Communication method and communication device

Abstract

The present application provides a communication method and a communication device for use in a scenario in which AI is integrated into a wireless network, particularly in a scenario in which an AI model for beam management is integrated into a wireless network. According to the method provided in the present application, a network device transmits reference signal resource configuration information to a terminal device, the reference signal resource configuration information including instruction information indicating the purpose of the reference signal resource, so that the terminal device can identify the purpose of the reference signal resource based on the instruction information to prevent the terminal device from using the reference signal resource for other purposes. For example, if the instruction information indicates that the reference signal resource is for performing a first operation based on an AI model, the terminal device can prevent the terminal device from using the reference signal resource for a non-AI model-based operation. This ensures the performance of the AI ​​model and avoids resource waste.

Description

[Technical Field]

[0001] This application claims priority to Chinese Patent Application No. 202310156061.6, entitled "COMMUNICATION METHOD AND COMMUNICATION APPARATUS," filed with the State Intellectual Property Office of the People's Republic of China on February 16, 2023, which is incorporated herein by reference in its entirety.

[0002] The present application relates to the field of communication technologies, and more particularly to communication methods and devices. [Background technology]

[0003] Currently, artificial intelligence (AI) has been introduced into wireless communication networks and has been widely applied in various air interface technology application scenarios, such as AI-based channel state information (CSI) prediction, AI-based beam management, and AI-based CSI feedback, and is playing an increasingly important role.

[0004] In the beam management process using an AI model, the AI ​​model may be deployed on a training device for training and updating. After the AI ​​model training is completed, the training device can infer optimal beams based on the trained AI model. However, how to balance the performance and resource consumption of training an AI model and / or inferring based on the trained AI model remains an open question. Summary of the Invention [Means for solving the problem]

[0005] The present application provides a communication method and a communication device for balancing performance and resource consumption for training an AI model and / or for balancing performance and resource consumption for performing inference on an AI model.

[0006] According to a first aspect, there is provided a communication method. The method may be performed by a terminal device or may be performed by a component (e.g., a chip or circuit) of the terminal device. This is not limited to the present specification.

[0007] The method includes, by a terminal device, receiving first reference signal resource configuration information from a network device, the first reference signal resource configuration information including first indication information, wherein the first indication information indicates that a first reference signal resource configured based on the first reference signal resource configuration information is for performing a first operation based on an AI model, or the first indication information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for acquiring a dataset for the first operation based on the AI ​​model.

[0008] According to the technical solution, the terminal device can identify the purpose of the first reference signal resource based on the first indication information, which can prevent the terminal device from performing a non-AI model-based operation based on the first reference signal resource, thereby helping to balance the performance and resource consumption of performing model training on the AI ​​model and / or the performance and resource consumption of performing model inference based on the AI ​​model.

[0009] For example, the first operation is model training or model inference for an AI model.

[0010] For example, the AI ​​model is for beam management.

[0011] For example, the first reference signal resource configuration information may be carried in one or more of a CSI reporting configuration (CSI-ReportConfig) field, a CSI resource configuration (CSI-ResourceConfig) field, a CSI-RS resource set (ResourceSet) field, or a CSI-RS resource set list (ResourceSetList) field.

[0012] Referring to the first aspect, in some implementation forms of the first aspect, the method further includes the terminal device performing a first operation based on the AI ​​model and the first reference signal resource configuration information.

[0013] Based on the technical solution, when an AI model is deployed in a terminal device, the terminal device can perform a first operation based on the AI ​​model and the first reference signal resource configuration information.

[0014] Referring to the first aspect, in some implementation forms of the first aspect, the first operation is model inference for the AI ​​model, and the method further includes, by the terminal device, receiving second reference signal resource configuration information from the network device. The second reference signal resource configuration information includes second instruction information, wherein the second instruction information indicates that the second reference signal resource configured based on the second reference signal resource configuration information is for performing model training for the AI ​​model, or the second instruction information indicates that the second reference signal resource configured based on the second reference signal resource configuration information is for acquiring a dataset for performing model training for the AI ​​model.

[0015] According to the technical solution, the terminal device can determine, based on the second instruction information, that the second reference signal resource is for performing model training on the AI ​​model, which can prevent the terminal device from performing model inference on the AI ​​model based on the second reference signal resource, or prevent the terminal device from performing non-AI model-based operations based on the second reference signal resource, thereby helping to balance the performance of performing model training and model inference on the AI ​​model and resource consumption.

[0016] Referring to the first aspect, in some implementation forms of the first aspect, the first reference signal resource configuration information further includes third instruction information indicating an association relationship between the first reference signal resource and the second reference signal resource, or the second reference signal resource configuration information further includes fourth instruction information indicating an association relationship between the second reference signal resource and the first reference signal resource.

[0017] According to the technical solution, the terminal device can determine an association relationship between the first reference signal resource and the second reference signal resource based on the third indication information or the fourth indication information. In this way, the terminal device can determine that the first reference signal resource and the second reference signal resource correspond to the same AI model. Furthermore, the terminal device can determine that the first reference signal resource and the second reference signal resource are for processing the same AI model.

[0018] For example, the third instruction information indicates the association relationship between the first reference signal resource and the second reference signal resource in one or more of the following ways: the third instruction information indicates an identifier of a second reference signal resource having an association relationship with the first reference signal resource; the third instruction information indicates an identifier of reference signal resource set #2 having an association relationship with the first reference signal resource, and reference signal resource set #2 includes a second reference signal resource having an association relationship with the first reference signal resource; the third instruction information indicates an identifier of second reference signal resource configuration information having an association relationship with the first reference signal resource; or the third instruction information indicates an identifier of an AI model associated with the first reference signal resource, and is for the AI ​​model, the identifier indicated by the third instruction information is for the AI ​​model associated with the second reference signal resource and is the same as the identifier indicated by the fourth instruction information, and the fourth instruction information is included in the second reference signal resource configuration information.

[0019] For example, the fourth instruction information indicates the association relationship between the first reference signal resource and the second reference signal resource in one or more of the following manners: the fourth instruction information indicates an identifier of a first reference signal resource that has an association relationship with the second reference signal resource; the fourth instruction information indicates an identifier of reference signal resource set #1 that has an association relationship with the second reference signal resource, and reference signal resource set #1 includes the first reference signal resource that has an association relationship with the second reference signal resource; the fourth instruction information indicates an identifier of first reference signal resource configuration information that has an association relationship with the second reference signal resource; or the fourth instruction information indicates an identifier of an AI model associated with the second reference signal resource, and is for the AI ​​model, and the identifier indicated by the fourth instruction information is for the AI ​​model associated with the first reference signal resource and is the same as the identifier indicated by the third instruction information, and the third instruction information is included in the first reference signal resource configuration information.

[0020] Referring to the first aspect, in some implementation forms of the first aspect, the first operation is model inference for an AI model, the first reference signal resource is a reference signal resource corresponding to input information for model inference of the AI ​​model, and the first reference signal resource configuration information further includes first identifier information, the first identifier information corresponds to a target reference signal resource set, and the target reference signal resource set includes reference signal resources corresponding to labels in a model training process of the AI ​​model.

[0021] For example, a reference signal resource corresponding to a label in the model training process of an AI model is denoted as reference signal resource #1. The correspondence between the label in the model training process of an AI model and reference signal resource #1 can be described as follows: reference signal resource #1 is for obtaining the label in the model training process of the AI ​​model. When the label in the model training process of an AI model corresponds to reference signal resource #1, the terminal device can determine the AI ​​model based on reference signal resource #1 or determine the AI ​​model based on a target reference signal resource set.

[0022] Based on the technical solution, when the first reference signal resource configuration information includes first identifier information, the terminal device can determine a target reference signal resource set based on the first identifier information, and then determine an AI model based on the target reference signal resource set. In this way, the terminal device can determine that the first reference signal resource is for performing a first operation on the determined AI model.

[0023] For example, the first identifier information includes a sequence number of a target reference signal resource set in a first target reference signal resource set group, the first target reference signal resource set group being a reference signal resource set group configured for multiple cells or multiple network devices, the multiple network devices including a network device accessed by a terminal device, and the multiple cells including a cell in the network device accessed by the terminal device in which the terminal device executes an AI model.

[0024] Referring to the first aspect, in some implementation forms of the first aspect, the method further includes: receiving, by the terminal device, fifth reference signal resource configuration information from the network device, the fifth reference signal resource configuration information for configuring a target reference signal resource set, the fifth reference signal resource configuration information including first identifier information, and the fifth reference signal resource configuration information further including second identifier information, the second identifier information corresponding to the target reference signal resource set.

[0025] Based on the technical solution, when both the first reference signal resource configuration information and the fifth reference signal resource configuration information include first identifier information, the terminal device can determine an association relationship between the first reference signal resource configuration information and the fifth reference signal resource configuration information based on the first identifier information. Then, the terminal device can determine that the target reference signal resource set includes reference signal resources corresponding to output information of inference of the AI ​​model based on the second identifier information included in the fifth reference signal resource configuration information. Finally, the terminal device can determine an AI model corresponding to the target reference signal resource set based on the target reference signal resource set, and determine that the first reference signal resource is for performing model inference on the determined AI model.

[0026] For example, the second identifier information includes a sequence number of a target reference signal resource set in a first target reference signal resource set group, the first target reference signal resource set group being a reference signal resource set group configured for a plurality of cells or a plurality of network devices, the plurality of network devices including network devices accessed by a terminal device, and the plurality of cells including cells in the network devices accessed by the terminal device where the terminal device executes an AI model. The first identifier information includes a sequence number of a target reference signal resource set in a second target reference signal resource set group, the second target reference signal resource set group being a reference signal resource set group configured for a cell where the terminal device executes an AI model.

[0027] Referring to the first aspect, in some implementation forms of the first aspect, the first operation is model training for an AI model, and the first instruction information further indicates that the first reference signal resource is for obtaining labels for performing model training for the AI ​​model.

[0028] Based on the technical solution, when the resource for obtaining labels for performing model training on the AI ​​model and the resource for obtaining input information for performing model training on the AI ​​model are different resources, the terminal device can determine, based on the first instruction information, that the first reference signal resource is for obtaining labels for performing model training on the AI ​​model, and prevent the terminal device from using the first reference signal resource for another purpose.

[0029] Referring to the first aspect, in some implementation forms of the first aspect, the first reference signal resource configuration information further includes fifth instruction information, and the fifth instruction information indicates that the first reference signal resource is for obtaining labels for performing model training on an AI model.

[0030] Based on the technical solution, when the resource for obtaining labels for performing model training on the AI ​​model and the resource for obtaining input information for performing model training on the AI ​​model are different resources, the terminal device can determine, based on the fifth instruction information, that the first reference signal resource is for obtaining labels for performing model training on the AI ​​model, and prevent the terminal device from using the first reference signal resource for another purpose.

[0031] Referring to the first aspect, in some implementation forms of the first aspect, the method further includes: receiving, by the terminal device, third reference signal resource configuration information from the network device, the third reference signal resource configuration information including sixth instruction information, wherein the sixth instruction information indicates that the third reference signal resource configured based on the third reference signal resource configuration information is for obtaining input information for performing model training on the AI ​​model.

[0032] Referring to the first aspect, in some implementation forms of the first aspect, the first reference signal resource configuration information further includes seventh instruction information indicating an association relationship between the first reference signal resource and the third reference signal resource, or the third reference signal resource configuration information further includes eighth instruction information indicating an association relationship between the first reference signal resource and the third reference signal resource.

[0033] According to the technical solution, the terminal device can determine an association relationship between the first reference signal resource and the third reference signal resource based on the seventh or eighth indication information. In this way, the terminal device can determine that the first reference signal resource and the third reference signal resource correspond to the same AI model, and the terminal device can determine that the first reference signal resource and the third reference signal resource are for performing model training for the same AI model.

[0034] Referring to the first aspect, in some implementation forms of the first aspect, the method further includes the terminal device obtaining input information and a label based on an association relationship between the first reference signal resource and the third reference signal resource.

[0035] For example, the seventh instruction information indicates the association relationship between the first reference signal resource and the third reference signal resource in one or more of the following manners: the seventh instruction information indicates a time offset value of the first reference signal resource relative to the third reference signal resource; the seventh instruction information indicates an identifier of the third reference signal resource having an association relationship with the first reference signal resource; the seventh instruction information indicates an identifier of reference signal resource set #3 having an association relationship with the first reference signal resource, and reference signal resource set #3 includes the third reference signal resource having an association relationship with the first reference signal resource; the seventh instruction information indicates an identifier of third reference signal resource configuration information having an association relationship with the first reference signal resource; or the seventh instruction information indicates an identifier of an AI model associated with the first reference signal resource, and is for the AI ​​model, the identifier indicated by the seventh instruction information is for the AI ​​model associated with the third reference signal resource and is the same as the identifier indicated by the eighth instruction information, and the eighth instruction information is included in the third reference signal resource configuration information.

[0036] For example, the eighth instruction information indicates the association relationship between the first reference signal resource and the third reference signal resource in one or more of the following manners: the eighth instruction information indicates a time offset value of the third reference signal resource relative to the first reference signal resource; the eighth instruction information indicates an identifier of the first reference signal resource having an association relationship with the third reference signal resource; the eighth instruction information indicates an identifier of reference signal resource set #1 having an association relationship with the third reference signal resource, and reference signal resource set #1 includes the first reference signal resource having an association relationship with the third reference signal resource; the eighth instruction information indicates an identifier of first reference signal resource configuration information having an association relationship with the third reference signal resource; or the eighth instruction information indicates an identifier of an AI model associated with the third reference signal resource, and is for the AI ​​model, the identifier indicated by the eighth instruction information is for the AI ​​model associated with the first reference signal resource and is the same as the identifier indicated by the seventh instruction information, and the seventh instruction information is included in the first reference signal resource configuration information.

[0037] For example, the third reference signal resource corresponds to a synchronization signal and PBCH block (SSB) resource, and the first reference signal resource corresponds to a channel state information reference signal (CSI-RS) resource.

[0038] Referring to the first aspect, in some implementation forms of the first aspect, the first instruction information indicating that the first reference signal resource configured based on the first reference signal resource configuration information is for performing a first operation based on an AI model includes the first instruction information indicating that the first operation based on the AI ​​model is in an enabled state.

[0039] Referring to the first aspect, in some implementation forms of the first aspect, the method further includes: receiving, by the terminal device, fourth reference signal resource configuration information from the network device, wherein the fourth reference signal resource configuration information includes ninth indication information, and the ninth indication information indicates that the fourth reference signal resource configured based on the fourth reference signal resource configuration information is for performing a non-AI model-based operation.

[0040] Referring to the first aspect, in some implementation forms of the first aspect, the first operation is model training for an AI model, and the first instruction information further indicates that the first reference signal resource is for obtaining input information for performing model training for the AI ​​model.

[0041] Based on the technical solution, when the resource for obtaining labels to perform model training on the AI ​​model and the resource for obtaining input information to perform model training on the AI ​​model are different resources, the terminal device can determine, based on the first instruction information, that the first reference signal resource is for obtaining input information to perform model training on the AI ​​model, and prevent the terminal device from using the first reference signal resource for another purpose.

[0042] Referring to the first aspect, in some implementation forms of the first aspect, the first operation is model inference for an AI model, and the first instruction information further indicates that the first reference signal resource is for obtaining input information for performing model inference for the AI ​​model.

[0043] Based on this technical solution, the terminal device can determine, based on the first instruction information, that the first reference signal resource is for obtaining input information for performing model inference on an AI model, which can prevent the terminal device from performing model training on the AI ​​model based on the first reference signal resource or prevent the terminal device from performing a non-AI model-based operation based on the first reference signal resource, thereby helping to balance the performance of performing model training and model inference on the AI ​​model and resource consumption.

[0044] Referring to the first aspect, in some implementation forms of the first aspect, the method further includes the terminal device transmitting first information to the network device, the first information being for requesting a first action based on the AI ​​model.

[0045] For example, if the first operation is model training for an AI model, the first information may include one or more of an identifier of the AI ​​model, computing capability information of the terminal device, sparse beam instruction information, or a training mechanism of the AI ​​model.

[0046] For example, if the first operation is model inference for an AI model, the first information may include one or more of an identifier of the AI ​​model, instruction information #1, or sparse beam instruction information. The instruction information #1 indicates that training of the AI ​​model has been completed.

[0047] According to a second aspect, there is provided a communication method, which may be performed by a terminal device or may be performed by a component (e.g., a chip or circuit) of the terminal device, and this is not limited thereto.

[0048] The method includes, by a terminal device, receiving first reference signal resource configuration information from a network device, the first reference signal resource configuration information including first indication information, wherein the first indication information indicates that a first reference signal resource configured based on the first reference signal resource configuration information is for performing model training on an AI model, or the first indication information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for acquiring a dataset for performing model training on the AI ​​model.

[0049] According to the technical solution, the terminal device may identify the purpose of the first reference signal resource based on the first indication information, which can prevent the terminal device from performing a non-AI model-based operation based on the first reference signal resource, or can prevent the terminal device from performing model inference on an AI model based on the first reference signal resource, thereby helping to balance the performance and resource consumption of performing model training and model inference on an AI model.

[0050] For example, the AI ​​model is for beam management.

[0051] For example, the first reference signal resource configuration information may be carried in one or more of the following fields: a CSI reporting configuration field, a CSI resource configuration field, a CSI-RS resource set field, or a CSI-RS resource set list field.

[0052] Referring to the second aspect, in some implementation forms of the second aspect, the method further includes the terminal device performing model training on the AI ​​model based on the first reference signal resource configuration information.

[0053] Based on the technical solution, when an AI model is deployed in a terminal device, the terminal device can perform model training on the AI ​​model based on the first reference signal resource configuration information.

[0054] Referring to the second aspect, in some implementation forms of the second aspect, the first instruction information further indicates that the first reference signal resource is for obtaining labels for performing model training on the AI ​​model.

[0055] Based on the technical solution, when the resource for obtaining labels for performing model training on the AI ​​model and the resource for obtaining input information for performing model training on the AI ​​model are different resources, the terminal device can determine, based on the first instruction information, that the first reference signal resource is for obtaining labels for performing model training on the AI ​​model, and prevent the terminal device from using the first reference signal resource for another purpose.

[0056] Referring to the second aspect, in some implementation forms of the second aspect, the first reference signal resource configuration information further includes fifth instruction information, and the fifth instruction information indicates that the first reference signal resource is for obtaining labels for performing model training on an AI model.

[0057] Based on the technical solution, when the resource for obtaining labels for performing model training on the AI ​​model and the resource for obtaining input information for performing model training on the AI ​​model are different resources, the terminal device can determine, based on the fifth instruction information, that the first reference signal resource is for obtaining labels for performing model training on the AI ​​model, and prevent the terminal device from using the first reference signal resource for another purpose.

[0058] Referring to the second aspect, in some implementation forms of the second aspect, the method further includes: receiving, by the terminal device, third reference signal resource configuration information from the network device, the third reference signal resource configuration information including sixth instruction information, wherein the sixth instruction information indicates that the third reference signal resource configured based on the third reference signal resource configuration information is for obtaining input information for performing model training on the AI ​​model.

[0059] Referring to the second aspect, in some implementation forms of the second aspect, the first reference signal resource configuration information further includes seventh instruction information indicating an association relationship between the first reference signal resource and the third reference signal resource, or the third reference signal resource configuration information further includes eighth instruction information indicating an association relationship between the first reference signal resource and the third reference signal resource.

[0060] According to the technical solution, the terminal device can determine an association relationship between the first reference signal resource and the third reference signal resource based on the seventh or eighth indication information, and can determine that the first reference signal resource and the third reference signal resource correspond to the same AI model.

[0061] Referring to the second aspect, in some implementation forms of the second aspect, the method further includes the terminal device obtaining input information and a label based on an association relationship between the first reference signal resource and the third reference signal resource.

[0062] For example, the seventh instruction information indicates the association relationship between the first reference signal resource and the third reference signal resource in one or more of the following ways: the seventh instruction information indicates a time offset value of the first reference signal resource relative to the third reference signal resource; the seventh instruction information indicates an identifier of the third reference signal resource having an association relationship with the first reference signal resource; the seventh instruction information indicates an identifier of reference signal resource set #3 having an association relationship with the first reference signal resource, and reference signal resource set #3 includes the third reference signal resource having an association relationship with the first reference signal resource; the seventh instruction information indicates an identifier of third reference signal resource configuration information having an association relationship with the first reference signal resource; or the seventh instruction information indicates an identifier of an AI model associated with the first reference signal resource, and is for the AI ​​model, the identifier indicated by the seventh instruction information is for the AI ​​model associated with the third reference signal resource and is the same as the identifier indicated by the eighth instruction information, and the eighth instruction information is included in the third reference signal resource configuration information.

[0063] For example, the eighth instruction information indicates the association relationship between the first reference signal resource and the third reference signal resource in one or more of the following manners: the eighth instruction information indicates a time offset value of the third reference signal resource relative to the first reference signal resource; the eighth instruction information indicates an identifier of the first reference signal resource having an association relationship with the third reference signal resource; the eighth instruction information indicates an identifier of reference signal resource set #1 having an association relationship with the third reference signal resource, and reference signal resource set #1 includes the first reference signal resource having an association relationship with the third reference signal resource; the eighth instruction information indicates an identifier of first reference signal resource configuration information having an association relationship with the third reference signal resource; or the eighth instruction information indicates an identifier of an AI model associated with the third reference signal resource, and is for the AI ​​model, the identifier indicated by the eighth instruction information is for the AI ​​model associated with the first reference signal resource and is the same as the identifier indicated by the seventh instruction information, and the seventh instruction information is included in the first reference signal resource configuration information.

[0064] For example, the third reference signal resource corresponds to an SSB resource, and the first reference signal resource corresponds to a CSI-RS resource.

[0065] Referring to the second aspect, in some implementation forms of the second aspect, the method further includes the terminal device transmitting first information to the network device, the first information being for requesting to perform model training based on the AI ​​model, the first information including one or more of an identifier of the AI ​​model, computing capability information of the terminal device, sparse beam instruction information, or a training mechanism of the AI ​​model.

[0066] According to a third aspect, there is provided a communication method. The method may be performed by a terminal device or may be performed by a component (e.g., a chip or circuit) of the terminal device. This is not limited to the present specification.

[0067] The method includes, by a terminal device, receiving first reference signal resource configuration information from a network device, the first reference signal resource configuration information including first indication information, wherein the first indication information indicates that a first reference signal resource configured based on the first reference signal resource configuration information is for performing model inference on an AI model, or the first indication information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for acquiring a dataset for performing model inference on an AI model.

[0068] According to a technical solution, the terminal device may identify the purpose of the first reference signal resource based on the first indication information, which can prevent the terminal device from performing a non-AI model-based operation based on the first reference signal resource, or can prevent the terminal device from performing model training for an AI model based on the first reference signal resource, thereby helping to balance the performance of performing model inference for an AI model and resource consumption.

[0069] For example, the AI ​​model is for beam management.

[0070] For example, the first reference signal resource configuration information may be carried in one or more of the following fields: a CSI reporting configuration field, a CSI resource configuration field, a CSI-RS resource set field, or a CSI-RS resource set list field.

[0071] Referring to the third aspect, in some implementation forms of the third aspect, the method further includes the terminal device performing model inference based on the AI ​​model and the first reference signal resource configuration information.

[0072] Based on the technical solution, when an AI model is deployed in a terminal device, the terminal device can perform model inference based on the AI ​​model and the first reference signal resource configuration information.

[0073] Referring to a third aspect, in some implementation forms of the third aspect, the method further includes: receiving, by the terminal device, second reference signal resource configuration information from the network device, wherein the second reference signal resource configuration information includes second instruction information, wherein the second instruction information indicates that the second reference signal resource configured based on the second reference signal resource configuration information is for performing model training on the AI ​​model, or the second instruction information indicates that the second reference signal resource configured based on the second reference signal resource configuration information is for acquiring a dataset for performing model training on the AI ​​model.

[0074] According to the technical solution, the terminal device can determine, based on the second instruction information, that the second reference signal resource is for performing model training on the AI ​​model, which can prevent the terminal device from performing model inference on the AI ​​model based on the second reference signal resource, or prevent the terminal device from performing non-AI model-based operations based on the second reference signal resource, thereby helping to balance the performance of performing model training on the AI ​​model and resource consumption.

[0075] Referring to the third aspect, in some implementation forms of the third aspect, the first reference signal resource configuration information further includes third instruction information indicating an association relationship between the first reference signal resource and the second reference signal resource, or the second reference signal resource configuration information further includes fourth instruction information indicating an association relationship between the second reference signal resource and the first reference signal resource.

[0076] According to the technical solution, the terminal device can determine an association relationship between the first reference signal resource and the second reference signal resource based on the third indication information or the fourth indication information. In this way, the terminal device can determine that the first reference signal resource and the second reference signal resource correspond to the same AI model. Furthermore, the terminal device can determine that the first reference signal resource and the second reference signal resource are for processing the same AI model.

[0077] For example, the third instruction information indicates the association relationship between the first reference signal resource and the second reference signal resource in one or more of the following ways: the third instruction information indicates an identifier of a second reference signal resource having an association relationship with the first reference signal resource; the third instruction information indicates an identifier of reference signal resource set #2 having an association relationship with the first reference signal resource, and reference signal resource set #2 includes a second reference signal resource having an association relationship with the first reference signal resource; the third instruction information indicates an identifier of second reference signal resource configuration information having an association relationship with the first reference signal resource; or the third instruction information indicates an identifier of an AI model associated with the first reference signal resource, and is for the AI ​​model, the identifier indicated by the third instruction information is for the AI ​​model associated with the second reference signal resource and is the same as the identifier indicated by the fourth instruction information, and the fourth instruction information is included in the second reference signal resource configuration information.

[0078] For example, the fourth instruction information indicates the association relationship between the first reference signal resource and the second reference signal resource in one or more of the following manners: the fourth instruction information indicates an identifier of a first reference signal resource that has an association relationship with the second reference signal resource; the fourth instruction information indicates an identifier of reference signal resource set #1 that has an association relationship with the second reference signal resource, and reference signal resource set #1 includes the first reference signal resource that has an association relationship with the second reference signal resource; the fourth instruction information indicates an identifier of first reference signal resource configuration information that has an association relationship with the second reference signal resource; or the fourth instruction information indicates an identifier of an AI model associated with the second reference signal resource, and is for the AI ​​model, and the identifier indicated by the fourth instruction information is for the AI ​​model associated with the first reference signal resource and is the same as the identifier indicated by the third instruction information, and the third instruction information is included in the first reference signal resource configuration information.

[0079] Referring to the third aspect, in some implementation forms of the third aspect, the first reference signal resource is a reference signal resource corresponding to input information for model inference of the AI ​​model, and the first reference signal resource configuration information further includes first identifier information. The first identifier information corresponds to a target reference signal resource set, and the target reference signal resource set includes reference signal resources corresponding to labels in a model training process of the AI ​​model.

[0080] According to the technical solution, the terminal device can determine a target reference signal resource set based on the first identifier information, and then determine an AI model based on the target reference signal resource set. In this way, the terminal device can determine that the first reference signal resource is for performing a first operation on the determined AI model.

[0081] For example, the first identifier information includes a sequence number of a target reference signal resource set in a first target reference signal resource set group, the first target reference signal resource set group being a reference signal resource set group configured for multiple cells or multiple network devices, the multiple network devices including a network device accessed by a terminal device, and the multiple cells including a cell in the network device accessed by the terminal device in which the terminal device executes an AI model.

[0082] Referring to the third aspect, in some implementation forms of the third aspect, the method further includes: receiving, by the terminal device, fifth reference signal resource configuration information from the network device, the fifth reference signal resource configuration information for configuring a target reference signal resource set, the fifth reference signal resource configuration information including first identifier information, and the fifth reference signal resource configuration information further including second identifier information, the second identifier information corresponding to the target reference signal resource set.

[0083] Based on the technical solution, the terminal device can determine an association relationship between the first reference signal resource configuration information and the fifth reference signal resource configuration information based on the first identifier information. Then, based on the second identifier information included in the fifth reference signal resource configuration information, the terminal device can determine that the target reference signal resource set includes reference signal resources corresponding to output information of the inference of the AI ​​model. Finally, the terminal device can determine an AI model corresponding to the target reference signal resource set based on the target reference signal resource set, and determine that the first reference signal resource is for performing model inference on the determined AI model.

[0084] For example, the second identifier information includes a sequence number of a target reference signal resource set in a first target reference signal resource set group, the first target reference signal resource set group being a reference signal resource set group configured for a plurality of cells or a plurality of network devices, the plurality of network devices including network devices accessed by a terminal device, and the plurality of cells including cells in the network devices accessed by the terminal device where the terminal device executes an AI model. The first identifier information includes a sequence number of a target reference signal resource set in a second target reference signal resource set group, the second target reference signal resource set group being a reference signal resource set group configured for a cell where the terminal device executes an AI model.

[0085] Referring to the third aspect, in some implementations of the third aspect, the method further includes the terminal device transmitting first information to the network device. The first information is for requesting that model inference be performed on the AI ​​model, and the first information includes one or more of an identifier of the AI ​​model, instruction information #1, or sparse beam instruction information. The instruction information #1 indicates that training of the AI ​​model has been completed.

[0086] According to a fourth aspect, there is provided a communication method. The method may be performed by a network device or may be performed by a component (e.g., a chip or circuit) of the network device. This is not limited to the present specification.

[0087] The method includes: a network device transmitting first reference signal resource configuration information to a terminal device, the first reference signal resource configuration information including first indication information, the first indication information indicating that a first reference signal resource configured based on the first reference signal resource configuration information is for performing a first operation based on an AI model, or the first indication information indicating that the first reference signal resource configured based on the first reference signal resource configuration information is for acquiring a dataset for the first operation based on the AI ​​model.

[0088] For the beneficial effects of the fourth aspect and implementation forms of the fourth aspect, please refer to the description of the first aspect.

[0089] Referring to the fourth aspect, in some implementations of the fourth aspect, the first operation is model training or model inference for the AI ​​model.

[0090] Referring to the fourth aspect, in some implementation forms of the fourth aspect, the first operation is model inference for an AI model, and the method further includes the network device transmitting second reference signal resource configuration information to the terminal device. The second reference signal resource configuration information includes second instruction information, wherein the second instruction information indicates that the second reference signal resource configured based on the second reference signal resource configuration information is for performing model training for the AI ​​model, or the second instruction information indicates that the second reference signal resource configured based on the second reference signal resource configuration information is for acquiring a dataset for performing model training for the AI ​​model.

[0091] Referring to the fourth aspect, in some implementation forms of the fourth aspect, the first reference signal resource configuration information further includes third indication information indicating an association relationship between the first reference signal resource and the second reference signal resource, or the second reference signal resource configuration information further includes fourth indication information indicating an association relationship between the second reference signal resource and the first reference signal resource.

[0092] Referring to the fourth aspect, in some implementation forms of the fourth aspect, the first operation is model inference for an AI model, the first reference signal resource is a reference signal resource corresponding to input information for model inference of the AI ​​model, and the first reference signal resource configuration information further includes first identifier information, the first identifier information corresponds to a target reference signal resource set, and the target reference signal resource set includes reference signal resources corresponding to labels in a model training process of the AI ​​model.

[0093] For example, the first identifier information includes a sequence number of a target reference signal resource set in a first target reference signal resource set group, the first target reference signal resource set group being a reference signal resource set group configured for multiple cells or multiple network devices, the multiple network devices including a network device accessed by a terminal device, and the multiple cells including a cell in the network device accessed by the terminal device in which the terminal device executes an AI model.

[0094] Referring to the fourth aspect, in some implementation forms of the fourth aspect, the method further includes: the network device transmitting fifth reference signal resource configuration information to the terminal device, wherein the fifth reference signal resource configuration information is for configuring a target reference signal resource set, the fifth reference signal resource configuration information includes first identifier information, and the fifth reference signal resource configuration information further includes second identifier information, and the second identifier information corresponds to the target reference signal resource set.

[0095] For example, the second identifier information includes a sequence number of a target reference signal resource set in a first target reference signal resource set group, the first target reference signal resource set group being a reference signal resource set group configured for a plurality of cells or a plurality of network devices, the plurality of network devices including network devices accessed by a terminal device, and the plurality of cells including cells in the network devices accessed by the terminal device where the terminal device executes an AI model. The first identifier information includes a sequence number of a target reference signal resource set in a second target reference signal resource set group, the second target reference signal resource set group being a reference signal resource set group configured for a cell where the terminal device executes an AI model.

[0096] Referring to the fourth aspect, in some implementation forms of the fourth aspect, the first operation is model training for an AI model, and the first instruction information further indicates that the first reference signal resource is for obtaining labels for performing model training for the AI ​​model.

[0097] Referring to the fourth aspect, in some implementation forms of the fourth aspect, the first reference signal resource configuration information further includes fifth instruction information, and the fifth instruction information indicates that the first reference signal resource is for obtaining labels for performing model training on an AI model.

[0098] Referring to the fourth aspect, in some implementation forms of the fourth aspect, the method further includes: the network device transmitting third reference signal resource configuration information to the terminal device, wherein the third reference signal resource configuration information includes sixth instruction information, and the sixth instruction information indicates that the third reference signal resource configured based on the third reference signal resource configuration information is for obtaining input information for performing model training on the AI ​​model.

[0099] Referring to the fourth aspect, in some implementation forms of the fourth aspect, the first reference signal resource configuration information further includes seventh indication information indicating an association relationship between the first reference signal resource and the third reference signal resource, or the third reference signal resource configuration information further includes eighth indication information indicating an association relationship between the first reference signal resource and the third reference signal resource.

[0100] For example, the third reference signal resource corresponds to an SSB resource, and the first reference signal resource corresponds to a CSI-RS resource.

[0101] Referring to the fourth aspect, in some implementation forms of the fourth aspect, the first instruction information indicating that the first reference signal resource configured based on the first reference signal resource configuration information is for performing a first operation based on an AI model includes the first instruction information indicating that the first operation based on the AI ​​model is in an enabled state.

[0102] Referring to the fourth aspect, in some implementation forms of the fourth aspect, the method further includes: the network device transmitting fourth reference signal resource configuration information to the terminal device, wherein the fourth reference signal resource configuration information includes ninth indication information, and the ninth indication information indicates that the fourth reference signal resource configured based on the fourth reference signal resource configuration information is for performing a non-AI model-based operation.

[0103] Referring to the fourth aspect, in some implementation forms of the fourth aspect, the first operation is model training for an AI model, and the first instruction information further indicates that the first reference signal resource is for obtaining input information for performing model training for the AI ​​model.

[0104] Referring to the fourth aspect, in some implementation forms of the fourth aspect, the first operation is model inference for an AI model, and the first instruction information further indicates that the first reference signal resource is for obtaining input information for performing model inference for the AI ​​model.

[0105] Referring to the fourth aspect, in some implementations of the fourth aspect, the method further includes the network device receiving first information from the terminal device, the first information being for requesting a first action based on the AI ​​model.

[0106] For example, if the first operation is model training for an AI model, the first information may include one or more of an identifier of the AI ​​model, computing capability information of the terminal device, sparse beam instruction information, or a training mechanism of the AI ​​model.

[0107] For example, if the first operation is model inference for an AI model, the first information may include one or more of an identifier of the AI ​​model, instruction information #1, or sparse beam instruction information. The instruction information #1 indicates that training of the AI ​​model has been completed.

[0108] Referring to the fourth aspect, in some implementations of the fourth aspect, the first reference signal resource configuration information may be carried in one or more of the following fields: a CSI reporting configuration field, a CSI resource configuration field, a CSI-RS resource set field, or a CSI-RS resource set list field.

[0109] Referring to the fourth aspect, in some implementations of the fourth aspect, the AI ​​model is for beam management.

[0110] According to a fifth aspect, there is provided a communication method. The method may be performed by a network device or may be performed by a component (e.g., a chip or circuit) of the network device. This is not limited to the present specification.

[0111] The method includes: a network device transmitting first reference signal resource configuration information to a terminal device, the first reference signal resource configuration information including first instruction information, wherein the first instruction information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for performing model training on an AI model, or the first instruction information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for acquiring a dataset for performing model training on the AI ​​model.

[0112] For the beneficial effects of the fifth aspect and implementation forms of the fifth aspect, please refer to the description of the second aspect.

[0113] For example, the AI ​​model is for beam management.

[0114] For example, the first reference signal resource configuration information may be carried in one or more of the following fields: a CSI reporting configuration field, a CSI resource configuration field, a CSI-RS resource set field, or a CSI-RS resource set list field.

[0115] Referring to the fifth aspect, in some implementation forms of the fifth aspect, the first instruction information further indicates that the first reference signal resource is for obtaining labels for performing model training on an AI model.

[0116] Referring to the fifth aspect, in some implementation forms of the fifth aspect, the first reference signal resource configuration information further includes fifth instruction information, and the fifth instruction information indicates that the first reference signal resource is for obtaining labels for performing model training on an AI model.

[0117] Referring to the fifth aspect, in some implementation forms of the fifth aspect, the method further includes: the network device transmitting third reference signal resource configuration information to the terminal device, wherein the third reference signal resource configuration information includes sixth instruction information, and the sixth instruction information indicates that the third reference signal resource configured based on the third reference signal resource configuration information is for obtaining input information for performing model training on the AI ​​model.

[0118] Referring to the fifth aspect, in some implementation forms of the fifth aspect, the first reference signal resource configuration information further includes seventh instruction information indicating an association relationship between the first reference signal resource and the third reference signal resource, or the third reference signal resource configuration information further includes eighth instruction information indicating an association relationship between the first reference signal resource and the third reference signal resource.

[0119] For example, the third reference signal resource corresponds to an SSB resource, and the first reference signal resource corresponds to a CSI-RS resource.

[0120] Referring to the fifth aspect, in some implementation forms of the fifth aspect, the method further includes the network device receiving first information from the terminal device, the first information being for requesting to perform model training based on the AI ​​model, and the first information including one or more of an identifier of the AI ​​model, computing capability information of the terminal device, sparse beam instruction information, or a training mechanism of the AI ​​model.

[0121] According to a sixth aspect, there is provided a communication method. The method may be performed by a network device or may be performed by a component (e.g., a chip or circuit) of the network device. This is not limited herein.

[0122] The method includes a network device transmitting first reference signal resource configuration information to a terminal device, the first reference signal resource configuration information including first instruction information, wherein the first instruction information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for performing model inference on an AI model, or the first instruction information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for acquiring a dataset for performing model inference on the AI ​​model.

[0123] For the beneficial effects of the sixth aspect and implementation forms of the sixth aspect, please refer to the description of the third aspect.

[0124] For example, the AI ​​model is for beam management.

[0125] For example, the first reference signal resource configuration information may be carried in one or more of the following fields: a CSI reporting configuration field, a CSI resource configuration field, a CSI-RS resource set field, or a CSI-RS resource set list field.

[0126] Referring to a sixth aspect, in some implementation forms of the sixth aspect, the method further includes the network device transmitting second reference signal resource configuration information to the terminal device, wherein the second reference signal resource configuration information includes second instruction information, and the second instruction information indicates that the second reference signal resource configured based on the second reference signal resource configuration information is for performing model training on the AI ​​model, or the second instruction information indicates that the second reference signal resource configured based on the second reference signal resource configuration information is for acquiring a dataset for performing model training on the AI ​​model.

[0127] Referring to the sixth aspect, in some implementation forms of the sixth aspect, the first reference signal resource configuration information further includes third instruction information indicating an association relationship between the first reference signal resource and the second reference signal resource, or the second reference signal resource configuration information further includes fourth instruction information indicating an association relationship between the second reference signal resource and the first reference signal resource.

[0128] Referring to the sixth aspect, in some implementation forms of the sixth aspect, the first reference signal resource is a reference signal resource corresponding to input information for model inference of the AI ​​model, and the first reference signal resource configuration information further includes first identifier information. The first identifier information corresponds to a target reference signal resource set, and the target reference signal resource set includes reference signal resources corresponding to labels in a model training process of the AI ​​model.

[0129] For example, the first identifier information includes a sequence number of a target reference signal resource set in a first target reference signal resource set group, the first target reference signal resource set group being a reference signal resource set group configured for multiple cells or multiple network devices, the multiple network devices including a network device accessed by a terminal device, and the multiple cells including a cell in the network device accessed by the terminal device in which the terminal device executes an AI model.

[0130] Referring to the sixth aspect, in some implementation forms of the sixth aspect, the method further includes: the network device transmitting fifth reference signal resource configuration information to the terminal device, wherein the fifth reference signal resource configuration information is for configuring a target reference signal resource set, the fifth reference signal resource configuration information includes first identifier information, and the fifth reference signal resource configuration information further includes second identifier information, and the second identifier information corresponds to the target reference signal resource set.

[0131] For example, the second identifier information includes a sequence number of a target reference signal resource set in a first target reference signal resource set group, the first target reference signal resource set group being a reference signal resource set group configured for a plurality of cells or a plurality of network devices, the plurality of network devices including network devices accessed by a terminal device, and the plurality of cells including cells in the network devices accessed by the terminal device where the terminal device executes an AI model. The first identifier information includes a sequence number of a target reference signal resource set in a second target reference signal resource set group, the second target reference signal resource set group being a reference signal resource set group configured for a cell where the terminal device executes an AI model.

[0132] Referring to the sixth aspect, in some implementation forms of the sixth aspect, the method further includes the network device receiving first information from the terminal device. The first information is for requesting that model inference be performed on the AI ​​model, and the first information includes one or more of an identifier of the AI ​​model, instruction information #1, or sparse beam instruction information. The instruction information #1 indicates that training of the AI ​​model has been completed.

[0133] According to a seventh aspect, there is provided a communication device. The device is configured to perform a method according to any one of the first to sixth aspects. Specifically, the device may include a unit and / or module, such as a processing unit and / or a transceiver unit, configured to perform the method according to any one of the implementation forms of any one of the first to sixth aspects. The communication device may be a terminal device or a network device.

[0134] In one implementation, the apparatus is a communication device. When the apparatus is a communication device, the transceiver unit may be a transceiver or an input / output interface, and the processing unit may be at least one processor. Optionally, the transceiver may be a transceiver circuit. Optionally, the input / output interface may be an input / output circuit.

[0135] In another implementation, the apparatus is a chip, chip system, or circuit used in a communications device. When the apparatus is a chip, chip system, or circuit used in a communications device, the transceiver unit may be an input / output interface, interface circuit, output circuit, input circuit, pin, associated circuit, etc. within the chip, chip system, or circuit, and the processing unit may be at least one processor, processing circuit, logic circuit, etc.

[0136] According to an eighth aspect, there is provided a communication device. The device includes a memory configured to store a program, and at least one processor configured to execute a computer program or instructions stored in the memory to perform a method according to any one of the implementation forms of any one of the first to sixth aspects. The communication device may be a terminal device or a network device.

[0137] In one implementation, the apparatus is a communications device.

[0138] In another implementation, the apparatus is a chip, chip system, or circuit used in a communications device.

[0139] According to a ninth aspect, the present application provides a processor configured to perform the method provided in the above aspect.

[0140] Operations such as transmitting and acquiring / receiving related to a processor may be understood as operations such as output and input of the processor, or transmitting and receiving operations performed by radio frequency circuits and antennas, unless otherwise specified or provided that the operations do not contradict the actual function or internal logic of the operations in the relevant description. This is not a limitation in this application.

[0141] According to a tenth aspect, there is provided a computer-readable storage medium, the computer-readable storage medium storing program code to be executed by a device, the program code being used to perform any one of the implementation forms of any one of the methods according to the first to sixth aspects.

[0142] According to an eleventh aspect, there is provided a computer program product comprising instructions, which, when executed on a computer, cause the computer to perform a method according to any one of the implementations of any one of the first to sixth aspects.

[0143] According to a twelfth aspect, there is provided a chip, the chip including a processor and a communication interface, wherein the processor reads instructions stored in a memory via the communication interface and executes a method according to any one of the implementation forms of any one of the first to sixth aspects.

[0144] Optionally, in one implementation, the chip further includes a memory. The memory stores a computer program or instruction. The processor is configured to execute the computer program or instruction stored in the memory. When the computer program or instruction is executed, the processor is configured to perform a method according to any one of the implementations of any one of the first to sixth aspects.

[0145] According to a thirteenth aspect, there is provided a communication system including the terminal device and / or network device as described above. [Brief explanation of the drawings]

[0146] [Figure 1] 1 is a diagram of a wireless communication system 100 applicable to an embodiment of the present application. [Figure 2] FIG. 1 illustrates the structure of a neuron. [Figure 3] FIG. 1 is a diagram illustrating the layer relationship of a neural network. [Figure 4] FIG. 1 is a diagram of a framework for AI model training and inference according to the present application. [Figure 5] 1 is a diagram of a wide beam and a narrow beam applicable to an embodiment of the present application; [Figure 6] 6 is a schematic flowchart of a communication method 600 according to an embodiment of the present application. [Figure 7] 7 is a schematic flowchart of a communication method 700 according to another embodiment of the present application. [Figure 8] 8 is a schematic flowchart of a communication method 800 according to yet another embodiment of the present application. [Figure 9] 9 is a schematic flowchart of a communication method 900 according to yet another embodiment of the present application. [Figure 10] 1 is a diagram of a communication device 1000 according to an embodiment of the present application. [Figure 11] 11 is a diagram of another communication device 1100 according to an embodiment of the present application. [Figure 12] FIG. 12 is a diagram of a chip system 1200 according to an embodiment of the present application. DETAILED DESCRIPTION OF THE INVENTION

[0147] The following describes the technical solutions of the embodiments in this application with reference to the accompanying drawings.

[0148] The technical solutions provided in the present application may be applied to various communication systems, for example, future communication systems such as 5th generation (5G) or new radio (NR) systems, long term evolution (LTE) systems, LTE frequency division duplex (FDD) systems, LTE time division duplex (TDD) systems, wireless local area network (WLAN) systems, satellite communication systems, sixth generation mobile communication systems, or integrated systems of multiple systems. The technical solutions provided in the present application may further be applied to device-to-device (D2D) communication, vehicle-to-everything (V2X) communication, machine-to-machine (M2M) communication, machine type communication (MTC), internet of things (IoT) communication systems, or other communication systems.

[0149] The terminal device in the embodiments of the present application includes various devices having wireless communication capabilities, and the terminal device may be configured to be connected to a person, an object, a machine, etc. The terminal device may be widely used in various scenarios such as cellular communication, D2D, V2X, peer-to-peer (P2P), M2M, MTC, IoT, virtual reality (VR), augmented reality (AR), industrial control, autonomous driving, telemedicine, smart grid, smart furniture, smart office, smart wearable, smart transportation, smart city, unmanned aerial vehicles, robots, remote sensing, passive sensing, positioning, navigation and tracking, and autonomous delivery. The terminal device may be a terminal in any of the above-mentioned scenarios, such as an MTC terminal or an IoT terminal.The terminal device may be a user equipment (UE), terminal, fixed device, mobile station device or mobile device, subscriber unit, handheld device, in-vehicle device, wearable device, cellular phone, smartphone, session initiation protocol (SIP) phone, wireless data card, personal digital assistant (PDA), computer, tablet computer, notebook computer, wireless modem, handheld device, laptop computer, computer with radio transceiver capability, smartbook, vehicle, satellite, global positioning system (GPS), etc. The terminal device may be a GPS (Global Positioning System, GPS) device, a target tracking device, a flight device (e.g., an unmanned aerial vehicle, a helicopter, a multi-copter, a quad-copter, or an airplane), a ship, a remote control device, a smart home device, or an industrial device, a device built into the above-mentioned device (e.g., a communication module, modem, or chip in the above-mentioned device), or another processing device connected to a wireless modem. For ease of explanation, an example in which the terminal device is a terminal or UE is used hereinafter for explanation.

[0150] It should be understood that in some scenarios, the UE may further be used as a base station, for example, the UE may act as a scheduling entity providing sidelink signals between UEs in scenarios such as V2X, D2D, or P2P.

[0151] In an embodiment of the present application, an apparatus configured to implement the functions of a terminal device may be the terminal device or may be an apparatus capable of supporting the terminal device in implementing the functions, such as a chip system or a chip. The apparatus may be installed in the terminal device. In an embodiment of the present application, the chip system may include a chip, or may include a chip and other discrete components.

[0152] The network device in the embodiment of the present application may be a device configured to communicate with a terminal device. The network device may also be referred to as an access network device or a radio access network device. For example, the network device may be a base station. The network device in the embodiment of the present application may be a radio access network (RAN) node (or device) that connects the terminal device to a wireless network. The base station may broadly cover various names below, or alternatively may be named NodeB (NodeB), evolved NodeB (eNB), next generation NodeB (gNB), relay station, access point, transmitting and receiving point (TRP), transmitting point (TP), primary station, secondary station, motor slide retainer (MSR) node, home base station, network controller, access node, wireless node, access point (AP), transmitting node, transceiver node, baseband unit (BBU), remote radio unit (RRU), active antenna unit (AAU), radio frequency head (RRH), central unit (CU), distributed unit (DU), positioning node, etc. The base station may also be a macro base station, a micro base station, a relay node, a donor node, etc., or a combination thereof. Alternatively, the base station may be a communication module, modem, or chip disposed in the above-mentioned device or apparatus. Alternatively, the base station may be a mobile switching center, a device performing base station functions in D2D, V2X, and M2M communications, a network-side device in a 6G network, a device performing base station functions in future communication systems, etc.The base stations can support networks using the same access technology or different access technologies. The specific technology used for the network devices and the specific device configuration are not limited to the embodiments of the present application.

[0153] A base station may be fixed or mobile. For example, a helicopter or unmanned aerial vehicle may be configured as a mobile base station, and one or more cells may move based on the location of the mobile base station. In another example, a helicopter or unmanned aerial vehicle may be configured as a device for communicating with another base station.

[0154] In an embodiment of the present application, an apparatus configured to implement the functions of a network device may be a network device or an apparatus capable of supporting a network device in implementing the functions, such as a chip system or a chip. The apparatus may be installed in a network device. In an embodiment of the present application, a chip system may include a chip, or may include a chip and other discrete components.

[0155] The network devices and terminal devices may be deployed on land, on the water surface, or on airborne aircraft, balloons, and satellites, including indoor or outdoor devices, handheld devices, or vehicle-mounted devices. The scenarios in which the network devices and terminal devices are located are not limited in the embodiments of the present application. In addition, each of the terminal devices and network devices may be a hardware device, a software function running on dedicated hardware, a software function running on general-purpose hardware, for example, a virtualization function instantiated on a platform (e.g., a cloud platform), or an entity including dedicated or general-purpose hardware devices and software functions. The specific forms of the terminal devices and network devices are not limited in the present application.

[0156] A communication system applicable to embodiments of the present application will first be briefly described below.

[0157] FIG. 1 is a diagram of a wireless communication system 100 applicable to one embodiment of the present application. As shown in FIG. 1, the wireless communication system includes a radio access network 100. The radio access network 100 may be a next-generation (e.g., 6G or higher) radio access network or a conventional (e.g., 5G, 4G, 3G, or 2G) radio access network. One or more terminal devices (120a-120j, collectively referred to as 120) may be interconnected or connected to one or more network devices (110a and 110b, collectively referred to as 110) in the radio access network 100. FIG. 1 is merely a diagram. The wireless communication system may further include other devices, such as core network devices, wireless relay devices, and / or wireless backhaul devices, which are not shown in FIG. 1.

[0158] In practical applications, a wireless communication system may include multiple network devices (also referred to as access network devices) and multiple terminal devices. This is not limited. One network device may provide services to one or more terminal devices. One terminal device may also access one or more network devices. The number of terminal devices and network devices included in a wireless communication system is not limited in the embodiments of the present application.

[0159] The communication system may further include network elements having artificial intelligence capabilities. Phases related to AI model design, such as one or more of a data collection phase (e.g., collection of training data and / or inference data), a model training phase, and a model inference phase, may be performed by one or more network elements having artificial intelligence capabilities. In a possible design, the AI ​​capabilities may be configured in existing network elements (e.g., AI modules or AI entities) in the communication system to implement AI-related operations, such as AI model training and / or inference. For example, the existing network elements may be access network devices (e.g., gNBs), terminal devices, core network devices, network management systems, etc. The network management systems may detect network performance status, optimize network connectivity and performance, improve network performance stability, and reduce network maintenance costs. Alternatively, in another possible design, independent network elements may be introduced into the communication system to perform AI-related operations, such as AI model training. The independent network elements may be referred to as AI network elements, AI nodes, etc. The names are not limiting in this disclosure. An AI network element may be directly connected to an access network device in a communication system or indirectly connected to the access network device through a third-party network element, which may be, but is not limited to, a core network device, such as an authentication management function (AMF) network element or a user plane function (UPF) network element, a network management system, a cloud server, or another network element.Optionally, an AI network element may be a hardware device, a software function running on dedicated hardware, a software function running on general-purpose hardware, a virtualized function instantiated on a platform (e.g., a cloud platform), or an entity including dedicated or general-purpose hardware devices and software functions. Optionally, an AI network element may be located within a server, for example, within a host or cloud server of an OTT system.

[0160] In the present disclosure, one parameter or multiple parameters may be obtained through inference based on one model. Training processes for different models may be deployed on different devices or nodes, or on the same device or node. Inference processes for different models may be deployed on different devices or nodes, or on the same device or node. This is not a limitation in the present disclosure.

[0161] The present disclosure relates to model processing or operations, such as model training or model inference. Optionally, model training includes one or more of model initial training, retraining, model updating, model performance validation or monitoring, and the like.

[0162] To facilitate understanding of the embodiments of the present application, the following first provides a brief explanation of terms used in the embodiments of the present application.

[0163] 1. Artificial intelligence (AI): AI allows machines to learn, accumulate experience, and solve problems that can be solved by humans through experience, such as natural language understanding, image recognition, and chess playing. Artificial intelligence can be understood as intelligence displayed by machines created by humans. In general, artificial intelligence is the technology of using computer programs to exhibit human intelligence. The goal of artificial intelligence involves understanding intelligence through symbolic reasoning or building computer programs for reasoning.

[0164] 2. Machine learning: Machine learning is an implementation of artificial intelligence. Machine learning can be divided into supervised learning, unsupervised learning, and reinforcement learning.

[0165] In supervised learning, a mapping relationship between sample values ​​and sample labels is learned by using a machine learning algorithm based on collected sample values ​​and sample labels, and the learned mapping relationship is expressed by using a machine learning model. The process of training a machine learning model is a process of learning a mapping relationship. For example, in signal detection, a received signal containing noise is a sample, and an actual constellation corresponding to the signal is a label. In machine learning, it is expected that the mapping relationship between the samples and the labels will be learned through training, i.e., the machine learning model will be able to train a signal detector. During training, the model parameters are optimized by calculating the error between the model's predicted value and the actual label. Once the mapping relationship is learned, the learned mapping relationship can be used to predict the sample label of each new sample. The mapping relationship learned by supervised learning can include linear mapping and nonlinear mapping. Learning tasks may be classified into classification tasks and regression tasks based on the type of label.

[0166] In unsupervised learning, the internal patterns of samples are autonomously explored by using an algorithm based on collected sample values. In a specific type of unsupervised learning algorithm, samples are used as training signals, in other words, the model learns the mapping relationship between samples, which is called self-supervised learning. During training, the model parameters are optimized by calculating the error between the model's predicted values ​​and the samples. Self-supervised learning can be used for signal compression and decompression restoration. Common algorithms include autoencoders, generative adversarial networks, etc.

[0167] Unlike supervised learning, reinforcement learning is an algorithm that learns problem-solving strategies through interaction with the environment. Unlike supervised and unsupervised learning, reinforcement learning does not have a clear "correct" action label. The algorithm must interact with the environment, obtain a reward signal fed back by the environment, and adjust its decision behavior to obtain a larger reward signal value. For example, in downlink power control, a reinforcement learning model adjusts each user's downlink transmit power based on the total system throughput fed back by the wireless network in the hope of obtaining a higher system throughput. The goal of reinforcement learning is also to learn a mapping relationship between the environmental status and optimal decision behavior. However, it is not possible to obtain a "correct" action label in advance. Therefore, it is not possible to calculate the error between an action and the "correct action" and optimize the network. Reinforcement learning training is implemented through repeated interactions with the environment.

[0168] Deep learning is also an important branch of machine learning. Based on the above-mentioned learning algorithm, deep learning uses a neural network architecture to establish a mapping relationship between input data and network output. According to the universal approximation theorem, a neural network can theoretically approximate any continuous function, and therefore has the ability to learn any mapping. Traditional communication systems require extensive expert knowledge to design communication modules. However, deep learning-based communication systems can automatically discover implicit pattern structures from large datasets, establish mapping relationships between data, and achieve better performance than traditional modeling methods.

[0169] Deep learning algorithms can generally be divided into two major phases: the training phase and the inference phase. In the training phase, a large amount of data is required for the network to learn the mapping relationship between input and output. Therefore, the quality of the training data directly affects the performance of the AI ​​algorithm. After training is completed, the deep learning algorithm is used in the inference phase to infer corresponding output results based on the input data.

[0170] 3. Neural Network: A neural network is a specific embodiment of a machine learning method. A neural network is a mathematical model that processes information and mimics the behavioral characteristics of animal neural networks. The concept of a neural network comes from the neuron structure of brain tissue. Each neuron can perform a weighted sum operation on the neuron's input values ​​and outputs the result obtained from the weighted sum operation via an activation function.

[0171] Figure 2 shows the structure of a neuron. As shown in Figure 2, the input of a neuron is x = [x1, x2, ..., x n ], and the weights corresponding to the inputs are w=[w1,w2,…,w n] and the offset of the weighted sum is b. b may be an integer, a decimal, or a complex number, among other possible values. The form of the activation function may vary. In one example, the activation function of a neuron is y = f(z) = max(0,z), and the output of the neuron is

number

number

[0172] A neural network typically includes a multi-layer structure, and each layer may include one or more logical decision-making units, sometimes referred to as neurons. Increasing the depth and / or width of a neural network can improve the representational capabilities of the neural network, providing more powerful information extraction and abstraction modeling capabilities for complex systems. The depth of a neural network may be understood as the number of layers included in the neural network, and the number of neurons included in each layer may be referred to as the layer width.

[0173] FIG. 3 is a diagram showing the layer relationship of a neural network.

[0174] In a possible implementation, a neural network includes an input layer and an output layer. The input layer of a neural network performs neuron processing on the received input and passes the results to the output layer, which obtains the output result of the neural network.

[0175] In another possible implementation, as shown in Figure 3, a neural network includes an input layer, a hidden layer, and an output layer. The input layer of the neural network performs neuron processing on the received input and passes the result to an intermediate hidden layer. The hidden layer transmits the calculation result to the output layer or an adjacent hidden layer. 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 to this.

[0176] A loss function may be defined in the neural network training process. The loss function is used to measure the difference between the model's predicted value and the actual value. In the neural network training process, the loss function describes the gap or difference between the neural network's output value and an ideal target value. The neural network training process is a process of adjusting neural network parameters to make the value of the loss function less than a threshold or meet the target requirement. The neural network parameters may include at least one of the number of layers of the neural network, the width of the neural network, the weights of the neurons, or parameters in the neuron's activation function.

[0177] 4. AI Model: An AI model is an algorithm or computer program that can implement an AI function. An AI model represents a mapping relationship between the input and output of the model, or an AI model is an AI model that maps an input of a specific dimension to an output of a specific dimension. The parameters of the function model may be obtained through machine learning training. For example, f(x) = ax² + b is a quadratic function model and may be considered as an AI model, where a and b are parameters of the AI ​​model, and a and b may be obtained through machine learning training. For example, the AI ​​model referred to in the following embodiments of the present application is not limited to a neural network, a linear regression model, a decision tree model, a support vector machine (SVM), a Bayesian network, a Q-learning model, or another machine learning (ML) model.

[0178] It can be understood that the AI ​​model can be implemented using hardware circuitry, software, or a combination of software and hardware. This is not limited to this. Non-limiting examples of software include program code, programs, subprograms, instructions, instruction sets, code, code segments, software modules, applications, software applications, etc.

[0179] 5. AI model training and AI model inference

[0180] Any AI model needs to be trained before it can be used to solve a specific technical problem.

[0181] As shown in Figure 4, AI model training is the process of calculating training data using a specified initial model and adjusting the parameters of the initial model using a specific method based on the calculation results, so that the model gradually learns specific rules and has specific functions. After training, an AI model with stable functions can be used for inference. AI model inference is the process of calculating input data using a trained AI model to obtain predicted inference results.

[0182] In the training phase, we first need to build a training set for the deep learning model based on the objective. The training set includes multiple training data, each of which has a label. The label of the training data is the correct answer of the training data to a specific question, and the label can represent the objective of training the deep learning model using the training data.

[0183] When a deep learning model is trained, training data may be input to the deep learning model in batches after parameter initialization, and the deep learning model performs calculations (i.e., "inference") on the training data to obtain prediction results for the training data. The prediction results obtained by inference and the labels corresponding to the training data are used as data for calculating losses based on a loss function. The loss function is a function used in the model training phase to calculate the difference (i.e., "loss value") between the model's prediction results for the training data and the labels of the training data. The loss function can be implemented using different mathematical functions. Common expressions of the loss function include the mean squared error loss function, the logarithmic loss function, and the least squares method. Model training is a process of repeated iterations. In each iteration, different training data is inferred and the loss function value is calculated. The purpose of multiple iterations is to continuously update the parameters of the deep learning model and find a parameter configuration that minimizes or gradually stabilizes the loss value of the loss function.

[0184] 6. Training and inference data sets

[0185] A training dataset is used to train an AI model. The training dataset may include the input of the AI ​​model, or may include the input of the AI ​​model and the target output of the AI ​​model. The training dataset includes one or more training data. The training data may be training samples input to the AI ​​model or the target output of the AI ​​model. The target output may also be referred to as a label or label sample. The training dataset is one of the important parts of machine learning. Essentially, model training is learning several features from the training data so that the output of the AI ​​model is as close as possible to the target output, e.g., so that the difference between the output of the AI ​​model and the target output is minimized. The composition and selection of the training dataset can determine to some extent the performance of the trained AI model. Model performance can be measured, for example, by a "loss value" or "inference accuracy."

[0186] Also, in the training process of an AI model (e.g., a neural network), a loss function can be defined. The loss function describes the gap or difference between the output value of the AI ​​model and the target output value. The specific form of the loss function is not limited in this application. The training process of an AI model is a process in which the model parameters of the AI ​​model are adjusted so that the value of the loss function is less than a threshold or the value of the loss function meets the target requirement. For example, if the AI ​​model is a neural network, adjusting the model parameters of the neural network includes adjusting at least one of the following parameters: the number of layers of the neural network, the width of the neural network, the weights of the neurons, or parameters in the activation functions of the neurons.

[0187] The inference data may be used as input to a trained AI model and used for inference of the AI ​​model. During model inference, the inference data is input to the AI ​​model to obtain a corresponding output, i.e., an inference result.

[0188] 7. Beam: A beam is a communication resource. A beam in the NR protocol can be a spatial filter, a spatial filter, or a spatial parameter. A beam used to transmit a signal may be referred to as a transmission beam (Tx beam), or a spatial domain transmit filter or a spatial domain transmit parameter. A beam used to receive a signal may be referred to as a reception beam (Rx beam), or a spatial domain receiver filter or a spatial domain receive parameter.

[0189] A transmit beam may refer to the signal strength distribution formed in different directions in space after a signal is transmitted through an antenna, and a receive beam may refer to the signal strength distribution in different directions in space of a wireless signal received from an antenna.

[0190] It should be understood that the above-listed representations of beams in the NR protocol are merely examples and are not intended to constitute any limitation on the present application. The present application does not exclude the possibility that other terms may be defined in future protocols to indicate the same or similar meanings.

[0191] In addition, the beam may be a wide beam, a narrow beam, or another type of beam. The technology for forming the beam may be a beamforming technology or another technology. The beamforming technology may specifically be considered as different resources. The same information or different information may be transmitted through different beams.

[0192] For example, when a low band or a mid band is used, a signal can be transmitted omnidirectionally or at a wide angle. When a high band is used, due to the short carrier wavelength of a high-frequency communication system, an antenna array including multiple antenna elements can be arranged at the transmitting end and the receiving end. The transmitting end transmits a signal using a specific beamforming weight, allowing the transmitted signal to form a spatially directional beam, and the receiving end receives the signal via the antenna array using a specific beamforming weight, thereby increasing the received power of the signal at the receiving end and avoiding path loss.

[0193] 5 is a diagram of a wide beam and a narrow beam applicable to one embodiment of the present application. As shown in (a) of FIG. 5, a network device and a terminal device may communicate with each other by using a wide beam, or may communicate with each other by using a narrow beam. As shown in (b) of FIG. 5, a narrow beam has a similar spotlight effect, concentrating limited transmission energy in a narrow direction to significantly increase the coverage of the network device.

[0194] 8. Beam Sweeping: Beam sweeping refers to the transmission of beams in predefined directions with a fixed periodicity to cover a specific spatial area during a specific period or time period. For example, during initial access, the UE needs to synchronize with the system and receive minimal system information. Therefore, the bearer synchronization signal and physical broadcast channel (PBCH) block (SSB) are used to perform sweeping and transmission at a fixed period. The channel state information reference signal (CSI-RS) can also use beam sweeping techniques. However, if all predefined beam directions need to be covered, the overhead of the CSI-RS would be excessively high. In this case, the CSI-RS is transmitted only in a specific subset of predefined beam directions based on the location of the terminal device being served.

[0195] 9. Beam measurement: Beam measurement is a process in which a network device or a terminal device measures the quality and characteristics of a received beamformed signal. In the beam management process, the terminal device or the network device can obtain information such as reference signal received power (RSRP), reference signal received quality (RSRQ), and signal to interference plus noise ratio (SINR) by using SSB and CSI-RS to identify the optimal beam.

[0196] 10. Beam Management: Appropriate beam pairs are established and maintained between the network device and the terminal device. For downlink transmission, the network side needs to select an appropriate transmit beam, and the terminal side needs to select an appropriate receive beam to jointly form a beam pair to maintain a good wireless connection. The above beam selection process is sometimes referred to as serving beam selection.

[0197] Currently, beam selection is mainly completed based on reference signals and corresponding beam measurements. Specifically, the network side configures reference signal resources for the terminal side based on user capabilities and network resources. After the configuration is completed, the network side transmits reference signals to the terminal side based on the configuration. The terminal side measures the reference signals and feeds back the measurement results to the network side. The network side configures a transmission beam based on the measurement results reported by the terminal.

[0198] Currently, in the process of beam management combined with an AI model, the AI ​​model can be deployed on a training device or a terminal device for training and updating. Furthermore, after the AI ​​model training is completed, the terminal device can infer an optimal beam (there may be one or more optimal beams) based on the trained AI model. However, how to balance the performance and resource consumption of training an AI model and / or performing inference based on the trained AI model has not been resolved.

[0199] In consideration of this, the present application provides a communication method in which a terminal device can identify the purpose of a reference signal resource configured by a network device, and then the terminal device can perform model training or model inference based on the reference signal resource and the purpose of the reference signal resource. This helps to balance performance and resource consumption in training an AI model and / or performing inference based on an AI model.

[0200] It should be noted that in this application, "indication" may include a direct indication, an indirect indication, an explicit indication, or an implicit indication. One piece of indication information indicating A can be understood as follows: The indication information carries A and may directly indicate A, or may indirectly indicate A. An indirect indication may mean that the indication information directly indicates B and the correspondence between B and A to indicate A by using the indication information. The correspondence between B and A may be predefined in a protocol, pre-stored, or obtained by using a configuration between communication devices.

[0201] In this application, information indicated by the indication information is referred to as information to be indicated. In a specific implementation process, there are many ways to indicate the information to be indicated. For example, these ways include, but are not limited to, ways in which the information to be indicated, such as the information to be indicated or an index of the information to be indicated, can be directly indicated. Alternatively, the information to be indicated can be indirectly indicated by indicating other information, and there is an association relationship between the other information and the information to be indicated. Alternatively, only a portion of the information to be indicated can be indicated, and other portions of the information to be indicated can be known or agreed upon in advance. For example, specific information can alternatively be indicated by using a pre-agreed (e.g., protocol-defined) arrangement sequence of multiple pieces of information to reduce indication overhead to some extent. In addition, the information to be indicated can be transmitted as a whole or divided into multiple sub-information pieces to be transmitted separately, and the transmission periods and / or transmission opportunities of these sub-information pieces can be the same or different.

[0202] It should be further noted that "at least one (item)" in this application refers to one (item) or multiple (items). "Multiple (items)" means two (items) or more than two (items). The term "and / or" describes an association relationship for describing related objects and indicates that three relationships may exist. For example, A and / or B may represent the following three cases: when only A exists, when both A and B exist, and when only B exists. The character " / " generally indicates an "or" relationship between related objects. In addition, terms such as "first," "second," etc. may be used in this application to describe objects, but it should be understood that these objects are not limited by these terms. These terms are used merely to distinguish objects from each other.

[0203] The following describes in detail the method provided in the embodiments of the present application with reference to the accompanying drawings. The embodiments provided in the present application may be applied to, but are not limited to, the communication system shown in Figure 1.

[0204] It should be noted that the kth reference signal resource (e.g., the first reference signal resource, the second reference signal resource, the third reference signal resource, and the fourth reference signal resource) in the embodiments of the present application may be replaced with the kth reference signal resource set, and the kth reference signal resource set includes at least one reference signal resource.

[0205] 6 is a schematic flowchart of a communication method 600 according to an embodiment of the present application. The method 600 shown in FIG. 6 may include the following steps.

[0206] S610: The terminal device sends first information to the network device.

[0207] Correspondingly, in S610, the network device receives first information from the terminal device.

[0208] The first information is for requesting execution of a first operation based on an AI model. The first operation is model training for the AI ​​model or performing model inference based on the AI ​​model. In a possible implementation, the first information is for requesting a reference signal resource for acquiring a dataset, the dataset being a dataset for executing the first operation based on the AI ​​model.

[0209] For example, the AI ​​model is for beam management.

[0210] For example, if the first operation is model training for an AI model, the first information may include one or more of an identifier of the AI ​​model, computing capability information of the terminal device, sparse beam instruction information, or a training mechanism of the AI ​​model.

[0211] The AI ​​model identifier is for identifying the AI ​​model. When the first information includes the AI ​​model identifier, the network device may identify the AI ​​model corresponding to the first operation requested by the terminal device based on the AI ​​model identifier. For example, when the terminal device supports requesting the network device to perform the first operation on multiple different AI models (e.g., including AI model #1 and AI model #2), and the first information sent by the terminal device to the network device includes the identifier of AI model #1, the network device may identify that the first operation requested by the terminal device corresponds to AI model #1. Note that when the terminal device supports requesting the network device to perform the first operation on one AI model, the first information does not need to include the identifier of the AI ​​model.

[0212] The computing capability information of the terminal device refers to information about factors that may affect the computing capability of the terminal device. For example, the computing capability information of the terminal device may include one or more of the capabilities of the processor (e.g., a central processing unit (CPU), a graphics processing unit (GPU), a tensor processing unit (TPU), a neural network processing unit (NPU), and a field-programmable gate array (FPGA)) of the terminal device, the size of the storage space of the terminal device, the size of the memory of the terminal device, the power of the terminal device, etc. If the first information includes the computing capability information of the terminal device, the network device may identify the capability of the terminal device to support model training of the AI ​​model, for example, the batch size of one training session, based on the computing capability information of the terminal device. Note that if the terminal device reports the computing capability information of the terminal device to the network device before performing method 600 or the computing capability information of the terminal device is pre-configured on the network device, the first information may not include the computing capability information of the terminal device.

[0213] The at least one group of sparse beams indicated by the sparse beam indication information is a subset of all beams supported by the network device. All beams supported by the network device are all beams usable by the network device. It can be understood that the number of all beams supported by the network device may be fixed or may have multiple values ​​corresponding to multiple application scenarios of the network device, for example, 16, 32, 64, or 128. In scenario 1, the corresponding value is 64, and in scenario 2, the corresponding value is 128. The number of all beams supported by the network device may be predefined in a protocol or configured by the network device for the terminal device or training device. If the first information includes sparse beam indication information, the network device can determine at least one group of sparse beams corresponding to the AI ​​model based on the sparse beam indication information. Furthermore, in the process of performing model inference based on the AI ​​model, the network device can configure reference signal resources for the terminal device based on the at least one group of sparse beams. In this way, the terminal device can obtain measurement results of reference signals corresponding to at least one group of sparse beams based on the reference signal resources configured by the network device, and the measurement results can be used as input information in the model inference process of the AI ​​model. A group of sparse beams in the at least one group of sparse beams can also be referred to as a sparse beam pattern.

[0214] The form of the sparse beam indication information is not limited in the embodiments of the present application. For example, the sparse beam indication information indicates at least one group of sparse beams by indicating the index of the sparse beam of the at least one group. In another example, the sparse beam indication information indicates at least one sparse beam group by indicating the beam identifier of the beam included in each sparse beam group.

[0215] It should be noted that the first information may not include sparse beam indication information. For example, if the terminal device reports sparse beam indication information to the network device before performing method 600, or if at least one group of sparse beams is pre-configured on the network device, the first information may not include sparse beam indication information.

[0216] The training mechanism of the AI ​​model refers to an algorithm corresponding to the AI ​​model. For example, if the AI ​​model is for beam management, the algorithm corresponding to the AI ​​model may be a classification algorithm in which the beam identifier of the optimal beam is used as a training label, or the algorithm corresponding to the AI ​​model may be a regression algorithm in which signal measurement results of reference signals corresponding to all beams supported by the network device are used as training labels. If the algorithm corresponding to the AI ​​model is a classification algorithm, the output of the AI ​​model is the probability that each beam in all beams supported by the network device is an optimal beam. There may be one or more optimal beams. If the algorithm corresponding to the AI ​​model is a regression algorithm, the output of the AI ​​model is the signal measurement results of reference signals corresponding to all beams supported by the network device. The signal measurement results of the reference signals may be one or more of RSRP, RSRQ, SINR, etc. Note that if the terminal device reports the training mechanism of the AI ​​model to the network device or if the training mechanism of the AI ​​model is pre-configured on the network device before performing method 600, the first information may not include the training mechanism of the AI ​​model.

[0217] In another example, if the first operation is model inference for an AI model, the first information may include one or more of an identifier of the AI ​​model, instruction information #1, or sparse beam instruction information. The instruction information #1 indicates that training of the AI ​​model has been completed.

[0218] It should be noted that when the AI ​​model is deployed on the terminal device, the terminal device transmits the first information to the network device in S610. When the AI ​​model is deployed on a training device different from the terminal device, the terminal device transmits the first information to the network device, or the training device transmits the first information to the network device in S610. When the AI ​​model is deployed on the training device and the first information transmitted by the terminal device or the training device to the network device includes computing power information, the computing power information is the computing power information of the training device.

[0219] It should be further noted that S610 is an optional step. For example, whether the terminal device or the training device performs the first operation based on the AI ​​model is determined by the network device, in which case S610 may not be performed in method 600.

[0220] S620: The network device sends first reference signal resource configuration information to the terminal device.

[0221] In response, the terminal device receives first reference signal resource configuration information from the network device.

[0222] The first reference signal resource configuration information may be carried in one or more of a CSI reporting configuration (CSI-ReportConfig) field, a CSI resource configuration (CSI-ResourceConfig) field, a CSI-RS resource set (ResourceSet) field, or a CSI-RS resource set list (ResourceSetList) field.

[0223] The first reference signal resource configuration information includes first indication information, where the first indication information indicates a purpose of the first reference signal resource configured based on the first reference signal resource configuration information. Accordingly, the terminal device can determine the purpose of the first reference signal resource based on the first indication information. Alternatively, the first indication information indicates an operation corresponding to the first reference signal resource (e.g., a first operation based on an AI model). Accordingly, the terminal device can determine the operation corresponding to the first reference signal resource based on the first indication information. For example, the first indication information indicates that the first reference signal resource is for performing the first operation based on an AI model. Alternatively, the first indication information indicates that the first reference signal resource is for acquiring a dataset, where the dataset is a dataset for performing the first operation based on the AI ​​model. Alternatively, the first indication information indicates that the first reference signal resource is for performing a non-AI model-based operation.

[0224] In a possible implementation, the first reference signal resource configuration information may not include the first indication information. If the first reference signal resource configuration information does not include the first indication information, it indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for performing a non-AI model-based operation.

[0225] The specific format of the first instruction information is not limited in the embodiments of the present application.

[0226] For example, the operation corresponding to the first reference signal resource, i.e., the purpose of the first reference signal resource, may be indicated based on different values ​​of the first indication information, the correspondence between different operations based on the AI ​​model, and the specific value of the first indication information.

[0227] For example, the first instruction information is 1-bit information. When the value of the first instruction information is “1”, the first instruction information indicates that the first reference signal resource is for performing model training on an AI model (i.e., the first operation is model training on the AI ​​model), or when the value of the first instruction information is “0”, the first instruction information indicates that the first reference signal resource is for performing model inference on an AI model (i.e., the first operation is model inference on the AI ​​model). Alternatively, when the value of the first instruction information is “0”, the first instruction information indicates that the first reference signal resource is for performing model training on an AI model (i.e., the first operation is model training on the AI ​​model), or when the value of the first instruction information is “1”, the first instruction information indicates that the first reference signal resource is for performing model inference on an AI model (i.e., the first operation is model inference on the AI ​​model). It may be understood that the present disclosure is described by using multiple values ​​of the first instruction information corresponding to multiple “first operations”. Alternatively, when multiple values ​​of the first indication information correspond to multiple actions, the multiple actions may be referred to as a first action, a second action, etc. This is not limited to this specification.

[0228] In another example, the first instruction information is 2-bit information. When the value of the first instruction information is “00”, the first instruction information indicates that the first reference signal resource is for performing model training on an AI model (i.e., the first operation is model training on an AI model). When the value of the first instruction information is “01”, the first instruction information indicates that the first reference signal resource is for performing model inference on an AI model (i.e., the first operation is model inference on an AI model). When the value of the first instruction information is “10”, the first instruction information indicates that the first reference signal resource is for performing a non-AI model-based operation.

[0229] Optionally, the first reference signal resource configuration information further includes indication information #8, where the indication information #8 indicates an identifier of the AI ​​model. Correspondingly, the terminal device may determine the identifier of the AI ​​model based on the indication information #8, and then the terminal device determines that the first reference signal resource is for performing a first operation on the AI ​​model identified by the identifier of the AI ​​model.

[0230] Optionally, if the first operation is model inference for the AI ​​model, method 600 further includes the network device transmitting second reference signal resource configuration information to the terminal device. Correspondingly, the terminal device receives the second reference signal resource configuration information from the network device. The second reference signal resource configuration information includes second instruction information, where the second instruction information indicates that the second reference signal resource configured based on the second reference signal resource configuration information is for performing model training for the AI ​​model, or the second instruction information indicates that the second reference signal resource is for acquiring dataset #1, where dataset #1 is a dataset for performing model training for the AI ​​model. For a specific format of the second instruction information, please refer to the description of the specific format of the first instruction information.

[0231] The second reference signal resource configuration information may be carried in one or more of the following fields: a CSI reporting configuration field, a CSI resource configuration field, a CSI-RS resource set field, or a CSI-RS resource set list field.

[0232] Optionally, the second reference signal resource configuration information further includes instruction information #9, where the instruction information #9 indicates an identifier of the AI ​​model. Correspondingly, the terminal device may determine the identifier of the AI ​​model based on the instruction information #9, and then the terminal device determines that the second reference signal resource is for performing model training on the AI ​​model identified by the identifier of the AI ​​model.

[0233] Optionally, when the first operation is model inference for an AI model, the first reference signal resource is a reference signal resource corresponding to input information of the model inference of the AI ​​model, and the first reference signal resource configuration information further includes first identifier information. The first identifier information corresponds to a target reference signal resource set, and the target reference signal resource set includes reference signal resources corresponding to labels in the model training process of the AI ​​model. The reference signal resource corresponding to the labels in the model training process of the AI ​​model is denoted as reference signal resource #1. The correspondence between the labels in the model training process of the AI ​​model and reference signal resource #1 may be described as follows: Reference signal resource #1 is for obtaining labels in the model training process of the AI ​​model. For further description of the first identifier information, please refer to method 900 below. For conciseness, details will not be described here.

[0234] It should be understood that the target reference signal resource set refers to a reference signal resource set that includes reference signal resources corresponding to the labels in the model training process of the AI ​​model. The target reference signal resource set does not have any other special meaning. Alternatively, the target reference signal resource set may be replaced with another name, for example, reference signal resource set #A.

[0235] Optionally, if the first operation is model training for an AI model, the first instruction information further indicates that the first reference signal resource is for obtaining labels for performing model training for the AI ​​model.

[0236] For example, if the first instruction information further indicates that the first reference signal resource is for obtaining labels for performing model training on an AI model, the first instruction information may be 2-bit information. For example, if the value of the first instruction information is "00," the first instruction information indicates that the first reference signal resource is for obtaining labels for performing model training on an AI model. When the first instruction information indicates that the first reference signal resource is for obtaining labels for performing model training on an AI model, it can be understood that the first instruction information also indicates that the first reference signal resource is for performing a first operation based on the AI ​​model, and the first operation is to train the AI ​​model. In other words, the first instruction information is a field, and the value of the field corresponds to the first operation (i.e., the first objective) and a sub-objective of the first operation.

[0237] In another example, when the first instruction information further indicates that the first reference signal resource is for obtaining a label for performing model training on an AI model, the first instruction information may be two-bit information. For example, the first bit indicates that the first reference signal resource is for performing a first operation based on an AI model, and the second bit indicates whether the first reference signal resource is for obtaining a label for performing model training on an AI model. For example, if the value of the second bit is "0", the second bit indicates that the first reference signal resource is for obtaining a label for performing model training on an AI model. Alternatively, if the value of the second bit is "1", the second bit indicates that the first reference signal resource is for obtaining a label for performing model training on an AI model.

[0238] In another example, if the first instruction information further indicates that the first reference signal resource is for obtaining a label for performing model training on an AI model, the first instruction information may be three-bit information. For example, the first two bits indicate that the first reference signal resource is for performing a first operation based on an AI model, and the third bit indicates whether the first reference signal resource is for obtaining a label for performing model training on an AI model. For example, if the value of the second bit is "0", the second bit indicates that the first reference signal resource is for obtaining a label for performing model training on an AI model. Alternatively, if the value of the second bit is "1", the second bit indicates that the first reference signal resource is for obtaining a label for performing model training on an AI model.

[0239] In other words, the first instruction information includes two parts, Part 1 and Part 2. Part 1 indicates a first operation (i.e., a first objective), and Part 2 indicates a sub-objective of the first operation. Optionally, the number of bits in Part 1 is 1 or 2, and the number of bits in Part 2 is 1. Optionally, the first operation is model training, and the sub-objective of the first operation includes a label, an input, or an input and a label.

[0240] Optionally, when the first operation is model training for an AI model, the first reference signal resource configuration information further includes fifth instruction information, where the fifth instruction information indicates that the first reference signal resource is for obtaining a label for performing model training for the AI ​​model. Alternatively, the fifth instruction information indicates that the first reference signal resource is for obtaining an input for performing model training for the AI ​​model, or an input and a label. In other words, the multiple values ​​of the fifth instruction information separately indicate that the first reference signal resource is for obtaining a label, an input, or an input and a label for performing model training for the AI ​​model.

[0241] For example, the fifth instruction information may be 1-bit information. For example, if the value of the fifth instruction information is "0", the fifth instruction information indicates that the first reference signal resource is for obtaining labels for performing model training on an AI model. Alternatively, if the value of the fifth instruction information is "1", the fifth instruction information indicates that the first reference signal resource is for obtaining labels for performing model training on an AI model.

[0242] Optionally, if the first instruction information further indicates that the first reference signal resource is for obtaining labels for performing model training on an AI model, or if the first reference signal resource configuration information further includes fifth instruction information, and the fifth instruction information indicates that the first reference signal resource is for obtaining labels for performing model training on an AI model, method 600 further includes the network device sending third reference signal resource configuration information to the terminal device, and in response, the terminal device receiving the third reference signal resource configuration information from the network device.

[0243] For example, the first reference signal resource configured based on the first reference signal resource configuration information corresponds to a CSI-RS resource or the first reference signal resource is for transmitting a CSI-RS, and the third reference signal resource configured based on the third reference signal resource configuration information corresponds to an SSB resource or the third reference signal resource is for transmitting an SSB.

[0244] For example, reference signals transmitted on first reference signal resources configured based on the first reference signal resource configuration information correspond to all narrow beams supported by the network device, and all narrow beams supported by the network device are referred to as a narrow beam universal set. Reference signals transmitted on third reference signal resources configured based on the third reference signal resource configuration information correspond to all wide beams supported by the network device, and all wide beams supported by the network device are referred to as a wide beam universal set.

[0245] The third reference signal resource configuration information may be carried in one or more of the following fields: a CSI reporting configuration field, a CSI resource configuration field, a CSI-RS resource set field, or a CSI-RS resource set list field.

[0246] The third reference signal resource configuration information includes sixth instruction information, and the sixth instruction information indicates that the third reference signal resource configured based on the third reference signal resource configuration information is for obtaining input information for performing model training for the AI ​​model.

[0247] For example, the sixth instruction information may be 2-bit information. For example, if the value of the sixth instruction information is "01", the sixth instruction information indicates that the third reference signal resource is for obtaining input information for performing model training on an AI model. When the sixth instruction information indicates that the third reference signal resource is for obtaining input information for performing model training on an AI model, it can be understood that the sixth instruction information also indicates that the third reference signal resource is for performing a first operation based on the AI ​​model, and the first operation is to train the AI ​​model.

[0248] In another example, the sixth instruction information may be 1-bit information. For example, if the value of the sixth instruction information is "0", the sixth instruction information indicates that the third reference signal resource is to obtain input information for performing model training on the AI ​​model. Alternatively, if the value of the sixth instruction information is "1", the sixth instruction information indicates that the third reference signal resource is to obtain input information for performing model training on the AI ​​model.

[0249] Optionally, the third reference signal resource configuration information further includes instruction information #10, where the instruction information #10 indicates an identifier of the AI ​​model. Correspondingly, the terminal device can determine the identifier of the AI ​​model based on the instruction information #10, and then the terminal device determines that the third reference signal resource is for obtaining input information for performing model training on the AI ​​model identified by the identifier of the AI ​​model.

[0250] Furthermore, if the AI ​​model is deployed to a terminal device, S630a is performed in method 600. If the AI ​​model is deployed to a training device, S630b1 and S630b2 are performed in method 600.

[0251] S630a: The terminal device performs a first action based on the AI ​​model.

[0252] After the terminal device receives the first reference signal resource configuration information, if the first instruction information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for performing a first operation based on an AI model, or if the first instruction information indicates that the first reference signal resource is for acquiring a dataset and the dataset is a dataset for performing a first operation based on an AI model, the terminal device performs the first operation based on the AI ​​model based on the first reference signal resource configuration information.

[0253] The following uses an example in which the AI ​​model is for beam management to explain how the terminal device performs the first operation based on the AI ​​model.

[0254] For example, if the first operation is model training for an AI model, the terminal device performs model training for the AI ​​model based on the first reference signal resource configuration information in the following steps: (1) The terminal device measures reference signals transmitted on the first reference signal resource based on the first reference signal resource configuration configured based on the first reference signal resource configuration information, and obtains signal measurement results for the reference signals transmitted on the first reference signal resource (e.g., obtains the RSRP of the reference signals transmitted on the first reference signal resource), where the reference signals transmitted on the first reference signal resource correspond to all beams supported by the network device (all beams supported by the network device are hereinafter referred to as a beam universal set). (2) The terminal device determines, based on the RSRP of the reference signals transmitted on the first reference signal resource, a beam corresponding to a reference signal with the largest RSRP among the reference signals transmitted on the first reference signal resource as an optimal beam, and uses the beam identifier of the optimal beam as a label for performing model training for the AI ​​model. (3) The terminal device uses the RSRP of a reference signal corresponding to the group of sparse beams among the reference signals transmitted on the first reference signal resource as input information for model training and inputs the input information to the AI ​​model. (4) The terminal device obtains output information of the AI ​​model, which includes the probability that each beam in the beam universal set is an optimal beam. (5) The terminal device uses the beam in the beam universal set with the highest probability of being an optimal beam as the optimal beam predicted based on the AI ​​model. (6) The terminal device calculates a loss function value based on the model training label and the beam identifier of the optimal beam predicted based on the AI ​​model. If the loss function value is less than a predetermined threshold, it indicates that the training of the AI ​​model is complete. If the loss function value is greater than or equal to the predetermined threshold, the terminal device performs model training on the AI ​​model by referring to steps (1) to (6) until the training of the AI ​​model is complete.

[0255] In another example, if the first operation is model training for an AI model, the terminal device performs model training for the AI ​​model based on the first reference signal resource configuration information in the following steps: (a) The terminal device measures reference signals transmitted on the first reference signal resource based on the first reference signal resource configured based on the first reference signal resource configuration information, and obtains signal measurement results for the reference signals transmitted on the first reference signal resource (e.g., obtains RSRPs of the reference signals transmitted on the first reference signal resource), where the reference signals transmitted on the first reference signal resource correspond to a beam universal set; (b) The terminal device uses the RSRPs of the reference signals transmitted on the first reference signal resource obtained through the measurement as labels for performing model training for the AI ​​model; and (c) The terminal device uses the RSRPs of reference signals that are in the reference signal transmitted on the first reference signal resource and correspond to a group of sparse beams as input information for model training, and inputs the input information into the AI ​​model. (d) The terminal device acquires output information of the AI ​​model, the output information of the AI ​​model including RSRPs of reference signals corresponding to each beam in the beam universal set predicted based on the AI ​​model. (e) The terminal device calculates a loss function value based on the model training labels and the RSRPs of reference signals corresponding to each beam in the beam universal set predicted based on the AI ​​model. If the loss function value is less than a preset threshold, it indicates that training of the AI ​​model is complete. If the loss function value is greater than or equal to the preset threshold, the terminal device performs model training on the AI ​​model by referring to steps (a) to (e) until training of the AI ​​model is complete.

[0256] In another example, if the first operation is model training for an AI model, and the first instruction information further indicates that the first reference signal resource is for obtaining labels for performing model training for the AI ​​model, or the first reference signal resource configuration information further includes fifth instruction information, the terminal device performs the first operation based on the AI ​​model based on the first reference signal resource configuration information and the third reference signal resource configuration information.

[0257] For example, the terminal device performs a first operation based on an AI model that is based on first reference signal resource configuration information and third reference signal resource configuration information, in the following steps: (A) The terminal device measures reference signals transmitted on the first reference signal resource based on the first reference signal resource configured based on the first reference signal resource configuration information, and obtains signal measurement results of the reference signals transmitted on the first reference signal resource (e.g., obtains the RSRP of the reference signal transmitted on the first reference signal resource), where the reference signals transmitted on the first reference signal resource correspond to a narrow beam universal set; (B) The terminal device determines, based on the RSRP of the reference signals transmitted on the first reference signal resource, a beam corresponding to a reference signal with the largest RSRP among the reference signals transmitted on the first reference signal resource as an optimal beam, and uses the beam identifier of the optimal beam as a label for performing model training on the AI ​​model. (C) The terminal device measures the reference signal transmitted on the third reference signal resource based on the third reference signal resource configured based on the third reference signal resource configuration information, and obtains a signal measurement result of the reference signal transmitted on the third reference signal resource (e.g., obtains the RSRP of the reference signal transmitted on the third reference signal resource). (D) The terminal device uses the RSRP of the reference signal transmitted on the third reference signal resource as input information for model training for the AI ​​model. (E) The terminal device obtains output information of the AI ​​model, where the output information of the AI ​​model includes a probability that each beam in the narrow beam universal set is an optimal beam. (F) The terminal device uses the beam in the narrow beam universal set that has the highest probability of being an optimal beam as the optimal beam predicted based on the AI ​​model. (G) The terminal device calculates a loss function value based on the label of the model training and the beam identifier of the optimal beam predicted based on the AI ​​model. If the loss function value is smaller than a preset threshold, it indicates that training of the AI ​​model is complete. If the loss function value is greater than or equal to the preset threshold, the terminal device performs model training on the AI ​​model by referring to step (A) to step (G) until the training of the AI ​​model is completed.

[0258] In another example, the terminal device performs a first operation based on the AI ​​model based on the first reference signal resource configuration information and the third reference signal resource configuration information in the following steps: (i) The terminal device measures the reference signals transmitted on the first reference signal resources based on the first reference signal resources configured based on the first reference signal resource configuration information, and obtains signal measurement results for the reference signals transmitted on the first reference signal resources (e.g., obtains the RSRPs of the reference signals transmitted on the first reference signal resources), where the reference signals transmitted on the first reference signal resources correspond to a narrow beam universal set; (ii) The terminal device uses the RSRPs of the reference signals transmitted on the first reference signal resources as labels for performing model training on the AI ​​model; and (iii) The terminal device measures the reference signals transmitted on the third reference signal resources based on the third reference signal resource configured based on the third reference signal resource configuration information, and obtains signal measurement results for the reference signals transmitted on the third reference signal resources (e.g., obtains the RSRPs of the reference signals transmitted on the third reference signal resources). (iv) The terminal device uses the RSRP of the reference signal transmitted on the third reference signal resource as input information for model training for the AI ​​model. (v) The terminal device obtains output information of the AI ​​model, where the output information of the AI ​​model includes RSRPs predicted based on the AI ​​model for the reference signal corresponding to each beam in the narrow beam universal set. (vi) The terminal device calculates a loss function value based on the model training label and the RSRPs predicted based on the AI ​​model for the reference signal corresponding to each beam in the narrow beam universal set. If the loss function value is smaller than a preset threshold, it indicates that training of the AI ​​model is complete. If the loss function value is greater than or equal to the preset threshold, the terminal device performs model training for the AI ​​model by referring to steps (i) to (vi) until training of the AI ​​model is complete.

[0259] In the process in which the terminal device performs model training on the AI ​​model, the terminal device determines one or more optimal beams that can be used as labels in the model training process.

[0260] In yet another example, if the first operation is model inference for an AI model, the terminal device performs the first operation in the following steps based on the AI ​​model and first reference signal resource configuration information: (I) The terminal device measures a reference signal transmitted on a first reference signal resource based on a first reference signal resource configured based on the first reference signal resource configuration information, and obtains a signal measurement result of the reference signal transmitted on the first reference signal resource (e.g., obtains the RSRP of the reference signal transmitted on the first reference signal resource). (II) The terminal device uses the RSRP of the reference signal transmitted on the first reference signal resource as input information for model inference for the AI ​​model. (III) The terminal device obtains output information of the AI ​​model. The output information of the AI ​​model includes RSRPs of reference signals corresponding to each beam in the beam-universal set predicted based on the AI ​​model, or the output information of the AI ​​model includes a probability that each beam in the beam-universal set is an optimal beam.

[0261] S630b1: The terminal device transmits second information to the training device.

[0262] In response, the training device receives second information from the terminal device.

[0263] The second information includes a signal measurement result obtained by the terminal device by measuring a reference signal transmitted on the first reference signal resource, or the second information includes a dataset for performing a first operation based on the AI ​​model.

[0264] After the terminal device receives the first reference signal resource configuration information, if the first instruction information indicates that the first reference signal resource is for performing a first operation based on an AI model, or if the first instruction information indicates that the first reference signal resource is for acquiring a dataset, and the dataset is a dataset for performing a first operation based on an AI model, the terminal device acquires second information based on the first reference signal resource configuration information and transmits the second information to the training device.

[0265] For example, the terminal device obtains a signal measurement result by measuring a reference signal transmitted on a first reference signal resource, and transmits the signal measurement result to the training device.

[0266] If the first instruction information further indicates that the first reference signal resource is for obtaining labels for performing model training for an AI model, or if the first reference signal resource configuration information further includes fifth instruction information and the fifth instruction information indicates that the first reference signal resource is for obtaining labels for performing model training for an AI model, the second information further includes signal measurement results obtained by the terminal device by measuring a reference signal transmitted on the third reference signal resource.

[0267] In another example, the terminal device obtains signal measurement results by measuring a reference signal transmitted on a first reference signal resource, determines a dataset for performing a first operation based on the AI ​​model based on the signal measurement results, and transmits the dataset for performing the first operation based on the AI ​​model to the training device.

[0268] For example, if the first operation is model training for an AI model, the terminal device can refer to steps (1) to (3) or steps (a) to (c) to obtain labels and input information for performing model training for the AI ​​model, and thereby the terminal device sends the labels and input information for performing model training for the AI ​​model to the training device.

[0269] In another example, if the first operation is model inference on an AI model, the terminal device can refer to step (I) and step (II) to obtain input information for performing model inference on the AI ​​model, and then the terminal device sends the input information for performing model inference on the AI ​​model to the training device.

[0270] In another example, when the first instruction information further indicates that the first reference signal resource is for obtaining labels for performing model training on an AI model, or when the first reference signal resource configuration information further includes fifth instruction information, and the fifth instruction information indicates that the first reference signal resource is for obtaining labels for performing model training on an AI model, by referring to steps (A) to (D) or steps (i) to (iv), the terminal device can obtain labels and input information for performing model training on the AI ​​model, and then the terminal device sends the labels and input information for performing model training on the AI ​​model to the training device.

[0271] S630b2: The training device performs a first action based on the AI ​​model.

[0272] After receiving the second information from the terminal device, the training device can perform a first operation based on the AI ​​model and the second information.

[0273] For example, the first operation is model training for the AI ​​model, and the second information includes signal measurement results obtained by the terminal device by measuring a reference signal transmitted on the first reference signal resource. In this case, for the step of performing the first operation based on the AI ​​model by the training device, please refer to the description related to performing the first operation for the AI ​​model in steps (2) to (6) or steps (b) to (e).

[0274] In another example, the first operation is model training for the AI ​​model, and the second information includes signal measurement results obtained by the terminal device by measuring a reference signal transmitted on a first reference signal resource and a signal measurement result obtained by the terminal device by measuring a reference signal transmitted on a third reference signal resource. In this case, for the step of performing the first operation based on the AI ​​model by the training device, please refer to the description related to performing the first operation for the AI ​​model in step (B) and steps (D) to (G), or step (ii) and steps (iv) to (vi).

[0275] In another example, the first operation is model training for the AI ​​model, and the second information includes a dataset for performing model training for the AI ​​model. In this case, for the step of performing the first operation based on the AI ​​model by the training device, please refer to the description related to performing the first operation for the AI ​​model in steps (3) to (6), steps (c) to (e), steps (D) to (G), or steps (iv) to (vi).

[0276] In another example, if the first operation is model inference on an AI model, for the step of performing the first operation based on the AI ​​model by the training device, please refer to the description related to performing the first operation on the AI ​​model in step (II) and step (III).

[0277] Optionally, the method 600 further includes the terminal device receiving fourth reference signal resource configuration information from the network device, where the fourth reference signal resource configuration information includes ninth indication information, where the ninth indication information indicates that the fourth reference signal resource configured based on the fourth reference signal resource configuration information is for performing a non-AI model-based operation. The terminal device performs the non-AI model-based operation based on the fourth reference signal resource. The non-AI model-based operation may include measuring a signal quality of a reference signal transmitted on the fourth reference signal resource.

[0278] In this embodiment of the present application, the reference signal resource configuration information (e.g., first reference signal resource configuration information) transmitted by the network device to the terminal device includes instruction information (e.g., first instruction information) indicating the purpose of the reference signal resource. In this manner, the terminal device can identify the use of the reference signal resource based on the instruction information and perform a first operation on the AI ​​model based on the reference signal resource or perform a non-AI model-based operation based on the reference signal resource. If the terminal device can determine the use of the reference signal resource based on the instruction information included in the reference signal resource configuration information, the terminal device can prevent the terminal device from performing a non-AI model-based operation based on the reference signal resource for performing the first operation on the AI ​​model. This helps to balance the performance and resource consumption of performing model training on an AI model and / or the performance and resource consumption of performing model inference based on an AI model.

[0279] Referring to Figures 7 to 9, the following describes the communication method provided in the embodiments of the present application by using an example in which the AI ​​model is for beam management and the training of the AI ​​model can be completed on the terminal device.

[0280] 7 is a schematic flowchart of a communication method 700 according to an embodiment of the present application. The method 700 shown in FIG. 7 may include the following steps.

[0281] S710: The terminal device sends first information to the network device.

[0282] Correspondingly, in S710, the network device receives first information from the terminal device. The first information can be understood to be the expression of "first information" in S610.

[0283] The first information is for requesting that model training be performed on the AI ​​model, and the first information includes one or more of an identifier of the AI ​​model, computing capability information of the terminal device, sparse beam instruction information, or a training mechanism of the AI ​​model.

[0284] For further explanation of S710, please refer to S610. For the sake of brevity, the details will not be explained again here.

[0285] It should be further noted that S710 is an optional step. For example, whether the terminal device performs model training on the AI ​​model is determined by the network device, and S710 may not be performed in method 700.

[0286] S720: The network device sends second reference signal resource configuration information to the terminal device.

[0287] In response, the terminal device receives second reference signal resource configuration information from the network device.

[0288] The second reference signal resource configuration information includes second instruction information, where the second instruction information indicates that the second reference signal resource configured based on the second reference signal resource configuration information is for performing model training on the AI ​​model, or the second instruction information indicates that the second reference signal resource is for acquiring dataset #1, where dataset #1 is a dataset for performing model training on the AI ​​model. For a specific format of the second instruction information, see the description of the specific format of the first instruction information in S620. The reference signals transmitted on the second reference signal resource correspond to all beams supported by the network device.

[0289] Optionally, the model training described above may be replaced with model training inputs and labels.

[0290] The second reference signal resource configuration information may be carried in one or more of the following fields: a CSI reporting configuration field, a CSI resource configuration field, a CSI-RS resource set field, or a CSI-RS resource set list field.

[0291] Optionally, the second reference signal resource configuration information further includes fourth indication information indicating an association relationship between the first reference signal resource and the second reference signal resource.

[0292] If the second reference signal resource configuration information includes the fourth indication information, the terminal device can determine that the first reference signal resource and the second reference signal resource correspond to the same AI model based on the association relationship between the first reference signal resource and the second reference signal resource.

[0293] For example, the fourth indication information indicates the association relationship between the first reference signal resource and the second reference signal resource in one or more of the following manners:

[0294] For example, the fourth indication information indicates an identifier of a first reference signal resource having an association relationship with a second reference signal resource. The identifier of the first reference signal resource is an identifier (ID) of the first reference signal resource. Alternatively, the identifier of the first reference signal resource is a sequence number of the first reference signal resource in reference signal resource set #1. Alternatively, the identifier of the first reference signal resource is a first bitmap, the number of bits included in the first bitmap is the same as the number of reference signal resources included in reference signal resource set #1, and each bit included in the first bitmap corresponds to one reference signal resource in reference signal resource set #1. For example, when a bit indicates "0", the reference signal resource corresponding to the bit may be considered to be the first reference signal resource having an association relationship with the second reference signal resource. Alternatively, when a bit indicates "1", the reference signal resource corresponding to the bit may be considered to be the first reference signal resource having an association relationship with the second reference signal resource.

[0295] In another example, the fourth instruction information indicates an identifier of a reference signal resource set #1 having an association relationship with a second reference signal, and the reference signal resource set #1 includes a first reference signal resource having an association relationship with the second reference signal resource.

[0296] In another example, the fourth indication information indicates an identifier of first reference signal resource configuration information having an association relationship with the second reference signal resource. If the first reference signal resource configuration information is carried in a CSI reporting configuration field, the identifier of the first reference signal resource configuration information may be a report configuration identifier (ReportConfigId). If the first reference signal resource configuration information is carried in a CSI resource configuration field, the identifier of the first reference signal resource configuration information may be a resource configuration identifier (ResourceConfigId). If the first reference signal resource configuration information is carried in a CSI-RS resource set field, the identifier of the first reference signal resource configuration information may be a resource set identifier (ResourceSetId) field. If the first reference signal resource configuration information is carried in a CSI-RS resource set list field, the identifier of the first reference signal resource configuration information may be a resource set list identifier (ResourceSetListId).

[0297] In another example, the fourth instruction information indicates an identifier of an AI model corresponding to the second reference signal resource. The identifier of the AI ​​model indicated by the fourth instruction information is the same as the identifier of the AI ​​model indicated by the third instruction information, and the third instruction information indicates the identifier of the AI ​​model corresponding to the first reference signal resource, and the third instruction information is included in the first reference signal resource configuration information. If the identifier of the AI ​​model indicated by the third instruction information is the same as the identifier of the AI ​​model indicated by the fourth instruction information, the terminal device can determine that the first reference signal resource and the second reference signal resource correspond to the same AI model and that the first reference signal resource and the second reference signal resource have an association relationship.

[0298] S730: The terminal device performs model training on the AI ​​model.

[0299] After receiving the second reference signal resource configuration information, if the terminal device determines based on the second instruction information that the second reference signal resource configured based on the second reference signal resource configuration information is for performing model training on the AI ​​model, or if it determines based on the second instruction information that the second reference signal resource is for acquiring dataset #1, the terminal device performs model training on the AI ​​model based on the second reference signal resource configuration information.

[0300] For steps of performing model training on an AI model by a terminal device based on the second reference signal resource configuration information, please refer to steps (1) to (6) of S630a or steps (a) to (e) of S630a.

[0301] S740: The terminal device sends the third information to the network device.

[0302] In response, the network device receives third information from the terminal device. It can be understood that the third information is also an expression of “first information” in S610.

[0303] The third information is for requesting that the AI ​​model perform model inference, and the third information includes one or more of an identifier of the AI ​​model, instruction information #1, or sparse beam instruction information. The instruction information #1 indicates that training of the AI ​​model has been completed.

[0304] For further explanation of S740, see S610. For the sake of brevity, the details will not be explained again here.

[0305] It should be further noted that S740 is an optional step. For example, whether the terminal device performs model inference on the AI ​​model is determined by the network device, and S740 may not be performed in method 700.

[0306] S750: The network device sends first reference signal resource configuration information to the terminal device.

[0307] In response, the terminal device receives first reference signal resource configuration information from the network device.

[0308] The first reference signal resource configuration information may be carried in one or more of the following fields: a CSI reporting configuration field, a CSI resource configuration field, a CSI-RS resource set field, or a CSI-RS resource set list field.

[0309] The first reference signal resource configuration information includes first instruction information, and the first instruction information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for performing model inference on an AI model, or the first instruction information indicates that the first reference signal resource is for acquiring dataset #2, and dataset #2 is a dataset for performing model inference on an AI model. For a specific format of the first instruction information, see the description of S620. The reference signal transmitted on the first reference signal resource corresponds to a group of sparse beams.

[0310] Optionally, the first reference signal resource configuration information further includes third indication information indicating an association relationship between the first reference signal resource and the second reference signal resource.

[0311] If the first reference signal resource configuration information includes the third indication information, the terminal device can determine that the first reference signal resource and the second reference signal resource correspond to the same AI model based on the association relationship between the first reference signal resource and the second reference signal resource.

[0312] For example, the third indication information indicates the association relationship between the first reference signal resource and the second reference signal resource in one or more of the following manners:

[0313] For example, the third indication information indicates an identifier of a second reference signal resource having an association relationship with the first reference signal resource. The identifier of the second reference signal resource is an ID of the second reference signal resource. Alternatively, the identifier of the second reference signal resource is a sequence number of the second reference signal resource in reference signal resource set #2. Alternatively, the identifier of the second reference signal resource is a second bitmap, the number of bits included in the second bitmap is the same as the number of reference signal resources included in reference signal resource set #2, and each bit included in the second bitmap corresponds to one reference signal resource in reference signal resource set #2. For example, when a bit indicates "0", the reference signal resource corresponding to the bit may be considered to be the second reference signal resource having an association relationship with the first reference signal resource. Alternatively, when a bit indicates "1", the reference signal resource corresponding to the bit may be considered to be the second reference signal resource having an association relationship with the first reference signal resource. Optionally, in this manner, one of the fourth indication information or the third indication information may indicate an association relationship between the first reference signal resource and the second reference signal resource.

[0314] In another example, the third indication information indicates an identifier of a reference signal resource set #2 having an association relationship with the first reference signal, and the reference signal resource set #2 includes a second reference signal resource having an association relationship with the first reference signal resource. Optionally, in this manner, one of the fourth indication information or the third indication information may indicate an association relationship between the first reference signal resource and the second reference signal resource.

[0315] In another example, the third indication information indicates an identifier of second reference signal resource configuration information having an association relationship with the first reference signal resource. If the second reference signal resource configuration information is carried in a CSI reporting configuration field, the identifier of the second reference signal resource configuration information may be a report configuration identifier (ReportConfigId). If the second reference signal resource configuration information is carried in a CSI resource configuration field, the identifier of the second reference signal resource configuration information may be a resource configuration identifier (ResourceConfigId). If the second reference signal resource configuration information is carried in a CSI-RS resource set field, the identifier of the second reference signal resource configuration information may be a resource set identifier (ResourceSetId) field. If the second reference signal resource configuration information is carried in a CSI-RS resource set list field, the identifier of the second reference signal resource configuration information may be a resource set list identifier (ResourceSetListId). Optionally, in this scheme, one of the fourth indication information or the third indication information may indicate an association relationship between the first reference signal resource and the second reference signal resource.

[0316] In another example, the third instruction information indicates an identifier of an AI model corresponding to the first reference signal resource. The identifier of the AI ​​model indicated by the third instruction information is the same as the identifier of the AI ​​model indicated by the fourth instruction information, and the fourth instruction information indicates the identifier of the AI ​​model corresponding to the second reference signal resource, and the fourth instruction information is included in the second reference signal resource configuration information. If the identifier of the AI ​​model indicated by the third instruction information is the same as the identifier of the AI ​​model indicated by the fourth instruction information, the terminal device may determine that the first reference signal resource and the second reference signal resource correspond to the same AI model and that the first reference signal resource and the second reference signal resource have an association relationship. In this manner, the fourth instruction information and the third instruction information may together indicate the association relationship between the first reference signal resource and the second reference signal resource.

[0317] S760: The terminal device performs model inference based on the AI ​​model.

[0318] After receiving the first reference signal resource configuration information, if the terminal device determines based on the first instruction information that the first reference signal resource configured based on the first reference signal resource configuration information is for performing model inference on an AI model, or if it determines based on the first instruction information that the first reference signal resource is for acquiring dataset #2, the terminal device performs model inference on the AI ​​model based on the first reference signal resource configuration information.

[0319] For steps of performing model inference on the AI ​​model by the terminal device based on the first reference signal resource configuration information, please refer to steps (I) to (III) of S630a.

[0320] In this embodiment of the present application, the second reference signal resource configuration information sent by the network device to the terminal device includes second instruction information. In this way, the terminal device can determine that the second reference signal resource is for performing model training on the AI ​​model based on the second instruction information. This can prevent the terminal device from performing model inference on the AI ​​model based on the second reference signal resource, or prevent the terminal device from performing non-AI model-based operations based on the second reference signal resource, thereby helping to balance the performance of performing model training on the AI ​​model and resource consumption.

[0321] The first reference signal resource configuration information transmitted by the network device to the terminal device includes first instruction information. In this way, the terminal device can determine, based on the first instruction information, that the first reference signal resource is for performing model inference on an AI model. This can prevent the terminal device from performing model training on the AI ​​model based on the first reference signal resource, or prevent the terminal device from performing non-AI model-based operations based on the first reference signal resource, thereby helping to balance performance for performing model inference on an AI model and resource consumption.

[0322] 8 is a schematic flowchart of a communication method 800 according to an embodiment of the present application. The method 800 shown in FIG. 8 may include the following steps.

[0323] S810: The terminal device sends first information to the network device.

[0324] Correspondingly, in S810, the network device receives first information from the terminal device. The first information can be understood to be the expression of "first information" in S610.

[0325] The first information is for requesting that model training be performed on the AI ​​model, and the first information includes one or more of an identifier of the AI ​​model, computing capability information of the terminal device, sparse beam instruction information, or a training mechanism of the AI ​​model.

[0326] For further explanation of S810, please refer to S610. For the sake of brevity, the details will not be explained again here.

[0327] It should be noted that S810 is an optional step. For example, whether the terminal device performs model training on the AI ​​model is determined by the network device, and S810 may not be performed in the method 800.

[0328] S820: The network device sends first reference signal resource configuration information to the terminal device.

[0329] In response, the terminal device receives first reference signal resource configuration information from the network device.

[0330] The first reference signal resource configuration information may be carried in one or more of the following fields: a CSI reporting configuration field, a CSI resource configuration field, a CSI-RS resource set field, or a CSI-RS resource set list field.

[0331] The first reference signal resource configuration information includes first instruction information, and the first instruction information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for performing model training on the AI ​​model, or the first instruction information indicates that the first reference signal resource is for acquiring dataset #1, and dataset #1 is a dataset for performing model training on the AI ​​model. The reference signal transmitted on the first reference signal resource corresponds to all narrow beams supported by the network device, and all narrow beams supported by the network device are referred to as a narrow beam universal set.

[0332] The first instruction information further indicates that the first reference signal resource is for obtaining labels for performing model training on the AI ​​model. For a specific format of the first instruction information, please refer to the description of S620. Alternatively, the first reference signal resource configuration information further includes fifth instruction information, where the fifth instruction information indicates that the first reference signal resource is for obtaining labels for performing model training on the AI ​​model. For a specific format of the fifth instruction information, please refer to the description of S620.

[0333] Optionally, the first reference signal resource configuration information further includes seventh indication information indicating an association relationship between the first reference signal resource and the third reference signal resource.

[0334] If the first reference signal resource configuration information includes the seventh indication information, the terminal device can determine, based on the association relationship between the first reference signal resource and the third reference signal resource, that the first reference signal resource and the third reference signal resource correspond to the same AI model.

[0335] For example, the seventh indication information indicates the association relationship between the first reference signal resource and the third reference signal resource in one or more of the following manners:

[0336] For example, the seventh indication indicates a time offset value of the first reference signal resource relative to the third reference signal resource.

[0337] In another example, the seventh indication information indicates an identifier of a third reference signal resource having an association relationship with the first reference signal resource. The identifier of the third reference signal resource is an ID of the third reference signal resource. Alternatively, the identifier of the third reference signal resource is a sequence number of the third reference signal resource in reference signal resource set #3. Alternatively, the identifier of the third reference signal resource is a third bitmap, the number of bits included in the third bitmap is the same as the number of reference signal resources included in reference signal resource set #3, and each bit included in the third bitmap corresponds to one reference signal resource in reference signal resource set #3. For example, when a bit indicates "0", the reference signal resource corresponding to the bit may be considered to be the third reference signal resource having an association relationship with the first reference signal resource. Alternatively, when a bit indicates "1", the reference signal resource corresponding to the bit may be considered to be the third reference signal resource having an association relationship with the first reference signal resource.

[0338] In another example, the seventh instruction information indicates an identifier of a reference signal resource set #3 having an association relationship with the first reference signal resource, and the reference signal resource set #3 includes a third reference signal resource having an association relationship with the first reference signal resource.

[0339] In another example, the seventh instruction information indicates an identifier of an AI model corresponding to the first reference signal resource. The identifier of the AI ​​model indicated by the seventh instruction information is the same as the identifier of the AI ​​model indicated by the eighth instruction information, and the eighth instruction information indicates the identifier of the AI ​​model corresponding to the third reference signal resource, and the eighth instruction information is included in the third reference signal resource configuration information. If the identifier of the AI ​​model indicated by the seventh instruction information is the same as the identifier of the AI ​​model indicated by the eighth instruction information, the terminal device can determine that the first reference signal resource and the third reference signal resource correspond to the same AI model and that the first reference signal resource and the third reference signal resource have an association relationship.

[0340] Optionally, the first reference signal resource configuration information further includes indication information #2, where the indication information #2 indicates an association relationship between the first reference signal resource and the sixth reference signal resource. For further description of the indication information #2, please refer to the description of the third indication information in S750.

[0341] Optionally, when the seventh instruction information indicates an identifier of an AI model corresponding to the first reference signal resource, the identifier of the AI ​​model indicated by the seventh instruction information is the same as the identifier of the AI ​​model indicated by instruction information #3, and instruction information #3 indicates the identifier of the AI ​​model corresponding to the sixth reference signal resource, and instruction information #3 is included in the sixth reference signal resource configuration information.

[0342] S830: The network device sends third reference signal resource configuration information to the terminal device.

[0343] In response, the terminal device receives third reference signal resource configuration information from the network device.

[0344] The third reference signal resource configuration information includes sixth instruction information, which indicates that the third reference signal resource configured based on the third reference signal resource configuration information is for obtaining input information for performing model training on the AI ​​model. For a specific format of the sixth instruction information, see the description of S620. The reference signals transmitted on the third reference signal resource correspond to all wide beams supported by the network device, and all wide beams supported by the network device are referred to as a wide beam universal set.

[0345] The third reference signal resource configuration information may be carried in one or more of the following fields: a CSI reporting configuration field, a CSI resource configuration field, a CSI-RS resource set field, or a CSI-RS resource set list field.

[0346] Optionally, the third reference signal resource configuration information further includes eighth indication information indicating an association relationship between the first reference signal resource and the third reference signal resource.

[0347] If the third reference signal resource configuration information includes the eighth indication information, the terminal device can determine that the first reference signal resource and the third reference signal resource correspond to the same AI model based on the association relationship between the first reference signal resource and the third reference signal resource.

[0348] If the third reference signal resource configuration information includes the eighth instruction information, the terminal device may obtain labels and input information for performing model training for the AI ​​model based on the association relationship between the first reference signal resource and the third reference signal resource.

[0349] For example, the eighth indication information indicates the association relationship between the first reference signal resource and the third reference signal resource in one or more of the following manners:

[0350] For example, the eighth indication information indicates a time offset value of the third reference signal resource relative to the first reference signal resource. Optionally, in this manner, one of the seventh indication information or the eighth indication information may indicate an association relationship between the first reference signal resource and the third reference signal resource.

[0351] In another example, the eighth indication information indicates an identifier of a first reference signal resource having an association relationship with the third reference signal resource. For the identifier of the first reference signal resource, see the description of S750. Optionally, in this manner, one of the seventh indication information or the eighth indication information may indicate the association relationship between the first reference signal resource and the third reference signal resource.

[0352] In another example, the eighth indication information indicates an identifier of a reference signal resource set #1 having an association relationship with a third reference signal resource, and the reference signal resource set #1 includes a first reference signal resource having an association relationship with the third reference signal resource. Optionally, in this manner, one of the seventh indication information or the eighth indication information may indicate an association relationship between the first reference signal resource and the third reference signal resource.

[0353] In another example, the eighth indication information indicates an identifier of first reference signal resource configuration information having an association relationship with the third reference signal resource. For the identifier of the first reference signal resource configuration information, see the description of S720. Optionally, in this manner, one of the seventh indication information or the eighth indication information may indicate an association relationship between the first reference signal resource and the third reference signal resource.

[0354] In another example, the eighth instruction information indicates an identifier of an AI model corresponding to the third reference signal resource. The identifier of the AI ​​model indicated by the eighth instruction information is the same as the identifier of the AI ​​model indicated by the seventh instruction information, and the seventh instruction information indicates the identifier of the AI ​​model corresponding to the first reference signal resource, and the seventh instruction information is included in the first reference signal resource configuration information. If the identifier of the AI ​​model indicated by the eighth instruction information is the same as the identifier of the AI ​​model indicated by the seventh instruction information, the terminal device can determine that the first reference signal resource and the third reference signal resource correspond to the same AI model and that the first reference signal resource and the third reference signal resource have an association relationship. In this way, the seventh instruction information and the eighth instruction information may together indicate the association relationship between the first reference signal resource and the third reference signal resource.

[0355] Optionally, the third reference signal resource configuration information further includes indication information #4, where the indication information #4 indicates an association relationship between the third reference signal resource and the sixth reference signal resource. For further description of the indication information #4, please refer to the description of the third indication information in S750.

[0356] Optionally, when the eighth instruction information indicates an identifier of an AI model corresponding to the third reference signal resource, the identifier of the AI ​​model indicated by the eighth instruction information is the same as the identifier of the AI ​​model indicated by instruction information #3, and instruction information #3 indicates the identifier of the AI ​​model corresponding to the sixth reference signal resource, and instruction information #3 is included in the sixth reference signal resource configuration information.

[0357] S840: The terminal device performs model training on the AI ​​model.

[0358] After receiving the first reference signal resource configuration information and the third reference signal resource configuration information, if the terminal device determines, based on the first instruction information, that the first reference signal resource configured based on the first reference signal resource configuration information is for obtaining labels for performing model training on the AI ​​model, and determines, based on the sixth instruction information, that the third reference signal resource configured based on the third reference signal resource configuration information is for obtaining input information for performing model training on the AI ​​model, the terminal device performs model training on the AI ​​model based on the first reference signal resource configuration information and the third reference signal resource configuration information.

[0359] For steps of performing model training on the AI ​​model by the terminal device based on the first reference signal resource configuration information and the third reference signal resource configuration information, please refer to steps (A) to (G) of S630a or steps (i) to (iv) of S630a.

[0360] Optionally, the method 800 further includes S850 to S870.

[0361] S850: The terminal device transmits third information to the network device.

[0362] In response, the network device receives third information from the terminal device. It can be understood that the third information is also an expression of “first information” in S610.

[0363] The third information is for requesting that the AI ​​model perform model inference, and the third information includes one or more of an identifier of the AI ​​model, instruction information #1, or sparse beam instruction information. The instruction information #1 indicates that training of the AI ​​model has been completed.

[0364] For further explanation of S850, see S610. For the sake of brevity, the details will not be explained again here.

[0365] It should be noted that S850 is an optional step. For example, whether the terminal device performs model inference on the AI ​​model is determined by the network device, and S850 may not be performed in method 800.

[0366] S860: The network device sends sixth reference signal resource configuration information to the terminal device.

[0367] In response, the terminal device receives sixth reference signal resource configuration information from the network device.

[0368] The sixth reference signal resource configuration information may be carried in one or more of the following fields: a CSI reporting configuration field, a CSI resource configuration field, a CSI-RS resource set field, or a CSI-RS resource set list field.

[0369] The sixth reference signal resource configuration information includes instruction information #5, which indicates that the sixth reference signal resource configured based on the sixth reference signal resource configuration information is for performing model inference on an AI model, or the instruction information #5 indicates that the sixth reference signal resource is for acquiring dataset #2, which is a dataset for performing model inference on an AI model. The reference signal transmitted on the sixth reference signal resource corresponds to a wide beam universal set.

[0370] Optionally, the sixth reference signal resource configuration information further includes indication information #6, where the indication information #6 indicates an association relationship between the first reference signal resource and the sixth reference signal resource. For further description of the indication information #6, please refer to the description of the third indication information in S750.

[0371] Optionally, the sixth reference signal resource configuration information further includes indication information #7, where the indication information #7 indicates an association relationship between the third reference signal resource and the sixth reference signal resource. For further description of the indication information #7, please refer to the description of the third indication information in S750.

[0372] Optionally, the sixth reference signal resource configuration information further includes instruction information #3, where instruction information #3 indicates an identifier of an AI model corresponding to the sixth reference signal resource. The identifier of the AI ​​model indicated by instruction information #3 is the same as the identifier of the AI ​​model indicated by the seventh instruction information indicating the identifier of the AI ​​model corresponding to the first reference signal resource. Alternatively, the identifier of the AI ​​model indicated by instruction information #3 is the same as the identifier of the AI ​​model indicated by the eighth instruction information indicating the identifier of the AI ​​model corresponding to the third reference signal resource.

[0373] S870: The terminal device performs model inference based on the AI ​​model.

[0374] After receiving the sixth reference signal resource configuration information, if the terminal device determines based on the instruction information #5 that the sixth reference signal resource configured based on the sixth reference signal resource configuration information is for performing model inference on an AI model, or if it determines based on the instruction information #5 that the sixth reference signal resource is for acquiring dataset #2, the terminal device performs model inference on the AI ​​model based on the sixth reference signal resource configuration information.

[0375] For steps of performing model inference on the AI ​​model by the terminal device based on the sixth reference signal resource configuration information, please refer to steps (I) to (III) of S630a.

[0376] In this embodiment of the present application, even if the reference signal resource for acquiring labels for performing model training on the AI ​​model and the reference signal resource for acquiring input information for performing model training on the AI ​​model are different reference signal resources, the terminal device can still identify the purpose of the reference signal resource based on the instruction information (e.g., the first instruction information or the sixth instruction information) in the received reference signal resource configuration information to acquire labels and input information for performing model training on the AI ​​model. This helps to balance the performance of performing model training on the AI ​​model and resource consumption.

[0377] 9 is a schematic flowchart of a communication method 900 according to an embodiment of the present application. The method 900 shown in FIG. 9 may include the following steps.

[0378] S910: The terminal device sends first information to the network device #1.

[0379] In response, in S910, network device #1 receives first information from the terminal device.

[0380] The first information is for requesting that model inference be performed on the AI ​​model, and the first information includes one or more of an identifier of the AI ​​model, sparse beam instruction information, or instruction information #1. The instruction information #1 indicates that training of the AI ​​model has been completed.

[0381] For further explanation of S910, please refer to S610. For the sake of brevity, the details will not be explained again here.

[0382] It should be noted that S910 is an optional step. For example, whether the terminal device performs model training on the AI ​​model is determined by the network device, and S910 may not be performed in method 900.

[0383] S920: The network device #1 transmits first reference signal resource configuration information to the terminal device.

[0384] In response, the terminal device receives first reference signal resource configuration information from network device #1.

[0385] The first reference signal resource configuration information may be carried in one or more of the following fields: a CSI reporting configuration field, a CSI resource configuration field, a CSI-RS resource set field, or a CSI-RS resource set list field.

[0386] The first reference signal resource configuration information includes first instruction information, where the first instruction information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for performing model inference on an AI model, or the first instruction information indicates that the first reference signal resource is for acquiring dataset #2, where dataset #2 is a dataset for performing model inference on an AI model. For a specific format of the first instruction information, see the description of S620. The reference signal transmitted on the first reference signal resource corresponds to a group of sparse beams, or the reference signal transmitted on the first reference signal resource corresponds to all wide beams supported by network device #1, where all wide beams supported by network device #1 are referred to as a wide beam universal set.

[0387] For example, the first indication information is active state indication information of an AI model, and the first indication information indicates that the AI ​​model is in an active state or a usable state.

[0388] Optionally, the first reference signal resource configuration information further includes first identifier information. The first identifier information corresponds to a target reference signal resource set, and the target reference signal resource set includes reference signal resources corresponding to output information of the model inference of the AI ​​model, or the target reference signal resource set includes reference signal resources corresponding to labels in the training process of the AI ​​model, i.e., reference signal resources corresponding to all beams supported by the network device. All beams supported by the network device are different in different application scenarios. For example, different network devices support different numbers of all beams, and correspondingly, the AI ​​models obtained through training are also different. In the inference process, reference signal resources corresponding to all beams supported by the network device associated with this inference (e.g., the first reference signal resource) are indicated, so that the terminal device can perform inference based on the AI ​​model corresponding to the reference signal resources corresponding to all beams supported by the network device. The reference signal resource corresponding to the output information of the model inference of the AI ​​model is indicated as reference signal resource #1. The correspondence relationship between the output information of the model inference of the AI ​​model and reference signal resource #1 can be described as follows: The output information of the model inference of the AI ​​model represents a signal measurement result predicted based on the AI ​​model of the reference signal transmitted on reference signal resource #1. When the output information of the model inference of the AI ​​model corresponds to reference signal resource #1, the terminal device can determine the AI ​​model based on reference signal resource #1 or determine the AI ​​model based on the target reference signal resource set. Alternatively, the correspondence between the label and reference signal resource #1 in the model training process of the AI ​​model may be described as follows: Reference signal resource #1 is for obtaining the label in the model training process of the AI ​​model.

[0389] In a possible implementation, the first identifier information is for identifying a target reference signal resource set in a plurality of network devices or a plurality of cells, where the plurality of network devices includes a network device #1, and the plurality of cells includes a cell #1 in which the terminal device executes the AI ​​model in the network device #1. Optionally, the plurality of network devices belong to the same operator or the same device vendor. Optionally, the network devices to which the plurality of cells belong belong to the same operator or the same device vendor.

[0390] When the first identifier information is for identifying a target reference signal resource set in multiple network devices or multiple cells, the first identifier information may be referred to as a global identifier of the target reference signal resource set. For example, the first identifier information includes a sequence number of a target reference signal resource set in a first target reference signal resource set group, where the first target reference signal resource set group is a reference signal resource set group configured for multiple network devices or multiple cells.

[0391] In the following, an example of first reference signal resource configuration information and an example of resource configuration information for configuring a target reference signal resource set for multiple network devices or multiple cells will be described using an example in which the first reference signal resource is SetB and the target reference signal resource set is SetA.

[0392] Example A An example in which the first reference signal resource configuration information is carried in the CSI-ReportConfig field is shown below, and reportConfigId-SetA′ is an example of the first identifier information. CSI-ReportConfig ::= sequence(SEQUENCE) { reportConfigId-SetB Report configuration identifier (reportConfigId) reportConfigId-SetA' ...}

[0393] An example is given below in which resource configuration information for configuring target reference signal resource sets for multiple network devices or multiple cells is carried in the CSI-ReportConfig field. CSI-ReportConfig ::= SEQUENCE { reportConfigId-SetA' ...}

[0394] Example B An example in which the first reference signal resource configuration information is carried in the CSI-ResourceConfig field may be shown as follows, where ResourceConfigId-SetA′ is an example of the first identifier information: CSI-ResourceConfig ::= SEQUENCE { csi-ResourceConfigId-SetB Resource configuration identifier (CSI-ResourceConfigId) ResourceConfigId-SetA' csi-RS-ResourceSetList choice(CHOICE) { nzp-CSI-RS-SSB SEQUENCE { ...}

[0395] An example is given below in which resource configuration information for configuring target reference signal resource sets for multiple network devices or multiple cells is carried in the CSI-ResourceConfig field. CSI-ResourceConfig ::= SEQUENCE { ResourceConfigId-SetA' csi-RS-ResourceSetList CHOICE { nzp-CSI-RS-SSB SEQUENCE { ...}

[0396] Example C An example of the first reference signal resource configuration information being carried in the CSI-RS-ResourceSetList field may be shown as follows, where ResourceConfigId-SetA′ is an example of the first identifier information: csi-RS-ResourceSetList CHOICE { nzp-CSI-RS-SSB SEQUENCE { ResourceConfigId-SetB ResourceConfigId-SetA' ...}

[0397] An example is given below in which resource configuration information for configuring target reference signal resource sets for multiple network devices or multiple cells is carried in the CSI-RS-ResourceSetList field. csi-RS-ResourceSetList CHOICE { nzp-CSI-RS-SSB SEQUENCE { ResourceConfigId-SetA' ...}

[0398] It should be understood that when the first reference signal resource configuration information includes first identifier information, the terminal device may determine a target reference signal resource set configured for multiple network devices or multiple cells based on the first identifier information, determine an AI model corresponding to the target reference signal resource set based on the target reference signal resource set, and perform model inference based on the AI ​​model corresponding to the target reference signal resource set.

[0399] In a possible implementation, the first identifier information is for identifying a target reference signal resource set in network device #1 or multiple cells #1, where cell #1 is a cell in which a terminal device executes an AI model in network device #1. When the first identifier information is for identifying a target reference signal resource set in network device #1 or cell #1, the first identifier information may be referred to as a local identifier of the target reference signal resource set. For example, the first identifier information includes a sequence number of a target reference signal resource set in a second target reference signal resource set group, where the second target reference signal resource set group is a reference signal resource set group configured for network device #1 or cell #1.

[0400] When the first identifier information is a local identifier of the target reference signal resource set, method 900 further includes S930, in which the network device transmits fifth reference signal resource configuration information to the terminal device, the fifth reference signal resource configuration information being for configuring the target reference signal resource set in network device #1 or cell #1, the fifth reference signal resource configuration information including the first identifier information and second identifier information, the second identifier information being for identifying the target reference signal resource set in multiple network devices or multiple cells, i.e., the second identifier information being a global identifier of the target reference signal resource set. For example, the second identifier information includes a sequence number of the target reference signal resource set in a first target reference signal resource set group, the first target reference signal resource set group being a reference signal resource set group configured for multiple network devices or multiple cells.

[0401] In the following, an example of the first reference signal resource configuration information, an example of the fifth reference signal resource configuration information, and an example of resource configuration information for configuring a target reference signal resource set for multiple network devices or multiple cells will be described using an example in which the first reference signal resource is SetB and the target reference signal resource set is SetA.

[0402] Example 1 An example in which the first reference signal resource configuration information is carried in the CSI-ReportConfig field is shown below, and reportConfigId-SetA is an example of the first identifier information. CSI-ReportConfig ::= SEQUENCE { reportConfigId-SetB reportConfigId-SetA ...}

[0403] An example in which the fifth reference signal resource configuration information is carried in the CSI-ReportConfig field is shown below, where reportConfigId-SetA is an example of the first identifier information and reportConfigId-SetA' is an example of the second identifier information. CSI-ReportConfig ::= SEQUENCE { reportConfigId-SetA reportConfigId-SetA' ...}

[0404] An example is shown below in which resource configuration information for configuring target reference signal resource sets for multiple network devices or multiple cells is carried in the CSI-ReportConfig field, where reportConfigId-SetA′ is an example of second identifier information. CSI-ReportConfig ::= SEQUENCE { reportConfigId-SetA' ...}

[0405] Example 2 An example in which the first reference signal resource configuration information is carried in the CSI-ResourceConfig field may be shown as follows, where ResourceConfigId-SetA is an example of the first identifier information: CSI-ResourceConfig ::= SEQUENCE { csi-ResourceConfigId-SetB ResourceConfigId-SetA csi-RS-ResourceSetList CHOICE { nzp-CSI-RS-SSB SEQUENCE { ...}

[0406] An example in which fifth reference signal resource configuration information is carried in the CSI-ResourceConfig field is shown below, where ResourceConfigId-SetA is an example of first identifier information and ResourceConfigId-SetA' is an example of second identifier information. CSI-ResourceConfig ::= SEQUENCE { ResourceConfigId-SetA ResourceConfigId-SetA' csi-RS-ResourceSetList CHOICE { nzp-CSI-RS-SSB SEQUENCE { ...}

[0407] An example is shown below in which resource configuration information for configuring target reference signal resource sets for multiple network devices or multiple cells is carried in the CSI-ResourceConfig field, and ResourceConfigId-SetA′ is an example of second identifier information. CSI-ResourceConfig ::= SEQUENCE { ResourceConfigId-SetA' csi-RS-ResourceSetList CHOICE { nzp-CSI-RS-SSB SEQUENCE { ...}

[0408] Example 3 An example of the first reference signal resource configuration information being carried in the CSI-RS-ResourceSetList field may be shown as follows, where ResourceConfigId-SetA is an example of the first identifier information: csi-RS-ResourceSetList CHOICE { nzp-CSI-RS-SSB SEQUENCE { ResourceConfigId-SetB ResourceConfigId-SetA ...}

[0409] An example in which fifth reference signal resource configuration information is carried in the CSI-RS-ResourceSetList field is shown below, where ResourceConfigId-SetA is an example of first identifier information and ResourceConfigId-SetA' is an example of second identifier information. csi-RS-ResourceSetList CHOICE { nzp-CSI-RS-SSB SEQUENCE { ResourceConfigId-SetA ResourceConfigId-SetA' ...}

[0410] An example is shown below in which resource configuration information for configuring target reference signal resource sets for multiple network devices or multiple cells is carried in the CSI-RS-ResourceSetList field, where ResourceConfigId-SetA is an example of second identifier information. csi-RS-ResourceSetList CHOICE { nzp-CSI-RS-SSB SEQUENCE { ResourceConfigId-SetA' ...}

[0411] It should be understood that if the first reference signal resource configuration information includes first identifier information and the fifth reference signal resource configuration information includes first identifier information and second identifier information, the terminal device can determine an association relationship between the first reference signal resource configuration information and the fifth reference signal resource configuration information based on the first identifier information. In this manner, the terminal device can determine a target reference signal resource set configured for multiple network devices or multiple cells based on the second identifier information included in the fifth reference signal resource configuration information, and determine that the target reference signal resource set includes reference signal resources corresponding to the output information of the model inference of the AI ​​model. Finally, the terminal device can determine an AI model corresponding to the target reference signal resource set based on the target reference signal resource set, and perform model inference based on the AI ​​model corresponding to the target reference signal resource set.

[0412] S940: The terminal device performs model inference based on the AI ​​model.

[0413] After receiving the first reference signal resource configuration information, if the terminal device determines based on the first instruction information that the first reference signal resource configured based on the first reference signal resource configuration information is for performing model inference on an AI model, or if it determines based on the first instruction information that the first reference signal resource is for acquiring dataset #2, the terminal device performs model inference on the AI ​​model based on the first reference signal resource configuration information.

[0414] For steps of performing model inference on the AI ​​model by the terminal device based on the first reference signal resource configuration information, please refer to steps (I) to (III) of S630a.

[0415] Optionally, the method 900 further includes S950 to S970.

[0416] S950: The terminal device transmits information about the K beams to the network device #1.

[0417] In response, network device #1 receives information about the K beams from the terminal device.

[0418] After performing model inference based on the AI ​​model and obtaining the output information of the AI ​​model, the terminal device may determine K beams based on the output information of the AI ​​model, where K is a positive integer.

[0419] For example, if the output information of the AI ​​model includes the RSRP of the reference signal corresponding to each beam in the beam universal set predicted by the AI ​​model, the terminal device may select K beams having the maximum RSRP of the reference signal corresponding to the beam universal set.

[0420] In another example, if the output information of the AI ​​model includes the probability that each beam in the beam universal set is an optimal beam, the terminal device may select K beams in the beam universal set that have the highest probability of being optimal beams.

[0421] The value of K may be predefined in the protocol or may be indicated to the terminal device by network device #1, which is not limited in this embodiment of the present application.

[0422] S960: The network device #1 transmits the fourth reference signal resource configuration information to the terminal device.

[0423] In response, the terminal device receives fourth reference signal resource configuration information from the network device.

[0424] The fourth reference signal resource configuration information may be carried in one or more of the following fields: a CSI reporting configuration field, a CSI resource configuration field, a CSI-RS resource set field, or a CSI-RS resource set list field.

[0425] The fourth reference signal resource configuration information includes ninth instruction information, and the ninth instruction information indicates that the fourth reference signal resource configured based on the fourth reference signal resource configuration information is for performing a non-AI model-based operation. Network device #1 transmits the fourth reference signal resource configuration information to the terminal device based on information about the K beams, and the reference signals transmitted on the fourth reference signal resource correspond to the K beams.

[0426] For example, the ninth instruction information is active state instruction information of the AI ​​model, and the ninth instruction information indicates that the AI ​​model is in a deactive state or a disabled state.

[0427] S970: The terminal device performs a non-AI model-based operation.

[0428] After receiving the fourth reference signal resource configuration information, if the terminal device determines based on the ninth instruction information that the fourth reference signal resource configured based on the fourth reference signal resource configuration information is for performing a non-AI model-based operation, the terminal device performs a non-AI model-based operation based on the fourth reference signal resource configuration information.

[0429] The terminal device performs a non-AI model-based operation based on the fourth reference signal resource configuration information in the following steps: Based on the fourth reference signal resource configured based on the fourth reference signal resource configuration information, the terminal device measures the reference signal transmitted on the fourth reference signal resource and obtains a signal measurement result of the reference signal transmitted on the fourth reference signal resource (e.g., obtains the RSRP of the reference signal transmitted on the fourth reference signal resource). Based on the RSRP of the reference signal transmitted on the fourth reference signal resource, the terminal device determines, as an optimal beam, a beam corresponding to a reference signal having the largest RSRP among the reference signals transmitted on the fourth reference signal resource. The optimal beam may include one or more beams.

[0430] In this embodiment of the present application, the first reference signal resource configuration information sent by the network device to the terminal device includes first instruction information. In this way, the terminal device can determine that the first reference signal resource is for performing model inference on an AI model based on the first instruction information. This can prevent the terminal device from performing model training on the AI ​​model based on the first reference signal resource, or prevent the terminal device from performing non-AI model-based operations based on the first reference signal resource, thereby helping to balance the performance of performing model inference on the AI ​​model and resource consumption.

[0431] It may be further understood that some optional features in the embodiments of the present application may be independent of other features in some scenarios or may be combined with other features in some scenarios, without limitation.

[0432] It may be further understood that in some of the above-described embodiments, transmitting information is referred to multiple times. For example, A transmits information to B. A transmitting information to B may include A transmitting information directly to B, or may include A transmitting information to B through another device or network element. This is not limited thereto.

[0433] It can be further understood that the solutions in the embodiments of the present application may be appropriately combined for use, and the explanations or descriptions of terms in the embodiments may be mutually referenced or explained in the embodiments. This is not limited thereto.

[0434] It may be further understood that in the method embodiments described above, the methods and operations implemented by the network element may alternatively be implemented by components (such as chips or circuits) that make up the network element, without this being limiting.

[0435] The above describes in detail the method provided in the embodiment of the present application with reference to Figures 6 to 9. The device provided in the embodiment of the present application will be described in detail below with reference to Figures 10 to 12. It should be understood that the description of the device embodiment corresponds to the description of the method embodiment. Therefore, for the contents not described in detail, please refer to the method embodiment described above. For the sake of brevity, the details will not be described again here.

[0436] 10 is a diagram of a communication device 1000 according to an embodiment of the present application. The device 1000 includes a transceiver unit 1010 and a processing unit 1020. The transceiver unit 1010 may be configured to implement corresponding communication functions. The transceiver unit 1010 may also be referred to as a communication interface or a communication unit. The processing unit 1020 may be configured to perform processing, for example, to perform a first operation based on an AI model.

[0437] Optionally, the apparatus 1000 may further include a storage unit. The storage unit may be configured to store instructions and / or data. The processing unit 1020 may read the instructions and / or data in the storage unit so that the apparatus implements the above-described method embodiments.

[0438] In design, the apparatus 1000 is configured to perform steps or procedures performed by the device in the above-mentioned method embodiments, the transceiver unit 1010 is configured to perform transmission and reception-related operations on the part of the device in the above-mentioned method embodiments, and the processing unit 1020 is configured to perform processing-related operations on the part of the device in the above-mentioned method embodiments.

[0439] In a possible implementation, the apparatus 1000 is configured to perform the steps or procedures performed by a terminal device in the embodiments shown in Figures 6 to 9. Optionally, the transceiver unit 1010 is configured to receive first reference signal resource configuration information from the network device. The first reference signal resource configuration information includes first indication information, where the first indication information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for performing a first operation based on an AI model, or the first indication information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for acquiring a dataset for the first operation based on the AI ​​model. The processing unit 1020 is configured to perform the first operation based on the AI ​​model and the first reference signal resource configuration information.

[0440] 6 to 9. Optionally, the transceiver unit 1010 is configured to transmit first reference signal resource configuration information to the terminal device. The first reference signal resource configuration information includes first indication information, wherein the first indication information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for performing a first operation based on an AI model, or the first indication information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for acquiring a dataset for the first operation based on the AI ​​model.

[0441] It should be understood that the specific processes by which the units perform the corresponding steps described above have been described in detail in the above method embodiments, and for the sake of brevity, the details will not be described herein.

[0442] It should be understood that the apparatus 1000 herein is embodied in the form of a functional unit. The term "unit" herein may refer to an application-specific integrated circuit (ASIC), an electronic circuit, a processor (e.g., a shared processor, a dedicated processor, or a group processor) configured to execute one or more software or firmware programs, a memory, a merged logic circuit, and / or another appropriate component supporting the described functions. In an optional example, those skilled in the art may understand that the apparatus 1000 may specifically be a device (e.g., a terminal device or a network device) in the above-described embodiments, or may be configured to perform procedures and / or steps corresponding to the device in the above-described method embodiments. To avoid repetition, details will not be described again here.

[0443] The apparatus 1000 in the above solution has functions for implementing corresponding steps performed by a device (e.g., a terminal device or a network device) in the above method. The functions may be implemented by hardware, or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above functions. For example, a transceiver unit may be replaced by a transceiver (e.g., a transmitting unit in a transceiver unit may be replaced by a transmitter, and a receiving unit in a transceiver unit may be replaced by a receiver), and another unit, e.g., a processing unit, may be replaced by a processor to separately perform receiving and transmitting operations and related processing operations in the method embodiments.

[0444] Furthermore, the transceiver unit 1010 may alternatively be a transceiver circuit (eg, the transceiver circuit may include a receiver circuit and a transmitter circuit), and the processing unit may be a processing circuit.

[0445] It should be noted that the apparatus in Fig. 10 may be the device of the above-mentioned embodiment, or may be a chip or a chip system, for example, a system on chip (SoC). The transceiver unit may be an input / output circuit or a communication interface. The processing unit may be a processor, a microprocessor, or an integrated circuit integrated on a chip. This is not limited to this specification.

[0446] 11 is a diagram of another communication device 1100 according to an embodiment of the present application. The device 1100 includes a processor 1110. The processor 1110 is coupled to a memory 1120. The memory 1120 is configured to store computer programs or instructions and / or data. The processor 1110 is configured to execute computer programs or instructions stored in the memory 1120 or read data stored in the memory 1120 to perform the methods in the above-described method embodiments.

[0447] Optionally, there are one or more processors 1110 .

[0448] Optionally, one or more memories 1120 are present.

[0449] Optionally, the memory 1120 and the processor 1110 are integrated together or located separately.

[0450] 11, the apparatus 1100 further includes a transceiver 1130. The transceiver 1130 is configured to receive and / or transmit signals. For example, the processor 1110 is configured to control the transceiver 1130 to receive and / or transmit signals.

[0451] In one example, the processor 1110 may have the functionality of the processing unit 1020 shown in FIG. 10, the memory 1120 may have the functionality of the storage unit, and the transceiver 1130 may have the functionality of the transceiver unit 1010 shown in FIG.

[0452] In the solution, the apparatus 1100 is configured to implement the operations performed by a device (eg, a terminal device or a network device) in the above-described method embodiments.

[0453] For example, the processor 1110 is configured to execute computer programs or instructions stored in the memory 1120 to implement relevant operations of a device (e.g., a terminal device or a network device) in the above-described method embodiments.

[0454] It should be understood that the processor referred to in the embodiments of the present application may be a central processing unit (CPU), or may further be another general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, etc. The general-purpose processor may be a microprocessor, or the processor may be any conventional processor, etc.

[0455] It should be further understood that the memory referred to in the embodiments of the present application may be volatile memory and / or non-volatile memory. The non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory may be random access memory (RAM). For example, RAM may be used as an external cache. By way of example and not limitation, RAM includes multiple forms such as static random access memory (static RAM, SRAM), dynamic random access memory (dynamic RAM, DRAM), synchronous dynamic random access memory (synchronous DRAM, SDRAM), double data rate synchronous dynamic random access memory (double data rate SDRAM, DDR SDRAM), enhanced synchronous dynamic random access memory (enhanced SDRAM, ESDRAM), synchlink dynamic random access memory (synchlink DRAM, SLDRAM), and direct rambus random access memory (direct rambus RAM, DR RAM).

[0456] It should be noted that when the processor is a general-purpose processor, a DSP, an ASIC, an FPGA or other programmable logic device, a discrete gate or transistor logic device, or a discrete hardware component, the memory (storage module) may be integrated into the processor.

[0457] The apparatus in FIG. 11 may be the device of the above-described embodiment, or may be a chip or a chip system, for example, a system on chip (SoC). The transceiver may be an input / output circuit or a communication interface. The processor may be a processor, a microprocessor, or an integrated circuit integrated on a chip. This is not limited to this specification.

[0458] It should be further noted that memory as described herein is intended to include, without being limited to, these and any other suitable types of memory.

[0459] 12 is a diagram of a chip system 1200 according to one embodiment of the present application. The chip system 1200 (sometimes referred to as a processing system) includes a logic circuit 1210 and an input / output interface 1220.

[0460] The logic circuit 1210 may be a processing circuit within the chip system 1200. The logic circuit 1210 may be coupled and connected to a storage unit and call instructions in the storage unit, thereby enabling the chip system 1200 to implement the methods and functions in the embodiments of the present application. The input / output interface 1220 may be an input / output circuit within the chip system 1200, which outputs information processed by the chip system 1200 or inputs data or signaling information to be processed into the chip system 1200 for processing.

[0461] In particular, for example, when chip system 1200 is installed on a terminal device, logic circuit 1210 is coupled to input / output interface 1220, which can input the first reference signal resource configuration information to logic circuit 1210 for processing, for example, to perform a first operation based on the AI ​​model and the first reference signal resource configuration information. In another example, when chip system 1200 is installed on a network device, logic circuit 1210 is coupled to input / output interface 1220, which can output the first reference signal resource configuration information to the terminal device.

[0462] In the solution, the chip system 1200 is configured to implement the operations performed by a device (eg, a terminal device or a network device) in the above-described method embodiments.

[0463] For example, the logic circuit 1210 is configured to implement processing-related operations performed by a device (e.g., a terminal device or a network device) in the above-described method embodiments, and the input / output interface 1220 is configured to implement transmission and / or reception-related operations / operations performed by a device (e.g., a terminal device or a network device) in the above-described method embodiments.

[0464] An embodiment of the present application further provides a computer-readable storage medium, which stores computer instructions for implementing the method executed by a device (e.g., a terminal device or a network device) in the above-described method embodiments.

[0465] For example, when the computer program is executed by a computer, the computer becomes capable of implementing the method executed by a device (for example, a terminal device or a network device) in the above-described method embodiments.

[0466] An embodiment of the present application further provides a computer program product including instructions, which, when executed by a computer, implement the method performed by a device (e.g., a terminal device or a network device) in the above-described method embodiments.

[0467] An embodiment of the present application further provides a communication system, which includes the terminal device and / or the network device in the above-mentioned embodiments. For example, the system includes the terminal device and / or the network device in the embodiments shown in Figures 6 to 9.

[0468] For the description of the relevant contents and beneficial effects of any one of the above-provided devices, please refer to the corresponding method embodiments provided above, and the details will not be described again here.

[0469] In some embodiments provided in this application, it should be understood that the disclosed devices and methods may be implemented in other manners. For example, the described device embodiments are merely examples. For example, the division into units is merely a logical division of function, and other divisions may be used in actual implementation. For example, multiple units or components may be combined or integrated into other systems, or some features may be ignored or not implemented. Furthermore, the shown or described mutual couplings or direct couplings or communication connections may be implemented through some interfaces. Indirect couplings or communication connections between devices or units may be implemented in electronic, mechanical, or other forms.

[0470] All or part of the above-described embodiments may be implemented using software, hardware, firmware, or any combination thereof. When software is used to implement the embodiments, all or part of the embodiments may be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded into a computer and executed, the procedures or functions according to the embodiments of the present application are generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. For example, the computer may be a personal computer, a server, or a network device. The computer instructions may be stored in a computer-readable storage medium or transmitted from a computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, optical fiber, or digital subscriber line (DSL)) or wireless (e.g., infrared, radio, or microwave) transmission. The computer-readable storage medium may be any available medium accessible by a computer, or a data storage device integrating one or more available media, such as a server or a data center. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, or a magnetic tape), an optical medium (e.g., a DVD), a semiconductor medium (e.g., a solid-state drive (SSD)), etc. For example, the available medium may include, but is not limited to, any medium capable of storing program code, such as a USB flash drive, a removable hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0471] The above description is merely a specific implementation of the present application and is not intended to limit the scope of protection of the present application. Any modifications or substitutions that are easily understood by those skilled in the art within the technical scope disclosed in the present application shall fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be subject to the scope of protection of the claims. [Explanation of symbols]

[0472] 100 Radio access network, wireless communication system 600 Communication Methods 700 Communication Methods 800 Communication Methods 900 Communication Methods 1000 Communication Equipment 1010 Transceiver Unit 1020 Processing Unit 1100 Communication equipment 1110 processor 1120 memory 1130 transceiver 1200 Chip System 1210 Logic Circuits 1220 Input / Output Interface

Claims

1. 1. A communication method performed by a terminal device or a chip used in a terminal device, comprising: receiving first reference signal resource configuration information from a network device, the first reference signal resource configuration information including first instruction information, wherein the first instruction information indicates that a first reference signal resource configured based on the first reference signal resource configuration information is for performing a first operation based on an artificial intelligence (AI) model, or the first instruction information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for acquiring a dataset for a first operation based on an artificial intelligence (AI) model; A method comprising:

2. performing the first operation based on the AI ​​model and the first reference signal resource configuration information; 10. The method of claim 1, further comprising:

3. The method of claim 1 or 2, wherein the first operation is model training or model inference for the AI ​​model.

4. The first operation is model inference for the AI ​​model, and the method includes: receiving second reference signal resource configuration information from the network device, the second reference signal resource configuration information including second instruction information, wherein the second instruction information indicates that a second reference signal resource configured based on the second reference signal resource configuration information is for performing model training for the AI ​​model, or the second instruction information indicates that a second reference signal resource configured based on the second reference signal resource configuration information is for acquiring a dataset for performing model training for the AI ​​model; 4. The method of claim 3, further comprising:

5. The first reference signal resource configuration information further includes third indication information indicating an association relationship between the first reference signal resource and the second reference signal resource; or The method of claim 4 , wherein the second reference signal resource configuration information further includes fourth indication information indicating an association relationship between the second reference signal resource and the first reference signal resource.

6. the first operation is model inference for the AI ​​model; The first reference signal resource is a reference signal resource corresponding to input information for model inference of the AI ​​model; 4. The method of claim 3, wherein the first reference signal resource configuration information further includes first identifier information, the first identifier information corresponding to a target reference signal resource set, and the target reference signal resource set including reference signal resources corresponding to labels in a model training process of the AI ​​model.

7. 4. The method of claim 3, wherein the first operation is model training for the AI ​​model, and the first instruction information further indicates that the first reference signal resource is for obtaining labels for performing model training for the AI ​​model.

8. 4. The method of claim 3, wherein the first reference signal resource configuration information further includes fifth instruction information, and the fifth instruction information indicates that the first reference signal resource is for obtaining labels for performing model training on the AI ​​model.

9. receiving third reference signal resource configuration information from the network device, the third reference signal resource configuration information including sixth instruction information, the sixth instruction information indicating that a third reference signal resource configured based on the third reference signal resource configuration information is for obtaining input information for performing model training on the AI ​​model; 9. The method of claim 7 or 8, further comprising:

10. The first reference signal resource configuration information further includes seventh indication information indicating an association relationship between the first reference signal resource and the third reference signal resource; or The third reference signal resource configuration information further includes eighth indication information indicating an association relationship between the first reference signal resource and the third reference signal resource; The method comprises: obtaining the input information and the label based on the association relationship between the first reference signal resource and the third reference signal resource; 10. The method of claim 9, further comprising:

11. receiving fourth reference signal resource configuration information from the network device, the fourth reference signal resource configuration information including ninth indication information, the ninth indication information indicating that a fourth reference signal resource configured based on the fourth reference signal resource configuration information is for performing a non-AI model-based operation; performing, by the terminal device, the non-AI model-based operation based on the fourth reference signal resource configuration information; 11. The method of any one of claims 1 to 10, further comprising:

12. 4. The method of claim 3, wherein the first operation is model training for the AI ​​model, and the first instruction information further indicates that the first reference signal resource is for obtaining input information for performing model training for the AI ​​model.

13. 4. The method of claim 3, wherein the first operation is model inference based on the AI ​​model, and the first instruction information further indicates that the first reference signal resource is for obtaining input information for performing model inference on the AI ​​model.

14. 1. A communication method comprising: transmitting first reference signal resource configuration information to a terminal device, the first reference signal resource configuration information including first instruction information, wherein the first instruction information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for performing a first operation based on an artificial intelligence (AI) model, or the first instruction information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for acquiring a dataset of a first operation based on an artificial intelligence (AI) model; A method comprising:

15. The method of claim 14 , wherein the first operation is model training or model inference for the AI ​​model.

16. The first operation is model inference for the AI ​​model, and the method includes: transmitting second reference signal resource configuration information to the terminal device, the second reference signal resource configuration information including second instruction information, wherein the second instruction information indicates that the second reference signal resource configured based on the second reference signal resource configuration information is for performing model training for the AI ​​model, or the second instruction information indicates that the second reference signal resource configured based on the second reference signal resource configuration information is for acquiring a dataset for performing model training for the AI ​​model; 16. The method of claim 15, further comprising:

17. The first reference signal resource configuration information further includes third indication information indicating an association relationship between the first reference signal resource and the second reference signal resource; or The method of claim 16 , wherein the second reference signal resource configuration information further includes fourth indication information indicating an association relationship between the second reference signal resource and the first reference signal resource.

18. the first operation is model inference for the AI ​​model; The first reference signal resource is a reference signal resource corresponding to input information for model inference of the AI ​​model; 16. The method of claim 15, wherein the first reference signal resource configuration information further includes first identifier information, the first identifier information corresponding to a target reference signal resource set, and the target reference signal resource set including reference signal resources corresponding to labels in a model training process of the AI ​​model.

19. transmitting fifth reference signal resource configuration information to the terminal device, the fifth reference signal resource configuration information being for configuring the target reference signal resource set, the fifth reference signal resource configuration information including the first identifier information, the fifth reference signal resource configuration information further including second identifier information, and the second identifier information corresponding to the target reference signal resource set; 20. The method of claim 18, further comprising:

20. 16. The method of claim 15, wherein the first operation is model training for the AI ​​model, and the first instruction information further indicates that the first reference signal resource is for obtaining labels for performing model training for the AI ​​model.

21. 16. The method of claim 15, wherein the first reference signal resource configuration information further includes fifth instruction information, and the fifth instruction information indicates that the first reference signal resource is for obtaining labels for performing model training on the AI ​​model.

22. transmitting third reference signal resource configuration information to the terminal device, the third reference signal resource configuration information including sixth instruction information, the sixth instruction information indicating that a third reference signal resource configured based on the third reference signal resource configuration information is for obtaining input information for performing model training on the AI ​​model; 22. The method of claim 20 or 21, further comprising:

23. The first reference signal resource configuration information further includes seventh indication information indicating an association relationship between the first reference signal resource and the third reference signal resource; or The method of claim 22 , wherein the third reference signal resource configuration information further includes eighth indication information indicating an association relationship between the first reference signal resource and the third reference signal resource.

24. transmitting fourth reference signal resource configuration information to the terminal device, the fourth reference signal resource configuration information including ninth indication information, the ninth indication information indicating that a fourth reference signal resource configured based on the fourth reference signal resource configuration information is for performing a non-AI model-based operation; 24. The method of any one of claims 14 to 23, further comprising:

25. 16. The method of claim 15, wherein the first operation is model training for the AI ​​model, and the first instruction information further indicates that the first reference signal resource is for obtaining input information for performing model training for the AI ​​model.

26. 16. The method of claim 15, wherein the first operation is model inference for the AI ​​model, and the first instruction information further indicates that the first reference signal resource is for obtaining input information for performing model inference on the AI ​​model.

27. receiving first information from the terminal device, the first information being for requesting the first action based on the AI ​​model; 27. The method of any one of claims 14 to 26, further comprising:

28. 1. A communication method comprising: receiving first reference signal resource configuration information from a network device, the first reference signal resource configuration information including first instruction information, wherein the first instruction information indicates that a first reference signal resource configured based on the first reference signal resource configuration information is for performing model training on an AI model, or the first instruction information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for acquiring a dataset for performing model training on an AI model; A method comprising:

29. performing model training on the AI ​​model based on the first reference signal resource configuration information; 29. The method of claim 28, further comprising:

30. 30. The method of claim 28 or 29, wherein the first instruction information further indicates that the first reference signal resource is for obtaining labels for performing model training on the AI ​​model.

31. 30. The method of claim 28 or 29, wherein the first reference signal resource configuration information further includes fifth instruction information, and the fifth instruction information indicates that the first reference signal resource is for obtaining labels for performing model training on the AI ​​model.

32. receiving third reference signal resource configuration information from the network device, the third reference signal resource configuration information including sixth instruction information, the sixth instruction information indicating that a third reference signal resource configured based on the third reference signal resource configuration information is for obtaining input information for performing model training on the AI ​​model; 32. The method of any one of claims 28 to 31, further comprising:

33. The first reference signal resource configuration information further includes seventh indication information indicating an association relationship between the first reference signal resource and the third reference signal resource; or 33. The method of claim 32, wherein the third reference signal resource configuration information further includes eighth indication information indicating an association relationship between the first reference signal resource and the third reference signal resource.

34. obtaining the input information and the label based on the association relationship between the first reference signal resource and the third reference signal resource; 34. The method of claim 32 or 33, further comprising:

35. sending first information to the network device, the first information being for requesting to perform model training based on the AI ​​model, the first information including one or more of an identifier of the AI ​​model, computing capability information of the terminal device, sparse beam instruction information, or a training mechanism of the AI ​​model; 35. The method of any one of claims 28 to 34, further comprising:

36. 1. A communication method comprising: transmitting first reference signal resource configuration information to a terminal device, the first reference signal resource configuration information including first instruction information, wherein the first instruction information indicates that a first reference signal resource configured based on the first reference signal resource configuration information is for performing model training on an AI model, or the first instruction information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for acquiring a dataset for performing model training on an AI model; A method comprising:

37. 37. The method of claim 36, wherein the first instruction information further indicates that the first reference signal resource is for obtaining labels to perform model training on the AI ​​model.

38. 37. The method of claim 36, wherein the first reference signal resource configuration information further includes fifth instruction information, and the fifth instruction information indicates that the first reference signal resource is for obtaining labels for performing model training on the AI ​​model.

39. transmitting third reference signal resource configuration information to the terminal device, the third reference signal resource configuration information including sixth instruction information, the sixth instruction information indicating that a third reference signal resource configured based on the third reference signal resource configuration information is for obtaining input information for performing model training on the AI ​​model; 39. The method of any one of claims 36 to 38, further comprising:

40. The first reference signal resource configuration information further includes seventh indication information indicating an association relationship between the first reference signal resource and the third reference signal resource; or 40. The method of claim 39, wherein the third reference signal resource configuration information further includes eighth indication information indicating an association relationship between the first reference signal resource and the third reference signal resource.

41. receiving first information from the terminal device, the first information being for requesting to perform model training based on the AI ​​model, the first information including one or more of an identifier of the AI ​​model, computing capability information of the terminal device, sparse beam instruction information, or a training mechanism of the AI ​​model; 41. The method of any one of claims 36 to 40, further comprising:

42. 1. A communication method comprising: receiving first reference signal resource configuration information from a network device, the first reference signal resource configuration information including first instruction information, wherein the first instruction information indicates that a first reference signal resource configured based on the first reference signal resource configuration information is for performing model inference on an AI model, or the first instruction information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for acquiring a dataset for performing model inference on an AI model; A method comprising:

43. performing model inference based on the AI ​​model and the first reference signal resource configuration information; 43. The method of claim 42, further comprising:

44. receiving second reference signal resource configuration information from the network device, the second reference signal resource configuration information including second instruction information, wherein the second instruction information indicates that a second reference signal resource configured based on the second reference signal resource configuration information is for performing model training for the AI ​​model, or the second instruction information indicates that a second reference signal resource configured based on the second reference signal resource configuration information is for acquiring a dataset for performing model training for the AI ​​model; 44. The method of claim 42 or 43, further comprising:

45. The first reference signal resource configuration information further includes third indication information indicating an association relationship between the first reference signal resource and the second reference signal resource; or 45. The method of claim 42, wherein the second reference signal resource configuration information further includes fourth indication information indicating an association relationship between the second reference signal resource and the first reference signal resource.

46. 46. ​​The method of claim 42, wherein the first reference signal resource is a reference signal resource corresponding to input information for model inference of the AI ​​model, and the first reference signal resource configuration information further includes first identifier information, the first identifier information corresponds to a target reference signal resource set, and the target reference signal resource set includes reference signal resources corresponding to labels in a model training process of the AI ​​model.

47. receiving fifth reference signal resource configuration information from the network device, the fifth reference signal resource configuration information for configuring the target reference signal resource set, the fifth reference signal resource configuration information including the first identifier information, the fifth reference signal resource configuration information further including second identifier information, the second identifier information corresponding to the target reference signal resource set; 47. The method of claim 46, further comprising:

48. sending first information to the network device, the first information being for requesting that model inference be performed on the AI ​​model, the first information including one or more of an identifier of the AI ​​model, instruction information #1, or sparse beam instruction information, the instruction information #1 indicating that training of the AI ​​model has been completed; 48. The method of any one of claims 42 to 47, further comprising:

49. 1. A communication method comprising: transmitting first reference signal resource configuration information to a terminal device, the first reference signal resource configuration information including first instruction information, wherein the first instruction information indicates that a first reference signal resource configured based on the first reference signal resource configuration information is for performing model inference on an AI model, or the first instruction information indicates that the first reference signal resource configured based on the first reference signal resource configuration information is for acquiring a dataset for performing model inference on an AI model; A method comprising:

50. transmitting second reference signal resource configuration information to the terminal device, the second reference signal resource configuration information including second instruction information, wherein the second instruction information indicates that the second reference signal resource configured based on the second reference signal resource configuration information is for performing model training on the AI ​​model, or the second instruction information indicates that the second reference signal resource configured based on the second reference signal resource configuration information is for acquiring a dataset for performing model training on the AI ​​model; 50. The method of claim 49, further comprising:

51. The first reference signal resource configuration information further includes third indication information indicating an association relationship between the first reference signal resource and the second reference signal resource; or The method of claim 49 or 50, wherein the second reference signal resource configuration information further includes fourth indication information indicating an association relationship between the second reference signal resource and the first reference signal resource.

52. 52. The method of claim 49, wherein the first reference signal resource is a reference signal resource corresponding to input information for model inference of the AI ​​model, and the first reference signal resource configuration information further includes first identifier information, the first identifier information corresponds to a target reference signal resource set, and the target reference signal resource set includes reference signal resources corresponding to labels in a model training process of the AI ​​model.

53. transmitting fifth reference signal resource configuration information to the terminal device, the fifth reference signal resource configuration information being for configuring the target reference signal resource set, the fifth reference signal resource configuration information including the first identifier information, the fifth reference signal resource configuration information further including second identifier information, and the second identifier information corresponding to the target reference signal resource set; 53. The method of claim 52, further comprising:

54. receiving first information from the terminal device, the first information being for requesting that model inference be performed on the AI ​​model, the first information including one or more of an identifier of the AI ​​model, instruction information #1, or sparse beam instruction information, the instruction information #1 indicating that training of the AI ​​model has been completed; 54. The method of any one of claims 49 to 53, further comprising:

55. A communication device comprising a module or unit configured to carry out the method of any one of claims 1 to 54.

56. 55. A communications device comprising at least one processor coupled to a memory, the memory configured to store a program, and the at least one processor configured to execute the computer program or instructions stored in the memory to perform the method of any one of claims 1 to 54.

57. 55. A computer-readable storage medium storing program code for execution by a device, the program code being used to perform the method of any one of claims 1 to 54.

58. 55. A computer program product comprising instructions that, when run on a computer, cause the computer to perform the method of any one of claims 1 to 54.

59. 55. A chip comprising a processor and a communication interface, wherein the processor reads instructions stored in a memory via the communication interface to perform the method of any one of claims 1 to 54.

60. A communication system comprising a terminal device and / or a network device, wherein the terminal device is configured to perform the method of any one of claims 1 to 13, or to perform the method of any one of claims 28 to 35, or to perform the method of any one of claims 42 to 48, and the network device is configured to perform the method of any one of claims 14 to 27, or to perform the method of any one of claims 36 to 41, or to perform the method of any one of claims 49 to 54.

Images