Communication method and related apparatus

Abstract

A communication method and a related apparatus. In the method, M pieces of output data acquired by a first communication apparatus are obtained by processing first input data by means of N models. Then, the first communication apparatus can determine, on the basis of the M pieces of output data, label data corresponding to the first input data. The label data is used for determining a first model, and the first model is used for determining, from one or more pieces of output data corresponding to second input data, a target output corresponding to the second input data. In other words, the first model is used for determining a personalized target output, that is, the first model may be a personalized model, and the label data may be used for determining the personalized model. Therefore, the label data determined by the first communication apparatus by means of the M pieces of output data can be used for determining the personalized model, and the personalized model can provide a personalized target output, so as to improve user experience.

Description

Communication method and related apparatus

[0001] This application claims priority from the Chinese patent application No. 202411076579.X filed on August 6, 2024, and entitled "A communication method and related apparatus", the content of which is incorporated herein by reference in its entirety. TECHNICAL FIELD

[0002] The present application relates to the field of communication, and in particular, to a communication method and related apparatus. BACKGROUND

[0003] With the development of communication technology, in a communication system, in addition to traditional communication services, the services performed by a communication device can also include other new services, such as artificial intelligence (AI) services. Generally, a communication system capable of processing AI services can also be referred to as an AI system.

[0004] Currently, a communication device can serve as a participating node of an AI system, and the computing power of the communication device is applied to a certain link of the AI system. Generally, the AI function introduced in the communication network needs to rely on a model to be implemented. However, in the AI system, how to improve user experience is a technical problem to be solved. SUMMARY

[0005] The present application provides a communication method and related apparatus for improving user experience.

[0006] The first aspect of the present application provides a communication method, which is performed by a first communication apparatus. The first communication apparatus can be a communication device (such as a terminal device or a network device), or the first communication apparatus can be a part of the communication device (such as a circuit or a chip responsible for communication functions (such as a Modem chip, also known as a baseband chip, or a system on chip (SoC) chip or a system in package (SIP) chip containing a modem core), etc.), or the first communication apparatus can also be a logic module or software that can realize all or part of the functions of the communication device. In the method, the first communication apparatus obtains M output data, the M output data is obtained by processing a first input data through N models, N is a positive integer, and M is an integer greater than or equal to N; the first communication apparatus determines label data corresponding to the first input data based on the M output data, wherein the label data is used to determine a first model, and the first model is used to determine a target output corresponding to a second input data from one or more output data corresponding to the second input data.

[0007] Based on the above scheme, the M output data obtained by the first communication device is obtained by processing the first input data through the N models. Thereafter, the first communication device can determine the label data corresponding to the first input data based on the M output data. The label data is used to determine the first model, and the first model is used to determine the target output corresponding to the second input data from one or more output data corresponding to the second input data. In other words, the first model is used to determine the personalized target output, i.e., the first model can be a personalized model, and the above label data can be used to determine the personalized model. Thus, the label data determined by the first communication device through the M output data can be used to determine the personalized model, and the personalized target output can be provided through the personalized model to improve the user experience.

[0008] In addition, in the case where the M output data is a plurality of output data (for example, M is greater than 1), the above label data can be obtained by processing the corresponding diversified output data through the N models, which can improve the performance of the label data and further improve the performance of the first model based on the label data.

[0009] It should be understood that the label data is used to determine the first model, which can be understood as the label data being used to determine the first model through model processing. For example, the model processing can include one or more of model training, model generation, model updating, model fine-tuning, model adjustment, and model fine adjustment.

[0010] Optionally, taking the label data being used to determine the first model through model training as an example, since the label data is the label data corresponding to the first input data, the model training data of the first model can include the first input data in addition to the above label data during the model training process. In this way, the first model can be trained with more information to improve the performance of the first model.

[0011] In this application, the model can be replaced by other terms, such as AI model, neural network, neural network model, AI neural network model, machine learning model, or AI processing model, etc.

[0012] In this application, the target output can be replaced by other terms, such as target, desired output, ideal output, personalized output, or label, etc.

[0013] It should be understood that the M output data is obtained by processing the first input data through the N models, which can be understood as each model in the N models processing the first input data to obtain one or more output data, and the M output data can include one or more output data corresponding to each model in the N models.

[0014] It should be understood that any of the M output data can be a model output or an intermediate result of one of the N models. In other words, one or more output data corresponding to each of the N models can include a model output and / or an intermediate result of the each of the N models, etc.

[0015] For example, one model can include one or more neural networks, and a model output of the one model can be understood as a partial or full output of a last neural network of the one or more neural networks, e.g., the model output can include the partial or full output of the last neural network. In addition, an intermediate result of the one model can be understood as a partial or full output of at least one neural network other than the last neural network of the one or more neural networks, e.g., the intermediate result can include the partial or full output of each of the at least one neural network.

[0016] As an example, in the above scenario, any of the N models is configured to process the first input data to obtain one or more output data, i.e., the model processing performed by the any of the N models can include one or more of reasoning, prediction, derivation, identification, decision, etc. Accordingly, the any of the N models can be referred to as a reasoning model, a prediction model, a pre-trained model for the above model processing, or a pre-trained large model for the above model processing, etc.

[0017] As an example, in the above scenario, the first model is configured to determine the target output corresponding to the second input data from the one or more output data corresponding to the second input data; wherein the model processing performed by the first model can include evaluation, ranking, optimization, screening, selection, or filtering, etc. Accordingly, the first model can be referred to as an evaluation model, a personalization model, a personalized evaluation model, a reward model, or a scoring model, etc.

[0018] Optionally, the first model can include one or more sub-models, each of which can process the one or more output data corresponding to the second input data to obtain a corresponding output, and then the one or more outputs corresponding to the one or more sub-models can be used to determine the target output corresponding to the second input data. In other words, the first model can determine the target output corresponding to the second input data through the one or more sub-models.

[0019] In a possible implementation form of the first aspect, the method further includes: determining, by the first communication device, the first model based on the tag data.

[0020] Based on the above scheme, after the first communication device determines the label data based on the M output data, the first communication device can further determine the first model based on the label data, so that the first communication device can locally (or through an external device of the first communication device, such as a terminal device with AI function, a server, etc.) implement the determination of the first model, thereby reducing the transmission overhead.

[0021] In a possible implementation form of the first aspect, the method further includes: the first communication device sending first information, the first information being used to indicate the label data.

[0022] Based on the above scheme, the first communication device can send the first information, so that a receiver (for example, the second communication device) of the first information can obtain the label data and implement the determination of the first model based on the label data, thereby reducing the implementation complexity of the first communication device.

[0023] In a possible implementation form of the first aspect, the method further includes: the first communication device receiving or sending second information, the second information being used to indicate model information of the first model.

[0024] Based on the above scheme, after the communication device (for example, the first communication device or the second communication device) determines the first model based on the label data, the communication device can further indicate the model information of the first model through the second information, so that a receiver of the second information can obtain the model information of the first model through the second information, thereby facilitating the sender or the receiver of the second information to perform model management on the first model based on the model information.

[0025] Optionally, the model information can indicate one or more of a model identifier (or index), a model parameter, a model structure, a sampling parameter, and a sampling number. For example, the model parameter can include one or more of a model hyperparameter and a model capability level. For example, in the case that the first model is an evaluation model, the second information indicating the model information of the first model can be understood as the second information indicating the model information of the evaluation model, including but not limited to one or more of an evaluation model identifier (or index), an evaluation model parameter, an evaluation model structure, a sampling parameter, and a sampling number.

[0026] Optionally, the model management can include one or more of model scheduling, model updating, model switching, and function fallback.

[0027] For example, the sampling parameter can include at least one of the following:

[0028] Temperature coefficient: a coefficient for controlling probability normalization, for example, a higher temperature makes the output more random, and a lower temperature makes the output more deterministic;

[0029] Random coefficient: randomly sampling with a certain probability;

[0030] Top-K: sampling from the K outputs with the highest probabilities;

[0031] Top-P: sampling from the output set with the cumulative probability greater than or equal to p.

[0032] In a possible implementation of the first aspect, the first communication apparatus obtains the M output data by receiving the M output data.

[0033] According to the above scheme, the first communication apparatus can obtain the M output data by receiving the M output data, that is, the N models used for processing the first input data can be deployed on other communication apparatuses, so that the first communication apparatus does not need to deploy the N models (for example, N pre-trained large models), thereby saving the storage space of the first communication apparatus and reducing the processing complexity of the first communication apparatus.

[0034] In a possible implementation of the first aspect, the method further includes: the first communication apparatus sending the first input data.

[0035] According to the above scheme, the first communication apparatus can further send the first input data to one or more communication apparatuses that deploy the N models, so that the one or more communication apparatuses can process the first input data to obtain and send the M output data.

[0036] Optionally, the first input data can be preconfigured data, so that the first communication apparatus does not need to send the first input data, thereby reducing the overhead.

[0037] In a possible implementation of the first aspect, the method further includes: the first communication apparatus sending third information, the third information indicating a request for the M output data.

[0038] According to the above scheme, the first communication apparatus can further send the third information, so that the receiver (for example, the second communication apparatus) of the third information can provide the M output data for the first communication apparatus based on the request of the third information.

[0039] Optionally, the third information indicates at least one of the following: a task identifier, model information of part or all of the N models, a processing mode, a generation mode of the first input data, or task auxiliary information.

[0040] As an example, the third information can be used to request M output data, and the determined label data of the M output data can be used to determine a first model, where the processing of the first model corresponds to a different processing mode, and different output data (e.g., the first indication information, the second indication information, the third indication information, etc. contained in the sixth information described below) can be obtained. Accordingly, in the case where the third information indicates the processing mode, the receiver of the third information can provide the corresponding M output data based on the processing mode, so that the first communication device obtains the M output data corresponding to the specified processing mode and obtains the first model corresponding to the specified processing mode. Optionally, in the case where the above processing is inference, the processing mode can be replaced by an inference mode.

[0041] As an example, the generation manner of the first input data can indicate that the first input data is generated in an offline manner or an online manner.

[0042] As an example, the task auxiliary information can contain information required for model management. For example, in the case where the model is used for a communication task, the task auxiliary information can include the configuration of the communication parameters (e.g., resource parameter configuration, power control parameter configuration, etc.) of the communication device. For another example, in the case where the model is used for a classification task, the task auxiliary information can include the approximate category of the input data. For another example, in the case where the model is used for a generation task, the task auxiliary information can include a prompt word.

[0043] In a possible implementation of the first aspect, the method further includes: receiving, by the first communication device, fourth information, where the fourth information is used to indicate the configuration information corresponding to the request.

[0044] Based on the above scheme, the receiver (e.g., the second communication device) of the third information can determine the configuration information based on the third information after receiving the third information, and indicate the configuration information to the first communication device through the fourth information, so that the first communication device can determine the configuration information corresponding to the processing performed by the second communication device on the request based on the configuration information.

[0045] Optionally, the configuration information indicated by the fourth information can be the same as the content (e.g., the model information of part or all of the N models, the processing mode, or the task auxiliary information, etc. in the above) requested by the third information, or can be partially different or completely different from the content requested by the third information, which is not limited here.

[0046] Optionally, in a case where the configuration information indicated by the fourth information is different from the content part requested by the third information or completely different, the first communication apparatus can indicate rejection (e.g., rejection of processing of the first input data by the second communication apparatus based on the configuration information) to the second communication apparatus, so that the second communication apparatus does not need to process based on the configuration information that is not expected by the first communication apparatus, thereby reducing processing and transmission overhead. And / or, the first communication apparatus can determine / generate / acquire the next request (e.g., the third information sent next) based on the configuration information, thereby improving processing efficiency.

[0047] In a possible implementation of the first aspect, the configuration information indicates at least one of: a groupcast identifier corresponding to the first communication apparatus, part or all of the N models, or a processing mode; wherein the M output data are carried in groupcast information, and the groupcast information comprises the groupcast identifier.

[0048] Based on the above scheme, the configuration information indicated by the fourth information can comprise the at least one described above, thereby improving the flexibility of implementation of the scheme.

[0049] For example, in a case where the configuration information indicates the groupcast identifier corresponding to the first communication apparatus, the first communication apparatus can obtain the M output data through the received groupcast information, and in a case where the number of first communication apparatuses is greater than 1, the groupcast transmission mode can reduce the transmission overhead of the output data.

[0050] For another example, in a case where the configuration information indicates part or all of the N models, the first communication apparatus can provide first input data matched with the input of the part or all of the models through the indication.

[0051] For another example, in a case where the configuration information indicates a processing mode for processing of the first input data, the first communication apparatus can obtain a target output matched with the processing mode based on the indication of the configuration information.

[0052] In a possible implementation of the first aspect, the first communication apparatus obtains the M output data by processing the first input data based on the N models.

[0053] Based on the above scheme, the first communication apparatus can process the first input data based on the N models deployed locally (or through an external device of the first communication apparatus, such as a terminal device, a server, etc. with AI function) to obtain the M output data, thereby reducing transmission overhead.

[0054] In a possible implementation of the first aspect, the method further comprises: receiving, by the first communication apparatus, the first input data.

[0055] Based on the above scheme, the first communication device can receive first input data, so that the first communication device can obtain M output data by processing the first input data.

[0056] Optionally, the first input data can be preconfigured data, or the first data can be data collected by the first communication device itself, so as to reduce transmission overhead.

[0057] In a possible implementation of the first aspect, the first communication device determines label data corresponding to the first input data based on the M output data, including: the first communication device determines fifth information based on the M output data, the fifth information being used to indicate at least one of a score, an order, an optimal value (or a better value) of the M output data; wherein the fifth information is used to determine the label data corresponding to the first input data.

[0058] Based on the above scheme, the first communication device can determine the label data through the fifth information, which can be implemented in the above-mentioned multiple ways, so as to improve the flexibility of the scheme implementation.

[0059] In a possible implementation of the first aspect, the first model is used to determine a target output corresponding to second input data from one or more output data corresponding to the second input data, including:

[0060] The first model is used to determine sixth information from one or more output data corresponding to the second input data, the sixth information being used to determine a target output corresponding to the second input data; wherein the sixth information includes at least one of:

[0061] First indication information indicating an order of the one or more output data; wherein the target output is one of the one or more output data;

[0062] Second indication information indicating a first index, the first index being used to determine the target output in the one or more output data; wherein the target output is one of the one or more output data;

[0063] Third indication information indicating the target output; wherein the target output corresponding to the first input data is determined based on the one or more output data. For example, the target output is one of the one or more output data, or the target output is different from any of the one or more output data.

[0064] Based on the above scheme, in the process of model processing of the first model, the first model can determine the target output corresponding to the second input data in the above-mentioned multiple ways, so as to improve the flexibility of the scheme implementation.

[0065] The second aspect of the present application provides a communication method, which is performed by a second communication device. The second communication device can be a communication device (e.g., a terminal device or a network device), or the second communication device can be a part of the communication device (e.g., a circuit or a chip responsible for communication functions (e.g., a modem chip, also known as a baseband chip, or a SoC chip or a SIP chip containing a modem core, etc.), or the second communication device can also be a logic module or software capable of implementing all or part of the functions of the communication device. In the method, the second communication device receives first information, which is used to indicate label data corresponding to first input data; wherein the label data is determined based on M output data obtained by processing the first input data through N models, N is a positive integer, and M is an integer greater than or equal to N; the second communication device determines a first model based on the label data; wherein the first model is used to determine output data corresponding to second input data from one or more output data corresponding to the second input data.

[0066] Based on the above scheme, the first information received by the second communication device is used to indicate the label data corresponding to the first input data. The second communication device can use the label data to determine the first model, and the first model is used to determine the target output corresponding to the second input data from one or more output data corresponding to the second input data. In other words, the first model is used to determine the personalized target output, that is, the first model can be a personalized model, and the above label data can be used to determine the personalized model. Thus, the second communication device can determine the personalized model through the label data, and the personalized target output can be provided through the personalized model to improve the user experience.

[0067] In addition, in the case where the M output data is a plurality of output data (e.g., M is greater than 1), the above label data can be obtained by processing the corresponding diversified output data through N models, which can improve the performance of the label data and further improve the performance of the first model based on the label data.

[0068] In a possible implementation manner of the second aspect, the method further includes: the second communication device sends second information, which is used to indicate model information of the first model.

[0069] Based on the above scheme, after the second communication device determines the first model based on the label data, the second communication device can further indicate the model information of the first model through the second information, so that the receiver (e.g., the first communication device) of the second information can obtain the model information of the first model through the second information, so that the sender of the second information or the receiver of the second information can perform model management on the first model based on the model information.

[0070] In a possible implementation manner of the second aspect, the method further includes: the second communication apparatus sending the M output data.

[0071] According to the foregoing scheme, the second communication apparatus can send the M output data, that is, the N models used for processing the first input data can be deployed on the second communication apparatus or other communication apparatus connected to the second communication apparatus, so that the first communication apparatus does not need to deploy the N models (for example, N pre-trained large models), thereby saving the storage space of the first communication apparatus and reducing the processing complexity of the first communication apparatus.

[0072] In a possible implementation manner of the second aspect, the method further includes: the second communication apparatus receiving the first input data.

[0073] According to the foregoing scheme, the second communication apparatus can further receive the first input data, so that the second communication apparatus can process the first input data to obtain and send the M output data.

[0074] Optionally, the first input data can be preconfigured data, so that the second communication apparatus does not need to receive the first input data, thereby reducing the overhead.

[0075] In a possible implementation manner of the second aspect, the method further includes: the second communication apparatus receiving third information, the third information indicating a request for the M output data.

[0076] According to the foregoing scheme, the second communication apparatus can further receive the third information, so that the second communication apparatus can provide the M output data for the first communication apparatus based on the request of the third information.

[0077] Optionally, the third information indicates at least one of the following: a task identifier, model information of part or all of the N models, a processing mode for processing the first input data, a generation mode of the first input data, or task auxiliary information.

[0078] In a possible implementation manner of the second aspect, the method further includes: the second communication apparatus sending fourth information, the fourth information being used for indicating configuration information corresponding to the request.

[0079] According to the foregoing scheme, after receiving the third information, the second communication apparatus can determine the configuration information based on the third information, and indicate the configuration information to the first communication apparatus through the fourth information, so that the first communication apparatus can determine the configuration information corresponding to the processing performed by the second communication apparatus on the request based on the configuration information.

[0080] In a possible implementation of the second aspect, the configuration information indicates at least one of: a multicast identifier corresponding to the first communication device, part or all of the N models, or a processing mode; wherein the M output data is carried in multicast information, and the multicast information comprises the multicast identifier.

[0081] According to the above scheme, the configuration information indicated by the fourth information can comprise the at least one, so as to improve the flexibility of the scheme implementation.

[0082] For example, in the case where the configuration information indicates the multicast identifier corresponding to the first communication device, the first communication device can obtain the M output data through the received multicast information, and in the case where the number of the first communication devices is greater than 1, the multicast transmission mode can reduce the transmission overhead of the output data.

[0083] For another example, in the case where the configuration information indicates part or all of the N models, the first communication device can provide the first input data matched with the input of the part or all of the models through the indication.

[0084] For another example, in the case where the configuration information indicates the processing mode for processing the first input data, the first communication device can obtain the target output matched with the processing mode based on the indication of the configuration information.

[0085] In a possible implementation of the second aspect, the method further comprises: sending, by the second communication device, the first input data.

[0086] According to the above scheme, the second communication device can send the first input data to the first communication device, so that the first communication device can obtain the M output data through the processing of the first input data.

[0087] Optionally, the first input data can be preconfigured data, or the first data can be data collected by the first communication device itself, so as to reduce the transmission overhead.

[0088] In a possible implementation of the second aspect, the label data is determined based on the M output data obtained through the processing of the first input data by the N models, comprising: the label data is determined through fifth information, the fifth information is determined through the M data; wherein the fifth information is used to indicate at least one of a score, an order, an optimal value (or a better value) of the M output data.

[0089] According to the above scheme, the first communication device can determine the label data through the fifth information, and the fifth information can be implemented in the above-mentioned various ways, so as to improve the flexibility of the scheme implementation.

[0090] In a possible implementation of the second aspect, the first model is configured to determine the target output corresponding to the second input data from one or more output data corresponding to the second input data, including:

[0091] The first model is configured to determine sixth information from the one or more output data corresponding to the second input data, the sixth information being used to determine the target output corresponding to the second input data; and the sixth information includes at least one of the following:

[0092] First indication information indicating an order of the one or more output data; and the target output is one of the one or more output data.

[0093] Second indication information indicating a first index used to determine the target output from the one or more output data; and the target output is one of the one or more output data.

[0094] Third indication information indicating the target output; and the target output corresponding to the first input data is determined based on the one or more output data. For example, the target output is one of the one or more output data, or the target output is different from any of the one or more output data.

[0095] Based on the above scheme, in the process of model processing of the first model, the first model can determine the target output corresponding to the second input data in the above-mentioned multiple ways, so as to improve the flexibility of the scheme implementation.

[0096] The third aspect of the present application provides a communication method, which is performed by a first communication device. The first communication device can be a communication device (such as a terminal device or a network device), or the first communication device can be part of a communication device (for example, a circuit or a chip responsible for communication functions (such as a Modem chip, also known as a baseband chip, or a SoC chip or a SIP chip containing a modem core, etc.), or the first communication device can also be a logic module or software that can realize all or part of the functions of the communication device. In the method, the first communication device obtains M output data, the M output data being obtained by processing a first input data by N models, N being a positive integer, and M being an integer greater than or equal to N; the first communication device determines seventh information based on the M output data, the seventh information being used to determine a sampling parameter of a second input data and / or one or more models for processing the second input data, the second input data being used to be processed by the one or more models to obtain a target output corresponding to the second input data.

[0097] Based on the above scheme, the M output data obtained by the first communication apparatus is obtained by processing the first input data by the N models. Thereafter, the first communication apparatus can determine the seventh information based on the M output data. The seventh information is used to determine the second input data and / or the sampling parameters of one or more models for processing the second input data, so as to obtain the target output corresponding to the second input data. In other words, the second input data and / or the sampling parameters determined by the seventh information can be used to determine the personalized target output corresponding to the second input data. Thus, the seventh information determined by the first communication apparatus based on the M output data can be used to optimize the sampling parameters and / or the input data of the model, so that the model can obtain the personalized target output, thereby improving the user experience.

[0098] In addition, in the case where the M output data is a plurality of output data (for example, M is greater than 1), the above-mentioned seventh information can be obtained by processing the corresponding diversified output data by the N models, which can improve the performance of the seventh information and further improve the performance of the sampling parameters and / or the input data of the model based on the seventh information.

[0099] For example, the sampling parameters can include at least one of the following:

[0100] Temperature coefficient: a coefficient for controlling probability normalization, for example, a higher temperature makes the output more random, and a lower temperature makes the output more deterministic;

[0101] Random coefficient: random sampling with a certain probability;

[0102] Top-K: sampling from the K outputs with the highest probability;

[0103] Top-P: sampling from the output set with a cumulative probability greater than or equal to p.

[0104] In a possible implementation of the third aspect, the method further includes: determining, by the first communication apparatus, the second input data and / or the sampling parameters of one or more models for processing the second input data based on the seventh information.

[0105] Based on the above scheme, after the first communication apparatus determines the seventh information based on the M output data, the first communication apparatus can further determine the second input data and / or the sampling parameters of one or more models for processing the second input data based on the seventh information, so that the first communication apparatus can locally (or through an external device of the first communication apparatus, such as a terminal device with AI function, a server, etc.) determine the second input data and / or the sampling parameters, thereby reducing the transmission overhead.

[0106] In a possible implementation form of the third aspect, the method further includes: sending, by the first communication apparatus, the seventh information.

[0107] Based on the above scheme, the first communication apparatus can send the seventh information, so that the receiver (for example, the second communication apparatus) of the seventh information can obtain the seventh information and implement the determination of the second input data and / or the sampling parameter based on the seventh information, so as to reduce the implementation complexity of the first communication apparatus.

[0108] In a possible implementation form of the third aspect, the method further includes: receiving or sending, by the first communication apparatus, eighth information, the eighth information being used to indicate evaluation formula information, the evaluation formula information and the M output data being used to determine the seventh information.

[0109] Based on the above scheme, the first communication apparatus can further receive or send the eighth information used to indicate the evaluation formula information, so that the receiver of the eighth information can implement the determination of the seventh information.

[0110] It should be understood that the evaluation formula indicated by the above-mentioned evaluation formula information can be used to score the inference result, for example, the evaluation formula represents the matching degree of the inference result and the local individualized preference / indicator. For example, the evaluation formula can be implemented by a series of physical / mathematical meaning calculation processes.

[0111] Optionally, the above-mentioned evaluation formula information can be replaced by other implementations, for example, evaluation algorithm, evaluation model, evaluation module, or evaluation key performance indicator (KPI), etc.

[0112] For example, the seventh information is used to determine the second input data and / or the sampling parameter of one or more models for processing the second input data, which can be understood as that the seventh information can determine the second input data and / or the sampling parameter of one or more models for processing the second input data by at least one of the evaluation formula, the evaluation algorithm, the evaluation model, the evaluation module, or the evaluation KPI.

[0113] In a possible implementation form of the third aspect, the first communication apparatus obtains the M output data, including: receiving, by the first communication apparatus, the M output data.

[0114] Based on the above scheme, the first communication apparatus can obtain the M output data by receiving the M output data, that is, the N models used to process the first input data can be deployed on other communication apparatuses, so that the first communication apparatus does not need to deploy the N models (for example, N pre-trained large models), which can save the storage space of the first communication apparatus and reduce the processing complexity of the first communication apparatus.

[0115] In a possible implementation form of the third aspect, the method further includes: sending, by the first communication apparatus, the first input data.

[0116] Based on the above scheme, the first communication apparatus can further send the first input data to one or more communication apparatuses which deploy the N models, so that the one or more communication apparatuses can process the first input data to obtain and send the M output data.

[0117] Optionally, the first input data can be preconfigured data, so that the first communication apparatus does not need to send the first input data, and the overhead can be reduced.

[0118] In a possible implementation form of the third aspect, the method further includes: sending, by the first communication apparatus, ninth information, the ninth information indicating a request for the M output data.

[0119] Based on the above scheme, the first communication apparatus can further send the ninth information, so that a receiver (e.g., the second communication apparatus) of the ninth information can provide the M output data for the first communication apparatus based on the request of the ninth information.

[0120] Optionally, the ninth information indicates at least one of the following: a task identifier, model information of part or all of the N models, a processing mode, a generation manner of the first input data, or task auxiliary information.

[0121] As an example, the ninth information can be used to request the M output data, the M output data can be used to determine the seventh information, and the first communication apparatus (or the receiver of the seventh information) can process the seventh information to obtain the second input data and / or sampling parameters of one or more models processing the second input data. Wherein, the processing of the seventh information corresponds to different processing modes, and different processing results can be obtained. Accordingly, in the case where the ninth information indicates the processing mode, the receiver of the ninth information can provide the corresponding M output data based on the processing mode, so that the first communication apparatus obtains the M output data corresponding to the specified processing mode and obtains the seventh information of the specified processing mode.

[0122] Optionally, in the case where the processing of the seventh information is the processing of the evaluation formula, different processing modes can be understood as different processing of the evaluation formula.

[0123] Optionally, in the case where the processing of the seventh information is the processing of the evaluation model, different processing modes can be understood as different processing of the evaluation model.

[0124] As an example, the generation manner of the first input data can indicate that the first input data is generated by offline or online.

[0125] As an example, the task auxiliary information can contain information required for model management. For example, in the case where the model is used for a communication task, the task auxiliary information can include configuration of communication parameters (e.g., resource parameter configuration, power control parameter configuration, etc.) of the communication device. As another example, in the case where the model is used for a classification task, the task auxiliary information can include the approximate category of the input data. As another example, in the case where the model is used for a generation task, the task auxiliary information can include a prompt word.

[0126] In a possible implementation of the third aspect, the method further includes: receiving, by the first communication device, tenth information, the tenth information being used to indicate the configuration information corresponding to the request.

[0127] Based on the above scheme, the receiver (e.g., the second communication device) of the ninth information can determine the configuration information based on the ninth information after receiving the ninth information, and indicate the configuration information to the first communication device through the tenth information, so that the first communication device can determine the configuration information corresponding to the processing performed by the second communication device on the request based on the configuration information.

[0128] Optionally, the configuration information indicated by the tenth information can be the same as the content (e.g., model information of part or all of the N models, or task auxiliary information, etc.) requested by the ninth information, or can be partially different or completely different from the content requested by the ninth information, which is not limited here.

[0129] Optionally, in the case where the configuration information indicated by the tenth information is partially different or completely different from the content requested by the ninth information, the first communication device can indicate rejection (e.g., rejection of the second communication device to process the first input data based on the configuration information) to the second communication device, so that the second communication device does not need to process based on the configuration information not expected by the first communication device, to reduce processing and transmission overhead. And / or, the first communication device can determine / generate / acquire the next request (e.g., the ninth information sent next) based on the configuration information, to improve processing efficiency.

[0130] In a possible implementation of the third aspect, the configuration information indicates at least one of: a multicast identifier corresponding to the first communication device, part or all of the N models, or a processing mode; wherein the M output data is carried in multicast information, and the multicast information includes the multicast identifier.

[0131] Based on the above scheme, the configuration information indicated by the tenth information can include the at least one described above, to improve the flexibility of the scheme implementation.

[0132] For example, in a case where the configuration information indicates a multicast identifier corresponding to the first communication device, the first communication device can obtain the M output data from the received multicast information. In a case where the number of the first communication devices is greater than 1, the transmission overhead of the output data can be reduced by using the multicast transmission mode.

[0133] For another example, in a case where the configuration information indicates part or all of the N models, the first communication device can provide the first input data matched with the input of the part or all of the N models according to the indication of the configuration information.

[0134] For another example, in a case where the configuration information indicates a processing mode for processing the first input data, the first communication device can obtain the seventh information matched with the processing mode according to the indication of the configuration information.

[0135] In a possible implementation of the third aspect, the first communication device obtains the M output data by processing the first input data based on the N models.

[0136] According to the above scheme, the first communication device can process the first input data based on the N models deployed locally (or through an external device of the first communication device, such as a terminal device with AI function, a server, etc.) to obtain the M output data, so as to reduce the transmission overhead.

[0137] In a possible implementation of the third aspect, the method further includes that the first communication device receives the first input data.

[0138] According to the above scheme, the first communication device can receive the first input data, so that the first communication device can obtain the M output data by processing the first input data.

[0139] Optionally, the first input data can be preconfigured data, or the first data can be data collected by the first communication device itself, so as to reduce the transmission overhead.

[0140] The fourth aspect of the present application provides a communication method, which is performed by a second communication device. The second communication device can be a communication device (e.g., a terminal device or a network device), or the second communication device can be a part of the communication device (e.g., a circuit or a chip responsible for communication functions (e.g., a Modem chip, also known as a baseband chip, or a SoC chip or a SIP chip containing a modem core, etc.), or the second communication device can also be a logic module or software capable of implementing all or part of the functions of the communication device. In the method, the second communication device receives seventh information, which is obtained based on M output data obtained by processing N input data by N models, where N is a positive integer, and M is an integer greater than or equal to N; and the second communication device determines second input data and / or sampling parameters of one or more models for processing the second input data based on the seventh information, where the second input data is used for processing by the one or more models to obtain a target output corresponding to the second input data.

[0141] Based on the above scheme, after receiving the seventh information, the second communication device can determine the second input data and / or the sampling parameters of the one or more models for processing the second input data based on the seventh information, where the second input data is used for processing by the one or more models to obtain a target output corresponding to the second input data. In other words, the second input data and / or the sampling parameters determined based on the seventh information can be used to determine the personalized target output corresponding to the second input data. Thus, the second communication device can optimize the sampling parameters and / or the input data of the models based on the seventh information determined based on the M output data, so that the models can obtain personalized target outputs, thereby improving the user experience.

[0142] In addition, in the case where the M output data is a plurality of output data (e.g., M is greater than 1), the above-mentioned seventh information can be obtained by processing the corresponding diversified output data by the N models, which can improve the performance of the seventh information and further improve the performance of the sampling parameters and / or the input data of the models based on the seventh information.

[0143] It should be understood that the above-mentioned sampling parameters can include one or more of the sampling parameters of the input data, the sampling number of the input data, the sampling parameters of the output data, the sampling number of the output data, and the model capability level.

[0144] In a possible implementation manner of the fourth aspect, the method further includes: the second communication device receives or sends eighth information, where the eighth information is used to indicate evaluation formula information, and the evaluation formula information and the M output data are used to determine the seventh information.

[0145] Based on the above scheme, the second communication apparatus can further receive or send eighth information for indicating the evaluation formula information, so that the receiver of the eighth information can determine the seventh information.

[0146] In a possible implementation of the fourth aspect, the method further includes: the second communication apparatus sending the M output data.

[0147] Based on the above scheme, the second communication apparatus can further send the M output data to the first communication apparatus, so that the first communication apparatus can determine the seventh information based on the M output data.

[0148] In a possible implementation of the fourth aspect, the method further includes: the second communication apparatus receiving the first input data.

[0149] Based on the above scheme, the second communication apparatus can further receive the first input data, so that the second communication apparatus can determine the M output data by deploying the N models (for example, N pre-trained large models) in the second communication apparatus or other communication apparatus connected to the second communication apparatus, and so that the first communication apparatus does not need to deploy the N models, thereby saving the storage space of the first communication apparatus and reducing the processing complexity of the first communication apparatus.

[0150] In a possible implementation of the fourth aspect, the method further includes: the second communication apparatus receiving ninth information, the ninth information indicating a request for the M output data.

[0151] Based on the above scheme, the second communication apparatus can further receive the ninth information, so that the second communication apparatus can provide the M output data for the first communication apparatus based on the request of the ninth information.

[0152] Optionally, the ninth information indicates at least one of the following: a task identifier, model information of part or all of the N models, a processing mode, a generation mode of the first input data, or task auxiliary information.

[0153] In a possible implementation of the fourth aspect, the method further includes: the second communication apparatus sending tenth information, the tenth information being used for indicating configuration information corresponding to the request.

[0154] Based on the above scheme, after receiving the ninth information, the second communication apparatus can determine the configuration information based on the ninth information, and indicate the configuration information to the first communication apparatus through the tenth information, so that the first communication apparatus can determine the configuration information corresponding to the processing performed by the second communication apparatus on the request based on the configuration information.

[0155] Optionally, the configuration information indicated by the tenth information can be the same as the content of the third information request (e.g., the model information of part or all of the N models, or the task assistance information, etc.), or can be partially different or completely different from the content of the ninth information request, which is not limited here.

[0156] Optionally, in the case that the configuration information indicated by the tenth information is partially different or completely different from the content of the ninth information request, the first communication device can indicate the second communication device to reject (e.g., reject the second communication device to process the first input data based on the configuration information), so that the second communication device does not need to process based on the configuration information that the first communication device does not expect, to reduce processing and transmission overhead. And / or, the first communication device can determine / generate / acquire the next request (e.g., the ninth information sent next) based on the configuration information, to improve processing efficiency.

[0157] In a possible implementation of the fourth aspect, the configuration information indicates at least one of: a groupcast identifier corresponding to the first communication device, part or all of the N models, or a processing mode; wherein the M output data is carried in groupcast information, and the groupcast information includes the groupcast identifier.

[0158] Based on the above scheme, the configuration information indicated by the tenth information can include at least one of the above, to improve the flexibility of the scheme implementation.

[0159] For example, in the case that the above configuration information indicates the groupcast identifier corresponding to the first communication device, the first communication device can obtain the M output data through the received groupcast information, and in the case that the number of first communication devices is greater than 1, the transmission overhead of the output data can be reduced through the groupcast transmission mode.

[0160] For another example, in the case that the above configuration information indicates part or all of the N models, the first communication device can provide the first input data matched with the input of the part or all of the models through the indication.

[0161] For another example, in the case that the above configuration information indicates the processing mode for processing the first input data, the first communication device can obtain the target output matched with the processing mode based on the indication of the configuration information.

[0162] In a possible implementation of the fourth aspect, the first communication device obtains the M output data, including: the first communication device processes the first input data based on the N models to obtain the M output data.

[0163] Based on the above scheme, the first communication device can process the first input data through N models deployed locally (or through an external device of the first communication device, such as a terminal device with AI function, a server, etc.) to obtain M output data, so as to reduce transmission overhead.

[0164] In a possible implementation manner of the fourth aspect, the method further includes: the second communication device sends the first input data.

[0165] Based on the above scheme, the first communication device can send the first input data to the first communication device, so that the first communication device can obtain M output data through processing of the first input data.

[0166] Optionally, the first input data can be preconfigured data, or the first data can be data collected by the first communication device itself, so as to reduce transmission overhead.

[0167] The fifth aspect of the present application provides a communication device, which is a first communication device, and the device includes a processing unit; the processing unit is configured to obtain M output data, the M output data being obtained by processing first input data through N models, N being a positive integer, and M being an integer greater than or equal to N; the processing unit is further configured to determine label data corresponding to the first input data based on the M output data, wherein the label data is used to determine a first model, and the first model is used to determine a target output corresponding to second input data from one or more output data corresponding to the second input data.

[0168] In the fifth aspect of the present application, the component modules of the communication device can also be used to perform the steps performed in the various possible implementation manners of the first aspect and achieve the corresponding technical effects. For details, please refer to the first aspect, which will not be described here.

[0169] The sixth aspect of the present application provides a communication device, which is a second communication device, and the device includes a transceiver and a processing unit; the transceiver is configured to receive first information, the first information being used to indicate label data corresponding to first input data; wherein the label data is determined based on M output data obtained by processing the first input data through N models, N being a positive integer, and M being an integer greater than or equal to N; the processing unit is configured to determine a first model based on the label data; wherein the first model is used to determine output data corresponding to second input data from one or more output data corresponding to the second input data.

[0170] In the sixth aspect of the present application, the component modules of the communication device can also be used to perform the steps performed in the various possible implementation manners of the second aspect and achieve the corresponding technical effects. For details, please refer to the second aspect, which will not be described here.

[0171] The seventh aspect of the present application provides a communication device, which is a first communication device, comprising a transceiver unit and a processing unit; the processing unit is configured to obtain M output data, the M output data being obtained by processing a first input data by N models, N being a positive integer, and M being an integer greater than or equal to N; the processing unit is further configured to determine seventh information based on the M output data, the seventh information being used to determine a second input data and / or model parameters of one or more models used to process the second input data, the second input data being used to be processed by the one or more models to obtain a target output corresponding to the second input data.

[0172] In the seventh aspect of the present application, the constituent modules of the communication device can also be configured to perform the steps performed in the various possible implementation manners of the third aspect and achieve the corresponding technical effects, which can be referred to the third aspect for details and will not be described herein again.

[0173] The eighth aspect of the present application provides a communication device, which is a second communication device, comprising a transceiver unit and a processing unit; the transceiver unit is configured to receive the seventh information, the seventh information being obtained based on M output data obtained by processing a first input data by N models, N being a positive integer, and M being an integer greater than or equal to N; the processing unit is configured to determine a second input data and / or model parameters of one or more models used to process the second input data based on the seventh information, the second input data being used to be processed by the one or more models to obtain a target output corresponding to the second input data.

[0174] In the eighth aspect of the present application, the constituent modules of the communication device can also be configured to perform the steps performed in the various possible implementation manners of the fourth aspect and achieve the corresponding technical effects, which can be referred to the fourth aspect for details and will not be described herein again.

[0175] The ninth aspect of the present application provides a communication device, comprising at least one processor, the at least one processor being coupled with a memory; the memory is configured to store programs or instructions; the at least one processor is configured to execute the programs or instructions to enable the device to implement the method in any one of the possible implementation manners of any one of the first aspect to the fourth aspect. Optionally, the communication device can comprise the memory.

[0176] The tenth aspect of the present application provides a communication device, comprising at least one logic circuit and an input-output interface; the logic circuit is configured to execute the method in any one of the possible implementation manners of any one of the first aspect to the fourth aspect.

[0177] The eleventh aspect of the present application provides a communication system, comprising the first communication device and the second communication device.

[0178] The twelfth aspect of the present application provides a computer readable storage medium for storing one or more computer-executable instructions, which, when executed by a processor, cause the processor to perform the method of any possible implementation of any one of the first aspect to the fourth aspect.

[0179] The thirteenth aspect of the present application provides a computer program product (or computer program), which, when executed by a processor, causes the processor to perform the method of any possible implementation of any one of the first aspect to the fourth aspect.

[0180] The fourteenth aspect of the present application provides a chip or chip system, which includes at least one processor for supporting a communication apparatus to implement the method of any possible implementation of any one of the first aspect to the fourth aspect. For example, the chip can be a baseband chip, a modem chip, a system on chip (SoC) chip containing a modem core, a system in package (SIP) chip, or a communication module, etc.

[0181] In a possible design, the chip or chip system can further include a memory for storing necessary program instructions and data of the communication apparatus. The chip system can be composed of a chip, or can include a chip and other discrete devices. Optionally, the chip system further includes an interface circuit for providing program instructions and / or data for the at least one processor.

[0182] The technical effects brought by the fifth aspect to the fourteenth aspect can be referred to the technical effects brought by the first aspect to the fourth aspect, which will not be repeated here. BRIEF DESCRIPTION OF DRAWINGS

[0183] FIGS. 1a to 1c are schematic diagrams of a communication system provided by the present application;

[0184] FIGS. 2a to 2g are schematic diagrams of an AI processing procedure related to the present application;

[0185] FIG. 3 is an interaction schematic diagram of a communication method provided by the present application;

[0186] FIGS. 4a to 4c are some schematic diagrams of a model provided by the present application;

[0187] FIG. 5 is another interaction schematic diagram of a communication method provided by the present application;

[0188] FIG. 6 to FIG. 10 are schematic diagrams of the communication apparatus provided in the present application. DETAILED DESCRIPTION

[0189] First, some terms in the embodiments of the present application are explained to facilitate understanding by those skilled in the art.

[0190] (1) Terminal device: can be a wireless terminal device capable of receiving network device scheduling and indication information, the wireless terminal device can be a device that provides voice and / or data connectivity to a user, or a handheld device with a wireless connection function, or other processing devices connected to a wireless modem.

[0191] The terminal device can communicate with one or more core networks or the Internet via a radio access network (RAN), and the terminal device can be a mobile terminal device, such as a mobile phone (or called "cellular" phone, mobile phone), computer and data card, for example, can be portable, pocket-sized, handheld, computer built-in or vehicle-mounted mobile devices that exchange voice and / or data with the radio access network. For example, personal communication service (PCS) phones, cordless phones, session initiation protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistants (PDAs), tablets, wireless transceiver-enabled computers, and the like. The wireless terminal device can also be called a subscriber unit, a subscriber station, a mobile station, a remote station, an access point, a remote terminal, an access terminal, a user terminal, a user agent, a subscriber station, a customer premises equipment, a terminal, a user equipment, a mobile terminal, and the like.

[0192] By way of example and not limitation, in embodiments of the present application, the terminal device can also be a wearable device. The wearable device can also be referred to as a smart wearable device or a smart wearable device, etc., which is a general term for devices that apply wearable technology to the intelligent design and development of daily wear, such as glasses, gloves, watches, clothing, and shoes, etc. The wearable device is a portable device that can be directly worn on the body or integrated into the user's clothes or accessories. The wearable device is not just a hardware device, but also has powerful functions through software support and data interaction, cloud interaction. The broad sense of wearable smart devices includes devices with full functions, large sizes, and the ability to realize complete or partial functions without relying on smart phones, such as smart watches or smart glasses, etc., and devices that focus only on a certain application function and need to be used with other devices such as smart phones, such as various smart wristbands, smart helmets, smart jewelry, etc.

[0193] The terminal can also be a drone, a robot, a terminal in device-to-device (D2D) communication, a terminal in vehicle to everything (V2X), a virtual reality (VR) terminal device, an augmented reality (AR) terminal device, a wireless terminal in industrial control, a wireless terminal in self driving, a wireless terminal in telemedicine or telehealth services, a wireless terminal in smart grid, a wireless terminal in transportation safety, a wireless terminal in smart city, a wireless terminal in smart home, etc.

[0194] In addition, the terminal device can also be a terminal device in a future communication system (such as a 5G Advanced communication system, etc.) after the 5th generation (5G) communication system or a terminal device in a future evolved public land mobile network (PLMN), etc. For example, the 5G Advanced network can further expand the form and function of the 5G communication terminal, and the 5G Advanced terminal includes but is not limited to vehicles, cellular network terminals (with satellite terminal functions), drones, and internet of things (IoT) devices.

[0195] In the embodiments of the present application, the terminal device can also obtain an artificial intelligence (AI) service provided by the network device. Optionally, the terminal device can also have AI processing capability.

[0196] (2) Network device: can be a device in a wireless network, for example, the network device can be a RAN node (or device) for accessing the terminal device to the wireless network, which can also be referred to as a base station. At present, some examples of RAN devices are: base station, evolved NodeB (eNodeB), base station gNB (gNodeB) in 5G communication system, transmission reception point (TRP), evolved Node B (eNB), radio network controller (RNC), Node B (NB), home base station (for example, home evolved Node B, or home Node B, HNB), baseband unit (BBU) or wireless fidelity (Wi-Fi) access point (AP) and the like. In addition, in one network structure, the network device can include a central unit (CU) node, or a distributed unit (DU) node, or a RAN device including a CU node and a DU node.

[0197] Optionally, the RAN node can also be a macro base station, a micro base station or an indoor station, a relay node or a donor node, or a wireless controller in a cloud radio access network (CRAN) scenario. The RAN node can also be a server, a wearable device, a vehicle or a vehicle-mounted device, etc. For example, the access network device in vehicle-to-everything (V2X) technology can be a road side unit (RSU).

[0198] In another possible scenario, multiple RAN nodes cooperate to assist a terminal to implement wireless access, and different RAN nodes respectively implement part of functions of a base station. For example, a RAN node can be a CU, a DU, a CU-control plane (CP), a CU-user plane (UP), or a radio unit (RU), etc. The CU and the DU can be separately configured, or can be included in the same network element, for example, in a baseband unit (BBU). The RU can be included in a radio frequency device or a radio frequency unit, for example, in a remote radio unit (RRU), an active antenna unit (AAU), a radio head (RH), or a remote radio head (RRH).

[0199] In different systems, the CU (or CU-CP and CU-UP), the DU, or the RU can also have different names, but those skilled in the art can understand their meanings. For example, in an open RAN (O-RAN or ORAN) system, the CU can also be referred to as an O-CU (open CU), the DU can also be referred to as an O-DU, the CU-CP can also be referred to as an O-CU-CP, the CU-UP can also be referred to as an O-CU-UP, and the RU can also be referred to as an O-RU. For the convenience of description, the CU, the CU-CP, the CU-UP, the DU, and the RU are taken as examples for description in this application. Any one of the CU (or the CU-CP, the CU-UP), the DU, and the RU in this application can be implemented by a software module, a hardware module, or a combination of a software module and a hardware module.

[0200] The communication between the access network device and the terminal device complies with a certain protocol layer structure. The protocol layer can include a control plane protocol layer and a user plane protocol layer. The control plane protocol layer can include at least one of the following: a radio resource control (RRC) layer, a packet data convergence protocol (PDCP) layer, a radio link control (RLC) layer, a media access control (MAC) layer, or a physical (PHY) layer, etc. The user plane protocol layer can include at least one of the following: a service data adaptation protocol (SDAP) layer, a PDCP layer, an RLC layer, a MAC layer, or a physical layer, etc.

[0201] For the correspondence between the network elements in the ORAN system and the protocol layer functions that can be implemented by the network elements, refer to Table 1 below.

[0202] Table 1

[0203] The network device can be another device that provides a wireless communication function for the terminal device. Embodiments of the present application do not limit the specific technology and specific device form adopted by the network device. For the convenience of description, embodiments of the present application do not limit.

[0204] The network device can also include a core network device, for example, a mobility management entity (MME) in a fourth generation (4G) network, a home subscriber server (HSS), a serving gateway (S-GW), a policy and charging rules function (PCRF), a public data network gateway (PDN gateway or P-GW), a network element such as an access and mobility management function (AMF) in a 5G network, a user plane function (UPF), or a session management function (SMF). In addition, the core network device can also include other core network devices in the 5G network and the next generation network of the 5G network.

[0205] In the embodiments of the present application, the network device described above can also be an AI-capable network node, which can provide AI services for terminals or other network devices, for example, AI nodes, computing power nodes, AI-capable RAN nodes, AI-capable core network elements, etc. on the network side (access network or core network).

[0206] In the embodiments of the present application, the device for implementing the function of the network device can be a network device or a device capable of supporting the network device to implement the function, such as a chip system, which can be arranged in the network device. In the technical solutions provided in the embodiments of the present application, the device for implementing the function of the network device is taken as an example to describe the technical solutions provided in the embodiments of the present application.

[0207] (3) Configuration and pre-configuration: in this application, configuration and pre-configuration will be used at the same time. Among them, configuration refers to that the network device / server sends some parameter configuration information or parameter values to the terminal through messages or signaling, so that the terminal determines the communication parameters or resource in transmission according to the values or information. Pre-configuration is similar to configuration, which can be parameter information or parameter values agreed by the network device / server and the terminal device in advance, or parameter information or parameter values adopted by the base station / network device or the terminal device according to the standard protocol, or parameter information or parameter values pre-stored in the base station / server or the terminal device. This application does not limit it.

[0208] Further, these values and parameters can be changed or updated.

[0209] (4) The terms "system" and "network" in the embodiments of the present application can be used interchangeably. "Multiple" means two or more. "And / or" describes the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B, which means that A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. The character " / " generally represents an "or" relationship between the associated objects. "At least one of the following" or similar expressions means any combination of these items, including any combination of single item or multiple items. For example, "at least one of A, B and C" includes A, B, C, AB, AC, BC or ABC. In addition, unless otherwise specified, the ordinal numbers "first", "second", etc. mentioned in the embodiments of the present application are used to distinguish multiple objects, and are not used to limit the order, time sequence, priority or importance of multiple objects.

[0210] (5) In the embodiments of the present application, "sending" and "receiving" represent the direction of signal transmission. For example, "sending information to XX" can be understood as that the destination of the information is XX, which can include direct sending through the air interface, or indirect sending through the air interface by other units or modules. "Receiving information from YY" can be understood as that the source of the information is YY, which can include direct receiving from YY through the air interface, or indirect receiving from YY through the air interface by other units or modules. "Sending" can also be understood as "output" of chip interface, and "receiving" can also be understood as "input" of chip interface.

[0211] In other words, sending and receiving can be carried out between devices, such as between network devices and terminal devices, or within devices, such as between components, modules, chips, software modules or hardware modules within devices through buses, wires or interfaces.

[0212] It can be understood that the information can be processed, such as encoding and modulation, between the source end and the destination end of the information transmission, but the destination end can understand the effective information from the source end. Similar expressions in this application can be similarly understood, and will not be repeated here.

[0213] (6) In the embodiments of the present application, “indication” can include direct indication and indirect indication, and can also include explicit indication and implicit indication. The information indicated by certain information (indication information described below) is referred to as to-be-indicated information. In the implementation process, there are many ways to indicate the to-be-indicated information, for example, but not limited to, the to-be-indicated information can be directly indicated, such as the to-be-indicated information itself or the index of the to-be-indicated information. The to-be-indicated information can also be indirectly indicated by indicating other information, where the other information and the to-be-indicated information have an association relationship. The to-be-indicated information can also be indicated only by a part, and the other part of the to-be-indicated information is known or agreed in advance. For example, the indication of a specific information can be achieved by means of the arrangement order of each information agreed in advance (for example, protocol predefined), thereby reducing the indication overhead to a certain extent. The present application does not limit the specific manner of indication. It can be understood that for the sender of the indication information, the indication information can be used to indicate the to-be-indicated information, and for the receiver of the indication information, the indication information can be used to determine the to-be-indicated information.

[0214] In the present application, the same or similar parts of each embodiment can be mutually referred to, unless otherwise specified. In the various embodiments of the present application, and the various methods / designs / implementation manners in each embodiment, the terms and / or descriptions of different embodiments, and the various methods / designs / implementation manners in each embodiment are consistent and can be mutually referred to, unless otherwise specified and logically conflicted. The technical features of different embodiments, and the various methods / designs / implementation manners in each embodiment can be combined to form new embodiments, methods, or implementation manners according to their inherent logical relationship. The implementation manners of the present application described below do not constitute a limitation on the protection scope of the present application.

[0215] The present application can be applied to a long term evolution (LTE) system, a new radio (NR) system, or a future communication system after 5G. The communication system includes at least one network device and / or at least one terminal device.

[0216] Referring to FIG. 1a, an architecture diagram of a communication system 1000 to which embodiments of the present application are applied is shown. As shown in FIG. 1a, the communication system can include a radio access network (RAN) 100, and optionally, the communication system 1000 can further include a core network 200 and an Internet 300. The RAN 100 includes at least one RAN node (e.g., 110a and 110b in FIG. 1a, collectively referred to as 110), and can further include at least one terminal (e.g., 120a-120j in FIG. 1a, collectively referred to as 120). The RAN 100 can further include other RAN nodes, such as a wireless relay device and / or a wireless backhaul device (not shown in FIG. 1a). The terminal 120 is connected to the RAN node 110 in a wireless manner, and the RAN node 110 is connected to the core network 200 in a wireless or wired manner. The core network device in the core network 200 and the RAN node 110 in the RAN 100 can be independent and different physical devices, or can be the same physical device integrated with the logical functions of the core network device and the logical functions of the RAN node. Terminals and terminals, and RAN nodes and RAN nodes can be connected to each other in a wired or wireless manner.

[0217] Taking the communication system shown in FIG. 1a as an example, different devices (including network devices and network devices, network devices and terminal devices, and / or terminal devices and terminal devices) can perform AI-related services in addition to performing communication-related services.

[0218] As shown in FIG. 1b, taking a base station as an example, the base station can perform communication-related services and AI-related services between one or more terminal devices, and communication-related services and AI-related services can also be performed between different terminal devices.

[0219] As shown in FIG. 1c, taking a terminal device including a television and a mobile phone as an example, the television and the mobile phone can also perform communication-related services and AI-related services.

[0220] The technical solutions provided in the present application can be applied to a wireless communication system (for example, the system shown in FIG. 1a, FIG. 1b or FIG. 1c), for example, an AI network element can be introduced in the communication system provided in the present application to implement part or all of the AI related operations. The AI network element can also be referred to as an AI node, an AI device, an AI entity, an AI module, an AI model, or an AI unit, etc. The AI network element can be built-in in a network element of the communication system. For example, the AI network element can be an AI module built-in in an access network device, a core network device, a cloud server, or an operation, administration and maintenance (OAM) to implement AI related functions. The OAM can be a network management of the core network device and / or a network management of the access network device. Alternatively, the AI network element can also be a network element independently arranged in the communication system. Optionally, an AI entity can also be included in a terminal or a chip built-in in the terminal to implement AI related functions.

[0221] Optionally, in the communication system, the AI application cases can include but are not limited to: channel state information (CSI) feedback enhancement, beam management enhancement, positioning accuracy enhancement, network energy saving, load balancing, and mobility optimization. The following will be described respectively.

[0222] 1. CSI feedback enhancement

[0223] CSI is the channel property of the communication link, and is the channel quality information reported by the terminal device to the network device. The terminal device reports the channel quality information to the network device, so as to select a suitable modulation and coding scheme (MCS) for the terminal device, so as to adapt to the changing wireless channel. For example, the terminal device performs channel estimation according to the received channel state information-reference signal (CSI-RS), and then feeds back the channel quality information to the network device. The information is used as the input of the model of the network device, so as to realize the AI model training of the network device. By applying AI to the CSI feedback enhancement, the overhead can be reduced, the accuracy can be improved, and the prediction can be realized.

[0224] CSI-RS feedback enhancement can include at least one sub-function, such as: CSI compression, CSI prediction, and CSI-RS configuration signaling reduction, respectively. CSI compression can further include CSI compression in at least one of spatial, time, and frequency domains.

[0225] 2. Beam management enhancement

[0226] Beam management enhancement is mainly to find the strongest transmit / receive beam pair. AI-based sparse beam prediction can improve accuracy. According to AI training and inference, it can include AI sparse beam prediction on the network side and AI sparse beam prediction on the terminal device side. Taking AI sparse beam prediction on the terminal device side as an example, the pre-trained AI model on the terminal device side can be delivered by the network side or pre-stored on the terminal device side. In the training phase, the network device scans all possible beams, and then the network reports the transmit beam pattern to the terminal device. When the model training is completed, the network device only needs to scan a small part of the beam, and then the terminal device feeds back the inference result to the network device. AI-based beam management can realize, for example, beam prediction in time and / or spatial domain to reduce overhead and delay and improve beam selection accuracy.

[0227] Beam management enhancement can include at least one sub-function, such as: beam scanning matrix prediction, and optimal beam prediction, respectively.

[0228] 3. Positioning accuracy enhancement

[0229] In line of sight (LOS) or non-line of sight (NLOS) scenarios, AI-based positioning can improve positioning accuracy with a smaller number of TRP antennas. Positioning enhancement can include at least one sub-function, such as: access network device-based positioning enhancement, positioning management function network element-based positioning enhancement, and terminal device-based positioning enhancement, respectively.

[0230] 4. Network energy saving

[0231] Network energy saving can be achieved through cell activation / deactivation, load reduction, improved coverage, or other RAN setting adjustments. AI technology can be used to optimize energy saving decisions by utilizing data collected in the RAN network. AI algorithms can predict the energy efficiency and load status of the next period, which can be used to assist in decision-making for cell activation / deactivation to save energy. Based on the predicted load, the system can dynamically configure energy saving strategies to maintain a balance between system performance and energy efficiency and reduce energy consumption.

[0232] 5. Load balancing

[0233] Load balancing can make the load evenly distributed among cells and among areas within a cell, or divert part of the traffic from congested cells, or split users among cells, carriers or access technologies to improve network performance. AI model based load balancing can provide higher quality user experience and improve system capacity.

[0234] 6. Mobility management

[0235] Mobility management is a solution to ensure service continuity during terminal device movement by minimizing dropped calls, radio link failure (RLF), unnecessary handover and ping-pong effect. AI can enhance mobility management, such as reducing the probability of unexpected events, predicting terminal device location / mobility / performance, and traffic steering, etc.

[0236] It should be understood that the definitions of the above technical terms are only examples. For example, as technology continues to evolve, the scope of the above definitions can also change, and the embodiments of the present application are not limited.

[0237] For example, an AI function can include multiple AI sub-functions.

[0238] Optionally, the AI application case is also referred to as an AI application scenario or an AI function.

[0239] As described above, AI can be widely used in CSI feedback enhancement, beam management, positioning accuracy enhancement, energy saving, mobility enhancement, load balancing and other aspects to improve network performance. AI models can be deployed on the network side and / or the terminal device side, and the training of AI models depends on the collection of training data, which can come from terminal device measurement and feedback.

[0240] The concepts that can be involved in the present application will be briefly introduced below.

[0241] AI can give machines human-like intelligence, such as allowing machines to apply computer hardware and software to simulate some intelligent behavior of humans. To achieve artificial intelligence, machine learning methods can be used. In machine learning methods, machines learn (or train) models using training data. The model represents the mapping between input and output. The learned model can be used for inference (or prediction), i.e., the model can be used to predict the output corresponding to a given input. The output can also be referred to as the inference result (or prediction result).

[0242] Machine learning can include supervised learning, unsupervised learning, and reinforcement learning. Among them, unsupervised learning can also be referred to as non-supervised learning.

[0243] Supervised learning learns the mapping relationship from sample values to sample labels according to the collected sample values and sample labels, and uses an AI model to express the learned mapping relationship. The process of training a machine learning model is the process of learning such a mapping relationship. In the training process, the sample values are input into the model to obtain the predicted values of the model, and the model parameters are optimized by calculating the error between the predicted values of the model and the sample labels (ideal values). After the mapping relationship is learned, the learned mapping can be used to predict new sample labels. The learned mapping relationship of supervised learning can include linear mapping or nonlinear mapping. According to the type of label, the learned task can be divided into classification tasks and regression tasks.

[0244] Unsupervised learning uses algorithms to discover the internal patterns of samples according to the collected sample values. In unsupervised learning, a class of algorithms uses the sample itself as a supervision signal, that is, the model learns the mapping relationship from the sample to the sample, which is called self-supervised learning. In training, the model parameters are optimized by calculating the error between the predicted values of the model and the sample itself. Self-supervised learning can be used for signal compression and decompression recovery applications. Common algorithms include autoencoders and generative adversarial networks.

[0245] Reinforcement learning is different from supervised learning, and is a class of algorithms that learn strategies to solve problems by interacting with the environment. Unlike supervised and unsupervised learning, reinforcement learning problems do not have clear "correct" action label data. The algorithm needs to interact with the environment to obtain the reward signal of the environment feedback, and then adjust the decision action to obtain a larger reward signal value. In the following power control, the reinforcement learning model adjusts the downlink transmission power of each user according to the system total throughput rate feedback by the wireless network, and then expects to obtain a higher system throughput rate. The goal of reinforcement learning is also to learn the mapping relationship between the environment state and the optimal (for example, the optimal) decision action. However, because the "correct action" label cannot be obtained in advance, the network cannot be optimized by calculating the error between the action and the "correct action". Reinforcement learning training is achieved through iterative interaction with the environment.

[0246] A neural network (NN) is a specific model in machine learning technology. According to the universal approximation theorem, a neural network can theoretically approximate any continuous function, so that the neural network has the ability to learn any mapping. Traditional communication systems need to use rich expert knowledge to design communication modules, while a deep learning communication system based on a neural network can automatically discover the implicit pattern structure from a large amount of data set, establish the mapping relationship between the data, and obtain better performance than traditional modeling methods.

[0247] The idea behind neural networks comes from the neuronal structure of the brain. For example, each neuron performs a weighted summation of its input values ​​and outputs the result through an activation function.

[0248] Figure 2a shows a schematic diagram of a neuron structure. Assume the input to the neuron is x = [x0, x1, ..., x...]. n The weights corresponding to each input are w = [w0, w1, ..., w] n ], where n is a positive integer, w i and x i It can be any possible type, such as a decimal, an integer (e.g., 0, a positive integer, or a negative integer), or a complex number. i As x i The weights are used to assign weights to x. i Weighting is applied. The bias for the weighted sum of the input values ​​is, for example, b. Activation functions can take many forms. Suppose the activation function of a neuron is: y = f(z) = max(0, z), then the output of that neuron is: For example, if the activation function of a neuron is y = f(z) = z, then the output of that neuron is: Here, b can be any possible type, such as a decimal, an integer (e.g., 0, a positive integer, or a negative integer), or a complex number. The activation functions of different neurons in a neural network can be the same or different.

[0249] Furthermore, neural networks generally consist of multiple layers, each of which may include one or more neurons. Increasing the depth and / or width of a neural network can improve its expressive power, providing more powerful information extraction and abstract modeling capabilities for complex systems. The depth of a neural network can refer to the number of layers it includes, and the number of neurons in each layer can be called the width of that layer. In one implementation, a neural network includes an input layer and an output layer. The input layer processes the received input information through neurons and passes the processing result to the output layer, which then obtains the output of the neural network. In another implementation, a neural network includes an input layer, hidden layers, and an output layer. The input layer processes the received input information through neurons and passes the processing result to the hidden layer. The hidden layer calculates the received processing result and passes the calculation result to the output layer or the next adjacent hidden layer, ultimately obtaining the output of the neural network. A neural network may include one hidden layer or multiple sequentially connected hidden layers, without limitation.

[0250] The neural network is, for example, a deep neural network (DNN). According to the construction manner of the network, the DNN can include a feedforward neural network (FNN), a convolutional neural network (CNN) and a recurrent neural network (RNN).

[0251] FIG. 2b is a schematic diagram of a FNN network. The FNN network is characterized by that the neurons in adjacent layers are fully connected to each other. This feature makes the FNN usually need a large amount of storage space and lead to a high computational complexity.

[0252] The CNN is a neural network specially designed to process data with a similar grid structure. For example, time series data (e.g. time axis discrete sampling) and image data (e.g. two-dimensional discrete sampling) can be considered as data with a similar grid structure. The CNN does not use all the input information for operation at one time, but uses a fixed size window to extract part of the information for convolution operation, which greatly reduces the calculation amount of model parameters. In addition, according to the different types of information extracted by the window (such as people and objects in the same image are different types of information), each window can use different convolution kernel operations, which makes the CNN better extract the features of the input data.

[0253] The RNN is a kind of neural network that uses feedback time series information. The input of the RNN includes the new input value at the current time and the output value of itself at the previous time. The RNN is suitable for obtaining sequence features with temporal correlation, such as speech recognition, channel coding and decoding applications.

[0254] In the above model training process of machine learning, a loss function can be defined. The loss function describes the gap or difference between the output value of the model and the ideal target value. The loss function can be embodied in various forms, and the specific form of the loss function is not limited. The model training process can be regarded as the following process: by adjusting part or all of the parameters of the model, the value of the loss function is less than the threshold value or meets the target demand.

[0255] The model can also be referred to as an AI model, a rule, or other names, etc. The AI model can be considered as a specific method to implement an AI function. The AI model represents a mapping relationship or a function between the input and the output of the model. The AI function can include one or more of the following: data collection, model training (or model learning), model information publishing, model inference (or model reasoning, reasoning, or prediction, etc.), model monitoring or model verification, or inference result publishing, etc. The AI function can also be referred to as an AI (related) operation, or an AI-related function.

[0256] The implementation process of the neural network will be described below with reference to the accompanying drawings.

[0257] 1. Fully connected neural network, also known as multilayer perceptron (MLP).

[0258] As shown in FIG. 2c, an MLP includes an input layer (left side), an output layer (right side), and multiple hidden layers (middle). Each layer of the MLP includes a number of nodes, referred to as neurons. The neurons of adjacent two layers are connected to each other.

[0259] Optionally, considering the neurons of adjacent two layers, the output h of the neuron of the next layer is the weighted sum of all the neurons x of the previous layer connected to it and is processed by an activation function, which can be represented as: h = f(wx + b).

[0260] where w is a weight matrix, b is a bias vector, and f is an activation function.

[0261] Further optionally, the output of the neural network can be recursively expressed as: y = f z (w z f z-1 (…)+b z ).

[0262] where z is the index of the layer of the neural network, z is greater than or equal to 1, and z is less than or equal to Z, where Z is the total number of layers of the neural network.

[0263] In other words, the neural network can be understood as a mapping relationship from a set of input data to a set of output data. Usually, the neural network is randomly initialized, and the process of obtaining this mapping relationship from the random w and b with the existing data is called training of the neural network.

[0264] Optionally, the specific way of training is to evaluate the output result of the neural network by using a loss function.

[0265] As shown in FIG. 2d, the error can be back-propagated, and the neural network parameters (including w and b) can be iteratively optimized by the method of gradient descent until the output of the loss function reaches a minimum value, i.e., the "better point (e.g., optimal point)" in FIG. 2d. It can be understood that the neural network parameters corresponding to the "better point (e.g., optimal point)" in FIG. 2d can be used as the neural network parameters in the trained AI model information.

[0266] Further optionally, the process of gradient descent can be represented as:

[0267] wherein θ is the parameter to be optimized (including w and b), L is the loss function, η is the learning rate, and controls the step size of gradient descent, represents the derivation operation, represents the derivative of L with respect to θ.

[0268] Further optionally, the process of back-propagation utilizes the chain rule of partial derivative.

[0269] As shown in FIG. 2e, the gradient of the parameters of the previous layer can be recursively calculated from the gradient of the parameters of the next layer, which can be expressed as:

[0270] wherein w ij is the weight of node j connected to node i, and s i is the input weighted sum on node i.

[0271] 2. Federated Learning (FL).

[0272] The concept of federated learning effectively solves the difficulties faced by the current development of artificial intelligence. Under the premise of fully guaranteeing the privacy and security of user data, the learning task of the model is efficiently completed by promoting the cooperation of various edge devices and central servers.

[0273] As shown in FIG. 2f, the FL architecture is the most widely used training architecture in the current FL field, and the FedAvg algorithm is the basic algorithm of FL. The algorithm process of FedAvg is roughly as follows:

[0274] (1) The central end initializes the model to be trained and broadcasts it to all clients.

[0275] (2) In the t-th round t∈[1, T], the client k∈[1, K] trains the received global model based on the local data set to obtain the local training result ​​Report it to the center node. In the example shown in Figure 2f, the local training results sent by the distributed nodes n, k, and m are respectively denoted as G n , k , m .

[0276] (3) The center node collects the local training results from all (or part) of the clients, assuming that the set of clients uploading the local model in the t-th round is The center end will obtain a new global model by weighted averaging with the sample number of the corresponding client as the weight, and the specific updating rule is After that, the center end broadcasts the latest version of the global model to all clients for a new round of training.

[0277] (4) Repeat steps (2) and (3) until the model converges or the number of training rounds reaches the upper limit.

[0278] Optionally, in addition to reporting the local model , the client can also report the trained local gradient , and the center node will average all the local gradients reported by the clients and update the global model according to the average gradient.

[0279] As can be seen, in the FL framework, the data set exists in the distributed nodes (such as the client), that is, the distributed nodes collect the local data set and perform local training, and report the local results (model or gradient) obtained by training to the center node. The center node itself may not have a data set, and can be responsible for fusing the training results of the distributed nodes to obtain a global model and issuing it to the distributed nodes.

[0280] 3. Decentralized learning.

[0281] As shown in Figure 2g, it is a completely distributed system without a center node. The design goal f(x) of the decentralized learning system is generally the average of the goals f i (x) of each node, that is, where n is the number of distributed nodes, and x is the parameter to be optimized. In machine learning, x is the parameter of the machine learning (such as neural network) model. Each node calculates the local gradient i using the local data and the local goal f , and then sends it to the neighbor nodes that are communicatively reachable. After receiving the gradient information sent by the neighbor nodes, any node can update the parameter x of the local model according to the following formula:

[0282] where, denotes the parameters of the local model of the i-th node after the k+1-th (k is a natural number) update, denotes the parameters of the local model of the i-th node after the k-th update (if k is 0, denotes the parameters of the local model of the i-th node that does not participate in the update) denotes the parameters of the local model of the i-th node that does not participate in the update), a k denotes a tuning coefficient, N i is a set of neighbor nodes of node i, |N i denotes the number of elements in the set of neighbor nodes of node i, that is, the number of neighbor nodes of node i. Through information interaction between nodes, the decentralized learning system will eventually learn a unified model.

[0283] The technical solutions provided by the present application can be applied in a communication system (such as the system shown in FIG. 1a or FIG. 1b or FIG. 1c). In the communication system, the communication nodes generally have signal transceiving capability and computing capability. Taking a network device with computing capability as an example, the computing capability of the network device is mainly to provide computing power support for the signal transceiving capability (such as: signal sending processing and receiving processing) to realize the communication task of the network device and other communication nodes.

[0284] With the development of communication technology, in the communication system, the services performed by the communication devices can include other new services in addition to traditional communication services, such as artificial intelligence (AI) services. Generally, a system capable of processing AI services, such as a communication system, can also be referred to as an AI system. However, in the AI system, how to improve user experience is a technical problem to be solved.

[0285] In a possible implementation manner, the complexity of the AI model deployed by the communication device is improved (such as increasing the number of parameters of the model, increasing the number of neural network layers contained in the model, etc.) by using the scaling law, which can effectively improve the performance of the model to improve the user experience.

[0286] As an example of the scaling law, the AI model deployed in the communication device can learn the knowledge of a pre-trained large model with high complexity deployed in the cloud. The former can be referred to as a student model and the latter can be referred to as a teacher model, so as to improve the performance of the student model (such as improving the inference accuracy, inference precision, etc. of the student model) by means of knowledge distillation (or knowledge transfer) to improve the user experience.

[0287] As another example of the scaling law, the AI model deployed in the communication device can be trained by a large amount of personalized data, which can improve the performance of the AI model (such as improving the inference accuracy, inference precision, etc. of the student model) to improve the user experience.

[0288] However, in the AI system, the use of the scaling law will inevitably increase the cost and power consumption, resulting in limited application scenarios of this method.

[0289] To solve the above problems, the present application provides a communication method and related devices, which will be described in detail below in conjunction with the accompanying drawings.

[0290] Please refer to FIG. 3, which is an implementation schematic diagram of the communication method provided by the present application. The method includes the following steps.

[0291] It should be noted that in FIG. 3 and the related implementation examples below, the first communication device and other communication devices (such as the second communication device) are taken as examples of the execution subject of the interaction schematic to illustrate the method, but the present application does not limit the execution subject of the interaction schematic. For example, the communication device can be a communication equipment, or a chip, a baseband chip, a modem chip, a system on chip (SoC) chip containing a modem core, a system in package (SIP) chip, a communication module, a chip system, a processor, a logic module or software in the communication equipment, etc. Optionally, the communication equipment can be a terminal device or a network device (such as an access network device, an access network element, a core network element, or a core network device, etc.).

[0292] As an example, the first communication device can be a terminal device and the second communication device can be a network device, or both the first communication device and the second communication device are network devices. For example, the network device can be an access network device or a communication device (such as at least one of a CU, a DU, or a RU) in an ORAN system.

[0293] As another example, both the first communication device and the second communication device are terminal devices, i.e., the scheme shown in FIG. 3 can be applied to a sidelink communication scenario.

[0294] S301. The first communication device obtains M output data. The M output data is obtained by processing the first input data through N models, N is a positive integer, and M is an integer greater than or equal to N.

[0295] S302. The first communication device determines the label data corresponding to the first input data based on the M output data. The label data is used to determine the first model, and the first model is used to determine the target output corresponding to the second input data from one or more output data corresponding to the second input data.

[0296] It should be understood that the label data is used to determine the first model, which can be understood as the label data is used to determine the first model in a manner of model processing. For example, the model processing can include one or more of model training, model generation, model updating, model fine-tuning, model adjustment, model fine adjustment.

[0297] Optionally, taking an example of the label data being used to determine the first model in a manner of model training, since the label data is the label data corresponding to the first input data, therefore, in the model training process, the model training data of the first model can include the first input data in addition to the label data. In this way, the first model can be trained with more information to improve the performance of the first model.

[0298] In this application, the model can be replaced by other terms, such as AI model, neural network, neural network model, AI neural network model, machine learning model, or AI processing model, etc.

[0299] In this application, the target output can be replaced by other terms, such as target, desired output, ideal output, personalized output, or label, etc.

[0300] It should be understood that the M output data is obtained by processing the first input data by the N models, which can be understood as each model in the N models processes the first input data to obtain one or more output data; the M output data can include one or more output data corresponding to each model in the N models.

[0301] It should be understood that any output data in the M output data can be a model output or an intermediate result of one of the N models. In other words, one or more output data corresponding to each model in the N models can include a model output and / or an intermediate result of the each model, etc.

[0302] For example, one model can include one or more layers of neural networks, and the model output of the one model can be understood as part or all of the output of the last layer of neural networks in the one or more layers of neural networks, for example, the model output can include part or all of the output of the last layer of neural networks. In addition, the intermediate result of the one model can be understood as part or all of the output of at least one layer of neural networks other than the last layer of neural networks in the one or more layers of neural networks, for example, the intermediate result can include part or all of the output of each layer of neural networks in the at least one layer of neural networks.

[0303] As an example, in the above scheme, any model in the N models is used to process the first input data to obtain one or more output data, i.e., the model processing performed by the any model can include one or more of inference, prediction, derivation, identification, decision-making. Correspondingly, the any model can be referred to as an inference model, a prediction model, a pre-trained model for the above model processing, or a pre-trained large model for the above model processing, etc.

[0304] Optionally, in the case where any model in the N models is an inference model, the first input data and the second input data can be the same or different inference data. Among them, the M output data corresponding to the first input data can be used for the determination of the inference model, and for this purpose, the first input data can be unlabeled data.

[0305] Based on the scheme shown in FIG. 3, the M output data obtained by the first communication device in step S301 is obtained by processing the first input data by the N models, and thereafter, the first communication device can determine the label data corresponding to the first input data based on the M output data in step S302. Among them, the label data is used to determine the first model, and the first model is used to determine the target output corresponding to the second input data from one or more output data corresponding to the second input data. In other words, the first model is used to determine the personalized target output, i.e., the first model can be a personalized model, and the above label data can be used to determine the personalized model. Thus, the label data determined by the first communication device through the M output data can be used to determine the personalized model, and through the personalized model, personalized target output can be provided to improve user experience.

[0306] In addition, in the case where the M output data is a plurality of output data (e.g., M is greater than 1), the above label data can be obtained by processing the corresponding diversified output data by the N models, which can improve the performance of the label data and further improve the performance of the first model based on the label data.

[0307] As an example, in the above scheme, the first model is used to determine the target output corresponding to the second input data from one or more output data corresponding to the second input data; wherein the model processing performed by the first model can include evaluation, ranking, optimization, screening, selection, or filtering, etc. Correspondingly, the first model can be referred to as an evaluation model, a personalized model, a personalized evaluation model, a reward model, or a scoring model, etc.

[0308] Optionally, the first model can include one or more sub-models, each of which can process one or more output data corresponding to the second input data to obtain a corresponding output, and then the output corresponding to each of the one or more sub-models can be used to determine the target output corresponding to the second input data. In other words, the first model can determine the target output corresponding to the second input data through one or more sub-models.

[0309] Some implementation examples will be provided below to exemplarily describe the implementation process of the first model. For example, the first model is used to determine the target output corresponding to the second input data from one or more output data corresponding to the second input data, which includes:

[0310] The first model is used to determine the sixth information from one or more output data corresponding to the second input data, and the sixth information is used to determine the target output corresponding to the second input data; wherein the sixth information includes at least one of the following:

[0311] The first indication information indicates the ranking of the one or more output data; wherein the target output is one of the one or more output data;

[0312] The second indication information indicates the first index, and the first index is used to determine the target output in the one or more output data; wherein the target output is one of the one or more output data;

[0313] The third indication information indicates the target output; wherein the target output corresponding to the first input data is determined based on the one or more output data. For example, the target output is one of the one or more output data, or the target output is different from any of the one or more output data.

[0314] Therefore, in the process of model processing of the first model, the first model can determine the target output corresponding to the second input data in the above-mentioned various ways to improve the flexibility of the scheme implementation.

[0315] Exemplarily, taking the second input data through one or more models (for example, the foregoing N models, pre-trained large models, and the following N models are taken as an example) to obtain K (K is a positive integer) output data as an example, it satisfies:

[0316] Wherein, x represents the second input data, P(x) represents the processing of the second input data by the N models, y0, y1, …, y K-1 respectively represent the output data (i.e. K output data) obtained by each of the N models based on the second input data.

[0317] The model processing performed by the first model is taken as an example of scoring. The first model can perform processing on the i-th output data y i to meet:

[0318] wherein, s i represents the scoring of the i-th output data y i processed by the first model. i s i represents the scoring obtained by the processing.

[0319] For example, in a case where the value of the scoring is positively correlated with the performance, the greater the value of the scoring represented by s i indicates that the corresponding performance of y i is higher, and vice versa, the smaller the value of the scoring represented by s i indicates that the corresponding performance of y i is lower. For another example, in a case where the value of the scoring is negatively correlated with the performance, the smaller the value of the scoring represented by s i indicates that the corresponding performance of y i is higher, and vice versa, the greater the value of the scoring represented by s i indicates that the corresponding performance of y i is lower. Hereinafter, the case where the value of the scoring is positively correlated with the performance is taken as an example for description.

[0320] Optionally, the optimal scoring y best meets:

[0321] The sixth information can be implemented in various manners.

[0322] In a first manner, the sixth information includes first indication information, and the first indication information indicates the ranking of the one or more output data; and the target output is one of the one or more output data.

[0323] For example, the one or more output data corresponding to the second input data is taken as the above K output data, the first communication device can determine K scores corresponding to the K output data based on the processing process of , and determine the ranking of the K output data indicated by the first indication information based on the K scores. For example, the ranking of the K scores from high to low can indicate the ranking of the performances of the K output data corresponding to the K scores from high to low. For another example, the ranking of the K scores from low to high can indicate the ranking of the performances of the K output data corresponding to the K scores from low to high.

[0324] As shown in FIG. 4a, which is an example of the first mode. In FIG. 4a, taking the value of K as 3 as an example, the first model can calculate scores for the three output data (output data 1, output data 2, and output data 3, respectively) to obtain three scores (score 1, score 2, and score 3, respectively), and sort the three scores to obtain the first indication information.

[0325] Thus, in the first mode, the first communication device (or the second communication device) can determine the performance of the output data corresponding to the higher (or the highest) performance of the K output data through the first indication information, and determine the output data as the target output (i.e., the personalized target output) corresponding to the second input data.

[0326] In the second mode, the sixth information includes second indication information, the second indication information indicating a first index, the first index being used to determine the target output in the one or more output data; and the target output being one of the one or more output data.

[0327] For example, taking the one or more output data corresponding to the second input data as the above-mentioned K output data as an example, the first communication device can determine the K scores corresponding to the K output data based on the processing process of , and determine the index of the output data corresponding to the highest value of the K scores as the first index indicated by the second indication information.

[0328] As shown in FIG. 4b, which is an example of the second mode. In FIG. 4b, taking the value of K as 3 as an example, the first model can process the three output data (output data 1, output data 2, and output data 3, respectively) to obtain the second indication information.

[0329] Thus, in the second mode, the first communication device (or the second communication device in the following step H) can determine the output data corresponding to the higher (or the highest) performance of the K output data through the second indication information, and determine the output data as the target output (i.e., the personalized target output) corresponding to the second input data.

[0330] In the third mode, the sixth information includes third indication information, the third indication information indicating the target output; and the target output being one of the one or more output data, or the target output being different from any of the one or more output data.

[0331] For example, taking the one or more output data corresponding to the second input data as the above-mentioned K output data as an example, the first communication device can process (e.g., sum average, weighted average, etc.) the K output data to determine the target output corresponding to the second input data. For example, the processing process can be that the first communication device determines the target output corresponding to the second input data based on determining K scores corresponding to the K output data, and processing based on the K scores to determine a target output corresponding to the second input data.

[0332] Optionally, the target output corresponding to the second input data can be obtained based on at least one of the following:

[0333] X scores corresponding to X weighted results of X output data in the K output data, X being less than or equal to K; or

[0334] a weighted result corresponding to scores of Y output data in the K output data, Y being less than or equal to K;

[0335] Wherein, the X output data and the Y output data can be completely different output data, or can be partially or completely overlapped output data, which is not limited here.

[0336] As shown in FIG. 4c, it is an example of the third mode. In FIG. 4c, taking the value of K as 3 as an example, the first model can process 3 output data (output data 1, output data 2 and output data 3 respectively) to obtain third indication information.

[0337] Thus, in the third mode, the first communication device (or the second communication device) can obtain the target output (i.e. personalized target output) corresponding to the second input data based on the K output data through the third indication information.

[0338] In a possible implementation, the method shown in FIG. 3 further includes that the first communication device determines the first model based on the label data. In other words, after the first communication device determines the label data based on the M output data, the first communication device can further determine the first model based on the label data, so that the first communication device can realize the determination of the first model locally (or through an external device of the first communication device, such as a terminal device with AI function, a server, etc.), so as to reduce the transmission overhead.

[0339] In a possible implementation, the method shown in FIG. 3 further includes that the first communication device sends first information, the first information being used to indicate the label data. Thus, the receiver (such as the second communication device) of the first information can obtain the label data, and realize the determination of the first model based on the label data, so as to reduce the implementation complexity of the first communication device.

[0340] In a possible implementation, the method shown in FIG. 3 further includes that the first communication device receives or sends second information, the second information being used to indicate model information of the first model. Thus, after the communication device (e.g., the first communication device or the second communication device) determines the first model based on the label data, the communication device can further indicate the model information of the first model through the second information, so that the receiver of the second information can obtain the model information of the first model through the second information, so as to facilitate the sender or the receiver of the second information to perform model management on the first model based on the model information.

[0341] Optionally, the model information involved in the present application can indicate one or more of a model identifier (or index), a model parameter, a model structure, a sampling parameter, and a sampling number. For example, the model parameter can include one or more of a model hyperparameter and a model capability level. For example, in the case that the first model is an evaluation model, the second information indicating the model information of the first model can be understood as the second information indicating the model information of the evaluation model, including but not limited to one or more of an evaluation model identifier (or index), an evaluation model parameter, an evaluation model structure, a sampling parameter, and a sampling number.

[0342] For example, the sampling parameter can include at least one of the following:

[0343] Temperature coefficient: a coefficient for controlling probability normalization, for example, a higher temperature makes the output more random, and a lower temperature makes the output more deterministic;

[0344] Random coefficient: random sampling with a certain probability;

[0345] Top-K: sampling from K outputs with the highest probability;

[0346] Top-P: sampling from an output set with a cumulative probability greater than or equal to p.

[0347] Optionally, the model management involved in the present application can include one or more of model scheduling, model updating, model switching, or function fallback.

[0348] In a possible implementation, in step S301, the first communication device can obtain M output data in various ways, which will be introduced below in combination with some examples.

[0349] In the first implementation example, in step S301, the first communication apparatus obtains the M output data, including: the first communication apparatus receiving the M output data. Thus, the first communication apparatus can obtain the M output data by receiving the M output data, i.e., the N models used for processing the first input data can be deployed on other communication apparatuses, so that the first communication apparatus does not need to deploy the N models (e.g., N pre-trained large models), thereby saving the storage space of the first communication apparatus and reducing the processing complexity of the first communication apparatus.

[0350] In a possible implementation of the first implementation example, the method further includes: the first communication apparatus sending the first input data. Thus, the first communication apparatus can further send the first input data to one or more communication apparatuses that deploy the N models, so that the one or more communication apparatuses can process the first input data to obtain and send the M output data.

[0351] Optionally, the first input data can be preconfigured data, so that the first communication apparatus does not need to send the first input data, thereby reducing the overhead.

[0352] In a possible implementation of the first implementation example, the method further includes: the first communication apparatus sending third information, the third information indicating a request for the M output data. Thus, the first communication apparatus can further send the third information, so that a receiver (e.g., a second communication apparatus) of the third information can provide the M output data for the first communication apparatus based on the request of the third information.

[0353] Optionally, the third information indicates at least one of: a task identifier, model information of part or all of the N models, a processing mode, a generation manner of the first input data, or task auxiliary information.

[0354] As an example, the third information can be used to request the M output data, and the label data determined by the M output data can be used to determine a first model, where the processing of the first model corresponds to different processing modes and can obtain different output data (e.g., first, second, and third indication information included in the sixth information described below). Correspondingly, in the case where the third information indicates the processing mode, the receiver of the third information can provide the corresponding M output data based on the processing mode, so that the first communication apparatus obtains the M output data corresponding to the specified processing mode and obtains the first model of the specified processing mode. Optionally, in the case where the above processing is inference, the processing mode can be replaced by an inference mode.

[0355] As an example, the generation manner of the first input data can indicate that the first input data is generated in an offline manner or an online manner.

[0356] As an example, the task auxiliary information can include information required for model management. For example, in the case where the model is used for a communication task, the task auxiliary information can include configuration of communication parameters (e.g., resource parameter configuration, power control parameter configuration, etc.) of the communication device. For another example, in the case where the model is used for a classification task, the task auxiliary information can include a rough category of the input data. For another example, in the case where the model is used for a generation task, the task auxiliary information can include a prompt word.

[0357] In a possible implementation of the example one, the method further includes: receiving, by the first communication device, fourth information, the fourth information being used to indicate the configuration information corresponding to the request. Thus, the receiver (e.g., the second communication device) of the third information can determine the configuration information based on the third information after receiving the third information, and indicate the configuration information to the first communication device through the fourth information, so that the first communication device can determine the configuration information corresponding to the processing performed by the second communication device on the request based on the configuration information.

[0358] Optionally, the configuration information indicated by the fourth information can be the same as the content (e.g., model information of part or all of the N models, processing mode, or task auxiliary information, etc.) requested by the third information, or can be partially different or completely different from the content requested by the third information, which is not limited here.

[0359] Optionally, in the case where the configuration information indicated by the fourth information is partially different or completely different from the content requested by the third information, the first communication device can indicate rejection (e.g., rejection of the second communication device to process the first input data based on the configuration information) to the second communication device, so that the second communication device does not need to process based on the configuration information not expected by the first communication device, to reduce processing and transmission overhead. And / or, the first communication device can determine / generate / acquire the next request (e.g., the third information sent next time) based on the configuration information, to improve processing efficiency.

[0360] Optionally, the configuration information indicates at least one of: a multicast identifier corresponding to the first communication device, part or all of the N models, or a processing mode; wherein the M output data is carried in multicast information, and the multicast information includes the multicast identifier. Thus, the configuration information indicated by the fourth information can include the at least one above, to improve the flexibility of the scheme implementation.

[0361] For example, in the case where the configuration information indicates the multicast identifier corresponding to the first communication device, the first communication device can obtain the M output data through the received multicast information, and in the case where the number of the first communication devices is greater than 1, the multicast transmission mode can reduce the transmission overhead of the output data.

[0362] For example, in a case where the configuration information indicates part or all of the N models, the first communication apparatus can provide first input data matching the part or all of the input of the models according to the indication.

[0363] For example, in a case where the configuration information indicates a processing mode for processing the first input data, the first communication apparatus can obtain a target output matching the processing mode based on the indication of the configuration information.

[0364] In the second implementation example, in step S301, the first communication apparatus obtains M output data by processing the first input data based on the N models.

[0365] In a possible implementation of the second implementation example, the method further includes: receiving, by the first communication apparatus, the first input data. Thus, the first communication apparatus can receive the first input data, so that the first communication apparatus can obtain the M output data by processing the first input data.

[0366] Optionally, the first input data can be preconfigured data, or the first data can be data collected by the first communication apparatus itself, so as to reduce transmission overhead.

[0367] In a possible implementation, the first communication apparatus determines the label data corresponding to the first input data based on the M output data, including: the first communication apparatus determines fifth information based on the M output data, the fifth information being used to indicate at least one of scores, rankings, optimal values (or better values) of the M output data; and the fifth information is used to determine the label data corresponding to the first input data. Thus, the first communication apparatus can determine the label data through the fifth information, and the fifth information can be implemented in various ways to improve the flexibility of the scheme implementation.

[0368] Referring to FIG. 5, another implementation schematic diagram of the communication method provided by the present application is shown, and the method includes the following steps.

[0369] S501. The first communication apparatus obtains M output data. The M output data is obtained by processing the first input data based on N models, N is a positive integer, and M is an integer greater than or equal to N.

[0370] S502. The first communication device determines seventh information based on the M output data. The seventh information is used to determine second input data and / or model parameters of one or more models used to process the second input data to obtain a target output corresponding to the second input data.

[0371] For example, the sampling parameter can include at least one of the following:

[0372] Temperature coefficient: a coefficient for controlling probability normalization, e.g., higher temperature makes the output more random, and lower temperature makes the output more deterministic;

[0373] Random coefficient: sampling with a certain probability;

[0374] Top-K: sampling from the K outputs with the highest probability;

[0375] Top-P: sampling from the output set with a cumulative probability greater than or equal to p.

[0376] Based on the scheme shown in FIG. 5, the M output data obtained by the first communication device in step S501 is obtained by processing the first input data through the N models. Thereafter, the first communication device can determine the seventh information based on the M output data in step S502. The seventh information is used to determine the second input data and / or the model parameters of one or more models used to process the second input data to obtain a target output corresponding to the second input data. In other words, the second input data and / or the model parameters determined by the seventh information can be used to determine the personalized target output corresponding to the second input data. Thus, the seventh information determined by the first communication device based on the M output data can be used to optimize the model parameters and / or the input data of the model, so that the model can obtain the personalized target output, thereby improving the user experience.

[0377] In addition, in the case where the M output data is multiple output data (e.g., M is greater than 1), the above-mentioned seventh information can be obtained by processing the corresponding diversified output data through the N models, which can improve the performance of the seventh information and further improve the performance of the model parameters and / or the input data of the model based on the seventh information.

[0378] In a possible implementation, the method further includes: determining, by the first communication apparatus, the second input data and / or the sampling parameters of the one or more models for processing the second input data based on the seventh information. Thus, after the first communication apparatus determines the seventh information based on the M output data, the first communication apparatus can further determine the second input data and / or the sampling parameters of the one or more models for processing the second input data based on the seventh information, so that the first communication apparatus can locally (or through an external device of the first communication apparatus, such as a terminal device with AI function, a server, etc.) implement the determination of the second input data and / or the sampling parameters, to reduce transmission overhead.

[0379] In a possible implementation, the method further includes: sending, by the first communication apparatus, the seventh information. Thus, the first communication apparatus can send the seventh information, so that a receiver (for example, the second communication apparatus) of the seventh information can obtain the seventh information and implement the determination of the second input data and / or the sampling parameters based on the seventh information, to reduce implementation complexity of the first communication apparatus.

[0380] In a possible implementation, the method further includes: receiving or sending, by the first communication apparatus, eighth information used to indicate evaluation formula information, the evaluation formula information and the M output data being used to determine the seventh information. Thus, the first communication apparatus can further receive or send the eighth information used to indicate the evaluation formula information, so that a receiver of the eighth information can implement the determination of the seventh information.

[0381] It should be understood that the evaluation formula indicated by the above-mentioned evaluation formula information can be used to score the inference result, for example, the evaluation formula represents a matching degree between the inference result and a local individualized preference / indicator.

[0382] Optionally, the above-mentioned evaluation formula information can be replaced by other implementations, for example, an evaluation algorithm, an evaluation module, or an evaluation key performance indicator (KPI), etc.

[0383] For example, the seventh information is used to determine the second input data and / or the sampling parameters of the one or more models for processing the second input data, which can be understood as that the seventh information can determine the second input data and / or the sampling parameters of the one or more models for processing the second input data through at least one of an evaluation formula, an evaluation algorithm, an evaluation model, an evaluation module, or an evaluation KPI.

[0384] In a possible implementation, in step S501, the first communication apparatus can obtain the M output data in multiple ways, which will be introduced below in combination with some examples.

[0385] In the third implementation example, in step S501, the first communication apparatus obtains the M output data, including: the first communication apparatus receives the M output data. Thus, the first communication apparatus can obtain the M output data by receiving the M output data, i.e., the N models used for processing the first input data can be deployed on other communication apparatuses, so that the first communication apparatus does not need to deploy the N models (e.g., N pre-trained large models), thereby saving the storage space of the first communication apparatus and reducing the processing complexity of the first communication apparatus.

[0386] In a possible implementation of the third implementation example, the method further includes: the first communication apparatus sends the first input data. Thus, the first communication apparatus can further send the first input data to one or more communication apparatuses on which the N models are deployed, so that the one or more communication apparatuses can process the first input data to obtain and send the M output data.

[0387] Optionally, the first input data can be preconfigured data, so that the first communication apparatus does not need to send the first input data, thereby reducing the overhead.

[0388] In a possible implementation of the third implementation example, the method further includes: the first communication apparatus sends ninth information, the ninth information indicating a request for the M output data. Thus, the first communication apparatus can further send the ninth information, so that a receiver (e.g., a second communication apparatus) of the ninth information can provide the M output data for the first communication apparatus based on the request of the ninth information.

[0389] Optionally, the ninth information indicates at least one of the following: a task identifier, model information of part or all of the N models, a processing mode, a generation manner of the first input data, or task auxiliary information.

[0390] As an example, the ninth information can be used to request the M output data, the M output data can be used to determine the seventh information, and the first communication apparatus (or a receiver of the seventh information) can process the seventh information to obtain the second input data and / or the sampling parameters of one or more models used for processing the second input data. Different processing modes correspond to different processing of the seventh information, and different processing results can be obtained. Accordingly, in the case where the ninth information indicates the processing mode, the receiver of the ninth information can provide corresponding M output data based on the processing mode, so that the first communication apparatus obtains the M output data corresponding to the specified processing mode and obtains the seventh information of the specified processing mode.

[0391] Optionally, in the case where the processing of the seventh information is the processing of the evaluation formula, different processing modes can be understood as different processing of the evaluation formula.

[0392] Optionally, in the case that the processing of the seventh information is the processing of the evaluation model, different processing modes can be understood as different processing of the evaluation model.

[0393] As an example, the generation manner of the first input data can indicate that the first input data is generated in an offline manner or an online manner.

[0394] As an example, the task auxiliary information can include information required for model management. For example, in the case that the model is used for a communication task, the task auxiliary information can include configuration of communication parameters (such as configuration of resource parameters, configuration of power control parameters, etc.) of the communication device. For another example, in the case that the model is used for a classification task, the task auxiliary information can include the approximate category of the input data. For another example, in the case that the model is used for a generation task, the task auxiliary information can include a prompt word.

[0395] In a possible implementation of the example three, the method further includes: receiving, by the first communication device, tenth information, the tenth information being used to indicate the configuration information corresponding to the request. In other words, the receiver (for example, the second communication device) of the ninth information can determine the configuration information based on the ninth information after receiving the ninth information, and indicate the configuration information to the first communication device through the tenth information, so that the first communication device can determine the configuration information corresponding to the processing performed by the second communication device on the request based on the configuration information.

[0396] Optionally, the configuration information indicated by the tenth information can be the same as the content (such as the model information of part or all of the N models, or the task auxiliary information, etc.) requested by the ninth information, or can be partially different or completely different from the content requested by the ninth information, which is not limited here.

[0397] Optionally, in the case that the configuration information indicated by the tenth information is partially different or completely different from the content requested by the ninth information, the first communication device can indicate rejection (for example, rejection of the second communication device to process the first input data based on the configuration information) to the second communication device, so that the second communication device does not need to process based on the configuration information not expected by the first communication device, to reduce processing and transmission overhead. And / or, the first communication device can determine / generate / acquire the next request (such as the next ninth information sent) based on the configuration information, to improve processing efficiency.

[0398] Optionally, the configuration information indicates at least one of: a multicast identifier corresponding to the first communication device, part or all of the N models, or a processing mode; wherein the M output data is carried in multicast information, and the multicast information includes the multicast identifier. Thus, the configuration information indicated by the tenth information can include the at least one described above, to improve the flexibility of the scheme implementation.

[0399] For example, if the configuration information indicates the multicast identifier corresponding to the first communication device, the first communication device can obtain M output data through the received multicast information. If the number of first communication devices is greater than 1, the transmission overhead of the output data can be reduced through multicast transmission.

[0400] For example, if the above configuration information indicates some or all of the N models, the first communication device can provide first input data that matches the input of the relevant part or all models through the indication.

[0401] For example, if the configuration information indicates a processing mode for processing the first input data, the first communication device can obtain seventh information that matches the processing mode based on the indication of the configuration information.

[0402] Example 4: In step S501, the process of the first communication device acquiring M output data includes: the first communication device processing the first input data based on the N models to obtain the M output data. Thus, the first communication device can process the first input data using N models deployed locally (or via external devices of the first communication device, such as AI-enabled terminal devices, servers, etc.) to obtain M output data, thereby reducing transmission overhead.

[0403] In one possible implementation of Example 4, the method further includes: the first communication device receiving the first input data. Thus, the first communication device can receive the first input data, enabling it to obtain M output data through processing the first input data.

[0404] Optionally, the first input data can be pre-configured data, or the first data can be data collected by the first communication device itself, in order to reduce transmission overhead.

[0405] Please refer to Figure 6. This application embodiment provides a communication device 600, which can realize the functions of the second communication device or the first communication device in the above method embodiments, and thus can also achieve the beneficial effects of the above method embodiments. In this application embodiment, the communication device 600 can be the first communication device (or the second communication device), or it can be an integrated circuit or component inside the first communication device (or the second communication device), such as a chip.

[0406] It should be noted that the transceiver unit 602 may include a transmitting unit and a receiving unit, which are used to perform transmitting and receiving respectively.

[0407] In a possible implementation, when the apparatus 600 is configured to perform the method performed by the first communication device in the foregoing embodiments, the apparatus 600 includes a processing unit 601; the processing unit 601 is configured to obtain M output data, the M output data being obtained by processing first input data by N models, N being a positive integer, and M being an integer greater than or equal to N; and the processing unit 601 is further configured to determine label data corresponding to the first input data based on the M output data, where the label data is used to determine a first model, and the first model is used to determine target output corresponding to second input data from one or more output data corresponding to the second input data.

[0408] Optionally, the processing unit 601 is configured to obtain the M output data by receiving, by the transceiver unit 602, the M output data.

[0409] In a possible implementation, when the apparatus 600 is configured to perform the method performed by the second communication device in the foregoing embodiments, the apparatus 600 includes a processing unit 601 and a transceiver unit 602; the transceiver unit 602 is configured to receive first information, the first information being used to indicate label data corresponding to first input data; where the label data is determined based on M output data obtained by processing the first input data by N models, N being a positive integer, and M being an integer greater than or equal to N; and the processing unit 601 is configured to determine a first model based on the label data; where the first model is used to determine output data corresponding to second input data from one or more output data corresponding to the second input data.

[0410] In a possible implementation, when the apparatus 600 is configured to perform the method performed by the first communication device in the foregoing embodiments, the apparatus 600 includes a processing unit 601; the processing unit 601 is configured to obtain M output data, the M output data being obtained by processing first input data by N models, N being a positive integer, and M being an integer greater than or equal to N; and the processing unit 601 is further configured to determine seventh information based on the M output data, the seventh information being used to determine model parameters of second input data and / or one or more models used to process the second input data, the second input data being used to be processed by the one or more models to obtain target output corresponding to the second input data.

[0411] Optionally, the processing unit 601 is configured to obtain the M output data by receiving, by the transceiver unit 602, the M output data.

[0412] In a possible implementation, when the apparatus 600 is configured to perform the method performed by the second communication apparatus in the foregoing embodiments, the apparatus 600 includes a processing unit 601 and a transceiver 602; the transceiver 602 is configured to receive seventh information, the seventh information being obtained based on M output data obtained by processing first input data by N models, N being a positive integer, and M being an integer greater than or equal to N; and the processing unit 601 is configured to determine second input data and / or model parameters of one or more models used for processing the second input data based on the seventh information, the second input data being used for processing by the one or more models to obtain a target output corresponding to the second input data.

[0413] It should be noted that the information execution process and the like of the units of the communication apparatus 600 are described in the foregoing method embodiments of the present application, and will not be described here.

[0414] Referring to FIG. 7, another schematic structural diagram of a communication apparatus 700 provided by the present application is shown, which includes a logic circuit 701 and an input-output interface 702. The communication apparatus 700 can be a chip or an integrated circuit.

[0415] The transceiver 602 shown in FIG. 6 can be a communication interface, which can be the input-output interface 702 shown in FIG. 7. The input-output interface 702 can include an input interface and an output interface. Alternatively, the communication interface can be a transceiver circuit, which can include an input interface circuit and an output interface circuit.

[0416] Optionally, the logic circuit 701 is configured to obtain M output data, the M output data being obtained by processing first input data by N models, N being a positive integer, and M being an integer greater than or equal to N; and the logic circuit 701 is further configured to determine label data corresponding to the first input data based on the M output data, wherein the label data is used to determine a first model, and the first model is used to determine a target output corresponding to second input data from one or more output data corresponding to the second input data.

[0417] Optionally, the input-output interface 702 is configured to receive first information, the first information being used to indicate label data corresponding to first input data; wherein the label data is determined based on M output data obtained by processing the first input data by N models, N being a positive integer, and M being an integer greater than or equal to N; and the logic circuit 701 determines a first model based on the label data; wherein the first model is used to determine output data corresponding to second input data from one or more output data corresponding to the second input data.

[0418] Optionally, the logic circuit 701 is configured to obtain M output data, the M output data being obtained by processing the first input data by N models, N being a positive integer, and M being an integer greater than or equal to N; and the logic circuit 701 is further configured to determine seventh information based on the M output data, the seventh information being used to determine the second input data and / or model parameters of one or more models used to process the second input data, the second input data being used to be processed by the one or more models to obtain a target output corresponding to the second input data.

[0419] Optionally, the input and output interface 702 is configured to receive the seventh information, the seventh information being obtained based on M output data obtained by processing the first input data by N models, N being a positive integer, and M being an integer greater than or equal to N; and the logic circuit 701 is configured to determine the second input data and / or model parameters of one or more models used to process the second input data based on the seventh information, the second input data being used to be processed by the one or more models to obtain a target output corresponding to the second input data.

[0420] The logic circuit 701 and the input and output interface 702 can also perform other steps performed by the first communication device or the second communication device in any of the embodiments and achieve corresponding beneficial effects, which will not be described here.

[0421] In a possible implementation, the processing unit 601 shown in FIG. 6 can be the logic circuit 701 in FIG. 7.

[0422] Optionally, the logic circuit 701 can be a processing device, and the functions of the processing device can be partially or entirely implemented by software.

[0423] Optionally, the processing device can include a memory and a processor, where the memory is configured to store a computer program, and the processor is configured to read and execute the computer program stored in the memory to perform the corresponding processing and / or steps in any one of the method embodiments.

[0424] Optionally, the processing device can only include the processor. The memory for storing the computer program is located outside the processing device, and the processor is connected with the memory through a circuit / wire to read and execute the computer program stored in the memory. The memory and the processor can be integrated together or can be physically independent of each other.

[0425] Optionally, the processing device can be one or more chips, or one or more integrated circuits. For example, the processing device can be one or more field-programmable gate arrays (FPGA), application specific integrated circuits (ASIC), system on chips (SoC), central processor units (CPU), network processors (NP), digital signal processors (DSP), micro controller units (MCU), programmable logic devices (PLD), or other integrated circuits, or any combination of the above chips or processors, etc.

[0426] Referring to FIG. 8, a communication device 800 involved in the above embodiments provided by the embodiments of the present application is shown, which can be the communication device as the terminal device in the above embodiments, and the communication device in the example shown in FIG. 8 is implemented by the terminal device (or components in the terminal device).

[0427] Optionally, the communication device 800 can include but is not limited to at least one processor 801 and a communication port 802.

[0428] Optionally, the transceiver unit 602 shown in FIG. 6 can be a communication interface, which can be the communication port 802 in FIG. 8, and the communication port 802 can include an input interface and an output interface. Alternatively, the communication port 802 can also be a transceiver circuit, which can include an input interface circuit and an output interface circuit.

[0429] Further optionally, the device can further include at least one of a memory 803 and a bus 804, and in the embodiments of the present application, the at least one processor 801 is configured to control and process the actions of the communication device 800.

[0430] Further, the processor 801 can be a central processing unit, a general purpose processor, a digital signal processor, an application specific integrated circuit, a field programmable gate array or other programmable logic device, transistor logic, hardware components, or any combination thereof. It can implement or execute various example logical blocks, modules, and circuits described in connection with the disclosure. The processor can also be a combination of computing functions, such as a combination of one or more microprocessors, a combination of a digital signal processor and a microprocessor, and the like. For the convenience and brevity of description, the specific working processes of the above-described system, device, and unit can be referred to the corresponding processes in the foregoing method embodiments, which will not be described herein.

[0431] It should be noted that the communication device 800 shown in FIG. 8 can be specifically used to implement the steps implemented by the terminal device in the foregoing method embodiments, and achieve the corresponding technical effects of the terminal device. The specific implementation of the communication device shown in FIG. 8 can be referred to the description in the foregoing method embodiments, which will not be described herein.

[0432] Please refer to FIG. 9, which is a structural schematic diagram of a communication device 900 involved in the foregoing embodiments provided by the embodiments of the present application. The communication device 900 can be specifically the communication device as the network device in the foregoing embodiments. The communication device in the example shown in FIG. 9 is implemented by a network device (or a component in the network device), and the structure of the communication device can be referred to the structure shown in FIG. 9.

[0433] The communication device 900 includes at least one processor 911 and at least one network interface 914. Further optionally, the communication device further includes at least one memory 912, at least one transceiver 913, and one or more antennas 915. The processor 911, the memory 912, the transceiver 913, and the network interface 914 are connected, for example, through a bus. In the embodiments of the present application, the connection can include various interfaces, transmission lines, or buses, etc., which are not limited in the embodiments of the present application. The antenna 915 is connected to the transceiver 913. The network interface 914 is used to enable the communication device to communicate with other communication devices through a communication link. For example, the network interface 914 can include a network interface between the communication device and a core network device, such as an S1 interface. The network interface can include a network interface between the communication device and other communication devices (such as other network devices or core network devices), such as an X2 or Xn interface.

[0434] The transceiver unit 602 shown in FIG. 6 can be a communication interface, which can be the network interface 914 in FIG. 9. The network interface 914 can include an input interface and an output interface. Alternatively, the network interface 914 can be a transceiver circuit, which can include an input interface circuit and an output interface circuit.

[0435] The processor 911 is mainly used for processing communication protocols and communication data, and controlling the whole communication device, executing software programs, processing data of the software programs, for example, for supporting the communication device to perform the actions described in the embodiments. The communication device can include a baseband processor mainly used for processing communication protocols and communication data, and a central processor mainly used for controlling the whole terminal device, executing software programs, and processing data of the software programs. The processor 911 in FIG. 9 can integrate the functions of the baseband processor and the central processor, and those skilled in the art can understand that the baseband processor and the central processor can also be independent processors interconnected by a bus or the like. Those skilled in the art can understand that the terminal device can include multiple baseband processors to adapt to different network modes, and the terminal device can include multiple central processors to enhance its processing capability, and various components of the terminal device can be connected by various buses. The baseband processor can also be referred to as a baseband processing circuit or a baseband processing chip. The central processor can also be referred to as a central processing circuit or a central processing chip. The function of processing communication protocols and communication data can be built into the processor, or stored in the memory in the form of a software program, and the processor executes the software program to realize the baseband processing function.

[0436] The memory is mainly used for storing software programs and data. The memory 912 can exist independently and be connected with the processor 911. Alternatively, the memory 912 can be integrated with the processor 911, for example, integrated in a chip. The memory 912 can store program codes for executing the technical solutions of the embodiments of the present application, and the processor 911 controls the execution. Various computer programs executed can also be regarded as a driver of the processor 911.

[0437] FIG. 9 only shows one memory and one processor. In actual terminal devices, there can be multiple processors and multiple memories. The memory can also be referred to as a storage medium or a storage device, etc. The memory can be a storage element on the same chip as the processor, that is, an on-chip storage element, or an independent storage element, and the embodiments of the present application do not limit this.

[0438] The transceiver 913 can be configured to support the receiving or transmitting of radio frequency signals between the communication device and a terminal. The transceiver 913 can be connected with the antenna 915. The transceiver 913 includes a transmitter Tx and a receiver Rx. Specifically, the one or more antennas 915 can receive radio frequency signals, the receiver Rx of the transceiver 913 is configured to receive the radio frequency signals from the antenna and convert the radio frequency signals into digital baseband signals or digital intermediate frequency signals, and provide the digital baseband signals or the digital intermediate frequency signals to the processor 911 for further processing, such as demodulation processing and decoding processing, by the processor 911. In addition, the transmitter Tx in the transceiver 913 is also configured to receive modulated digital baseband signals or digital intermediate frequency signals from the processor 911, and convert the modulated digital baseband signals or digital intermediate frequency signals into radio frequency signals, and transmit the radio frequency signals through the one or more antennas 915. Specifically, the receiver Rx can selectively perform one or more levels of down-mixing processing and analog-to-digital conversion processing on the radio frequency signals to obtain the digital baseband signals or the digital intermediate frequency signals, and the order of the down-mixing processing and the analog-to-digital conversion processing can be adjustable. The transmitter Tx can selectively perform one or more levels of up-mixing processing and digital-to-analog conversion processing on the modulated digital baseband signals or the digital intermediate frequency signals to obtain the radio frequency signals, and the order of the up-mixing processing and the digital-to-analog conversion processing can be adjustable. The digital baseband signals and the digital intermediate frequency signals can be collectively referred to as digital signals.

[0439] The transceiver 913 can also be referred to as a transceiving unit, a transceiver, a transceiving device, etc. Optionally, the devices in the transceiving unit for implementing the receiving function can be regarded as a receiving unit, and the devices in the transceiving unit for implementing the transmitting function can be regarded as a transmitting unit, i.e., the transceiving unit includes the receiving unit and the transmitting unit, the receiving unit can also be referred to as a receiver, an input port, a receiving circuit, etc., and the transmitting unit can be referred to as a transmitter, a transmitter, or a transmitting circuit, etc.

[0440] It should be noted that the communication device 900 shown in FIG. 9 can be specifically configured to implement the steps implemented by the network device in the foregoing method embodiments, and achieve the corresponding technical effects of the network device. The specific implementation mode of the communication device 900 shown in FIG. 9 can be referred to the description in the foregoing method embodiments, which will not be described here one by one.

[0441] Please refer to FIG. 10, which is a structural schematic diagram of a communication device involved in the above embodiments provided by the embodiments of the present application.

[0442] It can be understood that the communication apparatus 10 comprises, for example, modules, units, elements, circuits, or interfaces, and the like, which are properly configured together to perform the technical solutions provided in the present application. The communication apparatus 10 can be a terminal device or a network device as described above, or can be a component (for example, a chip) of the devices, to implement the methods described in the following method embodiments. The communication apparatus 10 comprises one or more processors 101. The processor 101 can be a general processor or a special-purpose processor, and the like. For example, it can be a baseband processor or a central processing unit. The baseband processor can be used to process communication protocols and communication data, and the central processing unit can be used to control the communication apparatus (such as a RAN node, a terminal, or a chip, and the like), execute software programs, and process data of the software programs.

[0443] Optionally, in one design, the processor 101 can include a program 103 (which can also be referred to as code or instructions at times) that can be run on the processor 101, so that the communication apparatus 10 performs the methods described in the following embodiments. In yet another possible design, the communication apparatus 10 comprises a circuit (not shown in FIG. 10).

[0444] Optionally, the communication apparatus 10 can comprise one or more memories 102, which have a program 104 (which can also be referred to as code or instructions at times) stored thereon, and the program 104 can be run on the processor 101, so that the communication apparatus 10 performs the methods described in the above method embodiments.

[0445] Optionally, the processor 101 and / or the memory 102 can comprise an AI module 107, 108, which is used to implement AI-related functions. The AI module can be implemented in software, hardware, or a combination of software and hardware. For example, the AI module can comprise a radio intelligence control (RIC) module. For example, the AI module can be a near-real-time RIC or a non-real-time RIC.

[0446] Optionally, the processor 101 and / or the memory 102 can also store data. The processor and the memory can be separately arranged, or can be integrated together.

[0447] Optionally, the communication apparatus 10 can also comprise a transceiver 105 and / or an antenna 106. The processor 101 can also be referred to as a processing unit, which controls the communication apparatus (such as a RAN node or a terminal). The transceiver 105 can also be referred to as a transceiving unit, a transceiver, a transceiving circuit, or a transceiver, and the like, which is used to realize the transceiving function of the communication apparatus through the antenna 106.

[0448] The processing unit 601 shown in FIG. 6 can be the processor 101. The transceiving unit 602 shown in FIG. 6 can be a communication interface, which can be the transceiver 105 in FIG. 10. The transceiver 105 can include an input interface and an output interface. Alternatively, the transceiver 105 can be a transceiving circuit, which can include an input interface circuit and an output interface circuit.

[0449] The embodiments of the present application further provide a computer readable storage medium for storing one or more computer-executable instructions, which, when executed by a processor, cause the processor to perform the method described in the possible implementation manners of the first communication device or the second communication device.

[0450] The embodiments of the present application further provide a computer program product (or computer program), which, when executed by a processor, causes the processor to perform the method described in the possible implementation manners of the first communication device or the second communication device.

[0451] The embodiments of the present application further provide a chip system, which includes at least one processor for supporting the communication device to implement the functions involved in the possible implementation manners of the communication device. Optionally, the chip system further includes an interface circuit for providing program instructions and / or data for the at least one processor. In a possible design, the chip system can further include a memory for storing necessary program instructions and data of the communication device. The chip system can be composed of a chip, or can include a chip and other discrete components. The communication device can be the first communication device or the second communication device in the method embodiments.

[0452] The embodiments of the present application further provide a communication system, which includes the first communication device and the second communication device in any of the above embodiments.

[0453] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other manners. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There can be another division manner for the actual implementation. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between the units can be indirect couplings or communication connections through some interfaces, devices or units, and can be electrical, mechanical or in other forms.

[0454] The units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed on multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0455] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit. When the integrated unit is realized in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application essentially or say the part that contributes or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various storage program codes.

Claims

1. A communication method characterized by comprising: The method comprises: obtaining M output data, the M output data being obtained by processing first input data by N models, N being a positive integer, and M being an integer greater than or equal to N; determining label data corresponding to the first input data based on the M output data, wherein the label data is used to determine a first model, and the first model is used to determine target output data corresponding to second input data from one or more output data corresponding to the second input data.

2. The method of claim 1, wherein, The method further comprises: determining the first model based on the label data.

3. The method of claim 1, wherein, The method further comprises: sending first information, the first information being used to indicate the label data.

4. The method according to any one of claims 1 to 3, characterized in that, The method further comprises: receiving or sending second information, the second information being used to indicate model information of the first model.

5. The method according to any one of claims 1 to 4, characterized in that, The obtaining of the M output data comprises: receiving the M output data.

6. The method of claim 5, wherein, The method further comprises: sending the first input data.

7. The method according to claim 5 or 6, characterized in that, The method further comprises: sending third information, the third information indicating a request for the M output data.

8. The method of claim 7, wherein, The third information indicates at least one of the following: a task identifier, model information of part or all of the N models, a processing mode, a generation manner of the first input data, or task auxiliary information.

9. The method according to any one of claims 5 to 8, characterized in that, The method further comprises: receiving fourth information, the fourth information being used to indicate configuration information corresponding to the request.

10. The method of claim 9, wherein, The fourth information indicates at least one of the following: model identifiers of part or all of the N models, a number of samples of inference results, or a processing mode.

11. The method according to any one of claims 1 to 4, characterized in that, The obtaining of the M output data comprises: processing the first input data based on the N models to obtain the M output data.

12. The method according to any one of claims 1 to 11, characterized in that, The determination of the label data corresponding to the first input data based on the M output data comprises: determining fifth information based on the M output data, the fifth information being used to indicate at least one of scores, rankings, or optimal values of the M output data; wherein the fifth information is used to determine the label data corresponding to the first input data.

13. A method of communication, comprising: The method comprises: receiving first information, the first information being used to indicate label data corresponding to first input data; wherein the label data is determined based on M output data obtained by processing the first input data by N models, N being a positive integer, and M being an integer greater than or equal to N; determining a first model based on the label data; wherein the first model is used to determine output data corresponding to second input data from one or more output data corresponding to the second input data.

14. The method of claim 13, wherein, The method further comprises: sending second information, the second information being used to indicate model information of the first model.

15. The method according to claim 13 or 14, characterized in that, The method further comprises: sending the M output data.

16. The method of claim 15, wherein, The method further comprises: receiving the first input data.

17. The method according to claim 15 or 16, characterized in that The method further comprises: receiving third information, the third information indicating a request for the M output data.

18. The method of claim 17, wherein, The third information indicates at least one of the following: a task identifier, model information of part or all of the N models, a processing mode, a generation manner of the first input data, or task auxiliary information.

19. The method according to any one of claims 15 to 18, characterized in that, The method further comprises: The fourth information is used for indicating configuration information corresponding to the request.

20. The method of claim 19, wherein, The fourth information indicates at least one of the following: a groupcast identifier corresponding to the first communication device, part or all of the N models, or a processing mode.

21. The method according to any one of claims 13 to 20, characterized in that, The label data is determined based on M output data obtained by processing first input data by the N models, and includes: The label data is determined by fifth information, and the fifth information is determined by the M data; wherein the fifth information is used for indicating at least one of the following: scores, rankings, and optimal values of the M output data.

22. The method according to any one of claims 1 to 21, characterized in that, The first model is used for determining a target output corresponding to the second input data from one or more output data corresponding to the second input data, including: The first model is used for determining sixth information from one or more output data corresponding to the second input data, and the sixth information is used for determining a target output corresponding to the second input data; wherein the sixth information includes at least one of the following: first indication information indicating a ranking of the one or more output data; wherein the target output is one of the one or more output data; second indication information indicating a first index, the first index being used for determining the target output in the one or more output data; wherein the target output is one of the one or more output data; third indication information indicating the target output; wherein the target output is determined based on the one or more output data.

23. A communications device, characterized by A module for performing the method of any one of claims 1 to 22.

24. A communications device, characterized by At least one processor is used to execute programs or instructions stored in the memory, so that the communication device implements the method of any one of claims 1 to 22.

25. The communication apparatus according to claim 24, wherein, The communication device further includes a processor.

26. The communication apparatus according to claim 24, wherein The communication device is a chip or a chip system.

27. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer programs or instructions, which, when executed by the communication device, implement the method of any one of claims 1 to 22.

28. A computer program product, characterised in that, The computer program or instruction is executed by the computer, and the method of any one of claims 1 to 22 is implemented. The computer program or instruction is executed by the computer, and the method of any one of claims 1 to 22 is implemented.

Images