 801: The source RAN node determines a target cell.

[0133] In a design, the source RAN node may receive a measurement report reported by the UE, where the measurement report includes cell identity information, cell quality information, etc. The cell may be a current serving cell, a neighboring cell, etc. This is not limited. In a design, the UE may periodically report a measurement report to the source RAN node. For example, the UE may periodically measure the signal quality of the serving cell and / or the signal quality of the neighboring cell, and then periodically report the measured information to the source RAN node. Alternatively, the UE may periodically measure the signal quality of the serving cell and / or the signal quality of the neighboring cell, and report the measured information, etc. to the source RAN node when a certain condition is met.

[0134] In the present application, the measurement report includes a cell identity. The cell identity may include at least one of the cell global identifier (CGI), physical cell identifier (PCI) and frequency, cell identity (cell ID), non-public network identifier (NPN ID), non-terrestrial network identifier (NTN ID), other cell identity, etc. The CGI may include a public land mobile network (PLMN ID), cell ID, etc. Optionally, the cell identity may further include a tracking area code (TAC) and / or an identity of a network device to which the cell belongs, e.g. a global network device identifier.

[0135] In this application, the measurement report includes the signal quality information of the cell. For example, the UE may obtain the cell quality information by measuring at least one of a downlink synchronization channel, a channel state information reference signal, a demodulation reference signal (DMRS), a cell-specific reference signal (CRS), a synchronization signal block (SSB), a synchronization signal / physical broadcast channel block, or other downlink signals. For example, the cell quality information may include at least one of a received signal code power (RSCP), a reference signal received power (RSRP), a reference signal received quality (RSRQ), a signal noise ratio (SNR), a signal to interference plus noise ratio (SINR), a reference signal strength indication (RSSI), or other signal qualities.

[0136] Optionally, the cell quality information may be at least one of a cell level, a beam level, a synchronization signal / physical broadcast channel block (SS / PBCH block) level, a channel state information reference signal (CSI-RS) level, a numerology level, a slicing level, or a bandwidth part (BWP) level. The level of the cell quality information is the granularity at which the cell quality information is measured. For example, the cell quality information is at a cell level means that the UE measures each cell among multiple measurement target cells to obtain the quality of each cell. Alternatively, the cell quality information is at a beam level means that the cell includes at least one beam, and the UE obtains the quality information of the cell by measuring the beam in the cell. For example, the cell includes three cells, and the UE may separately measure beam 1, beam 2, and beam 3 among the three beams to obtain the quality information of beam 1 and the quality information of beam 2 that satisfy the beam quality condition. Optionally, the UE may then obtain cell quality information based on the quality information of beam 1 and the quality information of beam 2, for example by taking the larger value, average, or weighted sum. Alternatively, the UE may report both the quality information of beam 1 and the quality information of beam 2. The cell qualities measured by the UE at other granularities are similar to those described above and will not be described again.

[0137] It should be noted that, when the cell quality information is at the beam level, SS / PBCH block level, CSI-RS level, air interface technology level, slicing level, BWP level, etc., the cell identity information further includes at least one of the corresponding beam identity information, SS / PBCH block identity information, CSI-RS identity information, air interface technology identity information, slice identity information, BWP identity information, etc.

[0138] In the design, when receiving the measurement report, the source RAN node may determine whether a serving cell needs to be handed over for the UE based on the measurement report. For example, if the quality of service of the source cell is lower than a certain threshold, it may be determined that the serving cell needs to be handed over for the UE. The source RAN node may select a target cell for the UE based on the received measurement report. For example, a cell whose signal quality is higher than a threshold and has the best signal quality may be selected as the target cell from the neighbor cells of the UE.

[0139] In another design, the source RAN node may determine the target cell in an AI manner. The source RAN node may predict the target cell by using an AI model or a machine learning (ML) model. Input data of the AI ​​model or the ML model includes at least one of the following: measurement reports of the UE, historical and / or real-time loads of neighboring cells, historical trajectory information of the UE, geographic coordinates of the UE at the current time, moving direction of the UE at the current time, moving speed of the UE, etc. Output data of the AI ​​model or the ML model includes at least one of the following: target cell of the UE, handover time stamp, etc.

[0140] Step 802: The source RAN node sends model information to a target RAN node corresponding to the target cell.

[0141] In the design, the source RAN node may send model information to the target RAN node, where the model information includes an index of the first model. Optionally, the model information may further include at least one of the following: application scenario of the first model, function of the first model, data that needs to be collected by the first model for model inference, etc. For example, in an AI-based mobility enhancement solution, a RAN node that performs model inference needs to request the UE's past handover success or failure cases from a RAN node that the UE previously accessed. In the present application, a process in which the source RAN node needs to inform the target RAN node of data that needs to be collected for model inference can be further described as follows: the RAN node sends first instruction information to the target RAN node, where the first instruction information indicates type information of input data of the first model. In this process, the source RAN node can inform the target RAN node of data that needs to be collected for model inference in the model inference process in advance, so that the target RAN node prepares for model inference to improve model inference efficiency.

[0142] Step 803: The source RAN node sends historical information of the UE to the target RAN node, where the historical information includes at least past data, past information, etc. collected from the UE by the source RAN node.

[0143] In design, the historical information of the UE may be information required to perform model inference by using the first model, etc. In other words, when the target RAN node performs model inference by using the first model, the historical information may be used as part of the input data of the first model. The above process may be further described as follows: the source RAN node sends past data information required for model inference to the target RAN node, and the past data information is for determining the input data of the first model. For example, the target RAN node may directly use the past data as an input of the first model to perform model inference, or the target RAN node may process the past data and use the processed past data as an input of the first model. This is not limited.

[0144] It should be noted that step 803 may be executed after step 802, or may be executed after step 804. This is not limited. In the procedure of FIG. 8, an example in which the procedure is executed after step 802 is used for explanation. For example, if the target cell is determined by an AI method, step 803 may be executed after step 802. Alternatively, if the target cell is determined by a non-AI method, step 803 may be executed after step 804. In the present application, the order of steps in all procedures is not limited.

[0145] Step 804: The UE is handed over from a source cell corresponding to a source RAN node to a target cell corresponding to a target RAN node. After the handover, the target cell becomes a serving cell of the UE.

[0146] Step 805: The target RAN node sends second indication information to the source RAN node, where the second indication information indicates at least one of the following: indicating that the UE is handed over to the target cell, instructing the source RAN node to delete the first model, or instructing the source RAN node to release corresponding computational resources.

[0147] For example, the source RAN node may allocate 1% of resources to the UE for model inference, and after the UE is handed over to the target cell, the source RAN node may release the 1% computational resources allocated to the UE, delete the first model, etc.

[0148] Step 806: The source RAN node deletes the first model and releases corresponding computing resources.

[0149] Step 807: The target RAN node sends a second model request message to the OAM, where the second model request message indicates an index of the first model.

[0150] Step 808: The OAM determines the first model based on the index of the first model.

[0151] In this application, there is a mapping relationship between the index of the first model and the first model, and the OAM can determine the first model based on the index of the first model.

[0152] Step 809: The OAM sends information about the first model to the target RAN node.

[0153] Step 8010: The target RAN node performs model inference based on the first model to obtain prediction information.

[0154] Step 8011: The target RAN node sends prediction information to the UE.

[0155] Step 8012: The UE sends second feedback information to the target RAN node.

[0156] Step 8013: The target RAN node sends first feedback information to the OAM.

[0157] Step 8014: The OAM updates the first model.

[0158] Step 8015: The OAM sends information about the updated first model to the target RAN node.

[0159] For the specific execution process of steps 809 to 8015, please refer to the explanation in FIG.

[0160] In this application, when the UE performs cell handover, the source RAN node sends the index of the first model to the target RAN node to ensure the continuity of the AI ​​service. Furthermore, after the UE successfully hands over, the target RAN node notifies the source RAN node to delete model-related information to release computational resources, so that the memory resource occupation and computational resource occupation of the RAN node can be reduced.

[0161] It can be understood that to implement the functions in the above methods, the OAM and RAN nodes include corresponding hardware structures and / or software modules for performing the functions. Those skilled in the art should easily realize that the present application can be implemented by hardware or a combination of hardware and computer software with reference to the units and method steps in the examples described in this application. Whether the functions are performed by hardware or hardware driven by computer software depends on the specific application scenario and design constraints of the technical solution.

[0162] 9 and 10 are diagrams of possible device structures according to the present application, respectively. These communication devices may be configured to implement the functions of OAM or RAN nodes in the above method, and thus can also realize the advantageous effects of the above method.

[0163] As shown in Fig. 9, the communication device 900 includes a processing unit 910 and a transceiver unit 920. The communication device 900 is configured to implement the functionality of OAM, a source RAN node, or a target RAN node in the method shown in Fig. 7 or 8.

[0164] When the communication device 900 is configured to implement the function of OAM in the method shown in Fig. 7 or 8, the transceiver unit 920 is configured to receive a first model request message from a first access network device. The processing unit 910 is configured to determine a first model based on the first model request message and a first mapping relationship, where the first mapping relationship includes a mapping relationship between the model, the model application scenario, and the model function, or the first mapping relationship includes a mapping relationship between the model, the model performance level, and the model function. The transceiver unit 920 is further configured to send information about the first model to the first access network device.

[0165] When the communication device 900 is configured to implement the functionality of the first RAN node or the source RAN node in the method shown in Figure 7 or Figure 8, the transceiver unit 920 is configured to send a first model request message to the Operations, Administration and Maintenance OAM, where the first model request message indicates a model application scenario and model capabilities of the first model, or the first model request message indicates model capabilities and model performance levels of the first model. The transceiver unit is further configured to receive information regarding the first model from the OAM.

[0166] When the communications apparatus 900 is configured to implement the functionality of a target RAN node in the method illustrated in Figure 8, the transceiver unit 920 is configured to receive an index of the first model from the first access network device and send a second model request message to the Operations, Administration and Maintenance OAM, where the second model request message indicates the index of the first model and receives information regarding the first model from the OAM.

[0167] For a more detailed description of the processing unit 910 and the transceiver unit 920, please directly refer to the relevant description of the method shown in Figure 5, Figure 7 or Figure 8. The details will not be described again here.

[0168] 10, the communication device 1000 includes a processor 1010 and an interface circuit 1020. The processor 1010 and the interface circuit 1020 are coupled to each other. It may be understood that the interface circuit 1020 may be a transceiver, an input / output interface, a pin, etc. Optionally, the communication device 1000 may further include a memory 1030 configured to store instructions to be executed by the processor 1010, or to store input data required for execution of the instructions by the processor 1010, or to store data generated after the processor 1010 executes the instructions.

[0169] When the communications device 1000 is configured to implement the above method, the processor 1010 is configured to implement the functionality of the processing unit 910 and the interface circuit 1020 is configured to implement the functionality of the transceiver unit 920.

[0170] If the communication device is a module used in OAM, the module in OAM implements the function of OAM in the above method. The module in OAM receives information from other modules in OAM (e.g., radio frequency module or antenna), where the information is sent by the RAN node to OAM, or sends information to other modules in OAM (e.g., radio frequency module or antenna), where the information is sent by OAM to the RAN node. The module in OAM here may be a baseband chip of OAM, or may be another module, etc.

[0171] When the communication device is a module used in a RAN node, the module in the RAN node implements the function of the RAN node in the above method. The module in the RAN node receives information from other modules (e.g., radio frequency modules or antennas) in the RAN node, where the information is sent by a terminal to the RAN node, or sends information to other modules (e.g., radio frequency modules or antennas) in the RAN node, where the information is sent by the RAN node to the terminal. The module in the RAN node here may be a baseband chip of the RAN node, or may be a DU or other module, etc. The DU here may be a DU in an open radio access network (O-RAN) architecture.

[0172] The processor in this application may be a central processing unit (CPU) or other general purpose processor, digital signal processor (DSP), application-specific integrated circuit (ASIC), field programmable gate array (FPGA), or other programmable logic device, transistor logic device, hardware component, or any combination thereof. The general purpose processor may be a microprocessor or any conventional processor.

[0173] The memory of the present application may be random access memory, flash memory, read-only memory, programmable read-only memory, erasable programmable read-only memory, electrically erasable programmable read-only memory, registers, hard disk drive, removable hard disk drive, CD-ROM, or any other form of storage medium well known in the art.

[0174] For example, a storage medium may be coupled to the processor such that the processor can read information from, and write information to, the storage medium. The storage medium may alternatively be components of the processor. The processor and the storage medium may reside in an ASIC. Further, the ASIC may reside in a base station or terminal. Indeed, the processor and the storage medium may reside as discrete components in a base station or terminal.

[0175] Part or all of the method of the present application may be implemented by software, hardware, firmware, or any combination thereof. When software is used to implement the method, part or all of the method may be implemented in the form of a computer program product. The computer program product includes one or more computer programs or instructions. When the computer programs or instructions are loaded and executed on a computer, all or part of the procedures or functions according to the present application are executed. The computer may be a general-purpose computer, a special-purpose computer, a computer network, a network device, a user equipment, a core network device, an OAM, or other programmable device. The computer program or instructions may be stored in a computer-readable storage medium or may be transmitted from a computer-readable storage medium to another computer-readable storage medium. For example, the computer program or instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center in a wired or wireless manner. The computer-readable storage medium may be any useful medium that can be accessed by a computer, or a data storage device, such as a server or a data center incorporating one or more useful media. The useful medium may be a magnetic medium, such as a floppy disk, a hard disk, or a magnetic tape, or an optical medium, such as a digital video disk, or a semiconductor medium, such as a solid-state drive. The computer-readable storage medium may be a volatile or non-volatile storage medium, or may include two types of storage media, i.e., a volatile storage medium and a non-volatile storage medium.

[0176] In this application, unless otherwise stated or there is no logical contradiction, the terms and / or descriptions in different embodiments may be consistent and cross-referenced, and the technical features in different embodiments may be combined based on their internal logical relationships to form a new embodiment.

[0177] In this application, "at least one" means one or more, and "plurality" means two or more. "And / or" describes an association relationship between related objects, and indicates that three relationships may exist. For example, A and / or B can indicate the following: A exists alone, A and B both exist, and B exists alone, and A and B may be singular or plural. In the text description of this application, the " / " character generally indicates an "or" relationship between related objects. In formulas in this application, the " / " character indicates a "division" relationship between related objects. "Comprising at least one of A, B, or C" may indicate including A, including B, including C, including A and B, including A and C, including B and C, or including A, B, and C.

[0178] It can be understood that various numbers used in this application are only distinguished for ease of description, and are not intended to limit the scope of this application. The sequential numbers of the above processes do not imply an execution order, and the execution order of the processes should be determined based on the functions and internal logic of the processes.

[0179] [CROSS REFERENCE TO RELATED APPLICATIONS] This application claims priority to Chinese Patent Application No. 202210096175.1, entitled "MODEL CONFIGURATION METHOD AND APPARATUS," filed with the State Intellectual Property Office of the People's Republic of China on January 26, 2022, the entire contents of which are incorporated herein by reference.

Claims

1. A model construction method, comprising: receiving a first model request message from a first access network device; determining a first model based on the first model request message and a first mapping relationship, wherein the first mapping relationship includes a mapping relationship between a model, a model application scenario, and a model function, or the first mapping relationship includes a mapping relationship between a model, a model performance level, and a model function; sending information regarding the first model to the first access network device; The method according to claim 1,

2. The first model request message indicates a model application scenario and a model function of the first model. The method of claim 1.

3. the first model request message indicates a model capability and a model performance level of the first model, or indicates a model capability and a model performance indicator of the first model; The method of claim 1.

4. The information about the first model includes the following about the first model: model index, model structure information, model parameters, model input format, model output format, model performance indicators, model application scenarios, model functions, or training parameters Indicate at least one of:

4. The method according to any one of claims 1 to 3.

5. receiving first feedback information from the first access network device; updating the first model based on the first feedback information; The method of any one of claims 1 to 4, further comprising:

6. The first feedback information indicates an actual accuracy of prediction information output by the first model, and updating the first model based on the feedback information includes: retraining the first model and updating the first model if the actual accuracy of the prediction information output by the first model is below a threshold. The method according to claim 5.

7. The first feedback information indicates a scenario change instruction, a model performance level change instruction, or a model performance indicator change instruction, and updating the first model based on the first feedback information includes: selecting a second model for the terminal device based on the scenario change instruction, the model performance level change instruction, or the model performance indicator change instruction; sending information regarding the second model to the first access network device; wherein the second model is for updating the first model. The method according to claim 5.

8. receiving a second model request message from a second access network device, the second model request message indicating an index of the first model; determining the first model based on the index of the first model; sending said information regarding said first model to said second access network device; The method of any one of claims 1 to 7, further comprising:

9. A model construction method, comprising: sending a first model request message to an operations, administration and maintenance (OAM), the first model request message indicating a model application scenario and a model capability of the first model, or the first model request message indicating a model capability and a model performance level of the first model, or indicating a model capability and a model performance indicator of the first model; receiving information regarding the first model from the OAM; The method according to claim 1,

10. Prior to sending the first model request message to an OAM, the method further comprises: and determining whether an artificial intelligence (AI) based enhanced service needs to be provided.

10. The method of claim 9.

11. The information about the first model includes the following about the first model: model index, model structure information, model parameters, model input format, model output format, model performance indicators, model application scenarios, model functions, or training parameters Indicate at least one of:

11. The method according to claim 9 or 10.

12. performing model inference based on the first model to obtain predictive information.

12. The method according to any one of claims 9 to 11.

13. and sending the prediction information including a prediction result to a terminal device. The method of claim 12.

14. The prediction information further includes an accuracy of predicting the prediction result.

14. The method according to claim 12 or 13.

15. and receiving second feedback information from the terminal device indicative of an actual accuracy of the prediction information.

15. The method according to claim 13 or 14.

16. sending first feedback information to the OAM indicative of the actual accuracy of the forecast information, an application scenario change instruction, a model performance level change instruction, or a model performance indicator change instruction.

16. The method according to any one of claims 9 to 15.

17. determining a target cell for the terminal device; sending an index of the first model to a second access network device corresponding to the target cell; 17. The method of any one of claims 9 to 16, further comprising:

18. sending a first indication to the second access network device corresponding to the target cell; The first instruction information indicates type information of the input data of the first model.

20. The method of claim 17.

19. and transmitting historical data information required for the model inference to the second access network device.

19. The method according to claim 17 or 18.

20. receiving second indication information from the second access network device, the second indication information indicating one of the following: indicating that the terminal is to be handed over to the target cell, instructing the first access network device to delete the first model, or instructing the first access network device to release corresponding computational resources; deleting the first model to free up the corresponding computational resources; 20. The method of any one of claims 17 to 19, further comprising:

21. A model construction method, comprising: receiving an index of a first model from a first access network device; sending a second model request message to an Operations, Administration and Maintenance OAM indicating the index of the first model; receiving information regarding the first model from the OAM; The method according to claim 1,

22. The information about the first model includes the following about the first model: model index, model structure information, model parameters, model input format, model output format, model performance indicators, model application scenarios, model functions, or training parameters Indicate at least one of:

22. The method of claim 21.

23. and receiving first indication information from the first access network device, the first indication information indicating type information of input data of the first model.

23. The method of claim 21 or 22.

24. and receiving historical data information necessary for model inference from the first access network device; In the model inference, the historical data information is for determining the input data of the first model.

24. A method according to any one of claims 21 to 23.

25. sending a second indication to the first access network device; The second indication information indicates one of the following: indicating that the terminal is handed over to a target cell, instructing the first access network device to delete the first model, or instructing the first access network device to release corresponding computational resources.

25. A method according to any one of claims 21 to 24.

26. A communication device comprising a unit adapted to carry out the method according to any one of claims 1 to 8.

27. A processor and a memory, the processor is coupled to the memory; The processor is configured to perform a method according to any one of claims 1 to 8. Communications equipment.

28. A communication device comprising a unit adapted to carry out a method according to any one of claims 9 to 20.

29. A processor and a memory, the processor is coupled to the memory; The processor is configured to perform a method according to any one of claims 9 to 20. Communications equipment.

30. A communication device comprising a unit adapted to carry out the method according to any one of claims 21 to 25.

31. A processor and a memory, the processor is coupled to the memory; The processor is configured to perform a method according to any one of claims 21 to 25. Communications equipment.

32. A communication device according to claim 26 or 27, A communication device according to claim 28 or 29; A communication device according to claim 30 or 31. A communication system having the above configuration.

33. Memorize the command, The instructions, when executed on a computer, are capable of causing the computer to carry out the method according to any one of claims 1 to 8, the method according to any one of claims 9 to 20, or the method according to any one of claims 21 to 25. A computer-readable storage medium.

34. Have orders, The instructions, when executed on a computer, are capable of causing the computer to carry out the method according to any one of claims 1 to 8, the method according to any one of claims 9 to 20, or the method according to any one of claims 21 to 25. Computer program products.

Images