Data transmission method and related apparatus

Abstract

Provided in the present application are a data transmission method and a related apparatus. The method provided in the present application comprises: a first apparatus generating at least one first simulated dataset on the basis of first data, wherein the first data is local data of the first apparatus, each of the at least one first simulated dataset has a corresponding weight, and the weight of each first simulated dataset is used for indicating the degree of importance of the first simulated dataset; and the first apparatus transmitting the first data, the at least one first simulated dataset and first indication information to a second apparatus, wherein the first indication information is used for indicating the weight of each of the at least one first simulated dataset. It can be seen therefrom that a first apparatus provides more data to a second apparatus, such that the second apparatus performs model training on the basis of first data and at least one first simulated dataset. Thus, the amount of data for model training is increased, thereby facilitating an improvement in the model performance and / or generalization performance of trained models.

Description

Data transmission method and related apparatus

[0001] The present application claims priority to the Chinese Patent Application No. 202410837539.6, filed on June 25, 2024, and entitled “Data transmission 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 technology, and in particular to a data transmission method and related apparatus. BACKGROUND

[0003] Currently, artificial intelligence (AI) technology is introduced into a wireless communication system. The AI technology can be used for compression and reconstruction of wireless channel information, beam management, and positioning enhancement, etc. An AI model is obtained based on data training to improve the performance of completing a wireless task through the AI model. As shown in FIG. 1, a wireless AI framework includes data collection, model training, model management, model inference, and model storage. A device can train a model based on collected data to improve the performance of the model.

[0004] However, the number of real data is limited (for example, the number of real data is limited due to time limit and / or collection cost limit), resulting in poor model performance and / or generalization performance of the trained model. SUMMARY

[0005] The present application provides a data transmission method and related apparatus, which are used for a first device to generate at least one first simulation data set according to first data, and to send the first data, the at least one first simulation data set, and first indication information to a second device. As can be seen, the first device provides more data to the second device, that is, the first device provides not only the first data but also the at least one first simulation data set. This facilitates the second device to perform model training based on the first data and the at least one first simulation data set. As can be seen, the amount of data used for model training is increased, which is beneficial to improve the model performance and / or generalization performance of the trained model. Further, the first indication information is used to indicate the weight of each first simulation data set in the at least one first simulation data set, and the weight of each first simulation data set is used to indicate the importance of the first simulation data set. This is beneficial to the second device to select appropriate data based on the weight of each first simulation data set to perform corresponding model training, thereby meeting the needs of different models.

[0006] The first aspect of the present application provides a data transmission method, which can be executed by a first device. The first device can be a terminal device, a network device (such as an access network device or a core network device), or a component (for example, a processor, a chip, or a chip system) in the terminal device or the network device, or a logic module or software capable of realizing all or part of the functions of the terminal device or the network device. The method comprises the following steps: the first device generates at least one first simulation data set according to first data, the first data being local data of the first device, each first simulation data set in the at least one first simulation data set having a corresponding weight, and the weight of each first simulation data set being used to indicate the importance of each first simulation data set; and the first device sends the first data, the at least one first simulation data set, and first indication information to a second device, the first indication information being used to indicate the weight of each first simulation data set in the at least one first simulation data set.

[0007] According to the above technical solution, the first device generates at least one first simulation data set according to the first data and sends the first data, the at least one first simulation data set, and first indication information to the second device. It can be known that the first device provides more data to the second device, that is, the first device provides not only the first data but also the at least one first simulation data set. This facilitates the second device to perform model training based on the first data and the at least one first simulation data set. Therefore, the amount of data used for model training is increased, which is beneficial to improving the model performance and / or the generalization performance of the model obtained through training. Furthermore, the first indication information is used to indicate the weight of each first simulation data set in the at least one first simulation data set, and the weight of each first simulation data set is used to indicate the importance of the first simulation data set. This is beneficial to the second device to select appropriate data for corresponding model training based on the weight of each first simulation data set, so as to meet the requirements of different models. For example, for a model with high model performance requirements, the second device can preferentially select a first simulation data set with a higher weight for model training. For a model with high generalization performance requirements, the second device can preferentially select a first simulation data set with a lower weight for model training.

[0008] Based on the first aspect, in a possible implementation manner, the method further comprises the following step: the first device sends the weight of the first data to the second device, the weight of the first data being used to indicate the importance of the first data. Therefore, the second device can obtain the weight of the first data and consider whether to select the first data for model training in combination with the weight. Especially for the case where the first device provides multiple local data, the second device can select appropriate data for model training in combination with the weights of the multiple local data. For example, for a model with high model performance requirements, the second device can select local data with a higher weight for model training. Therefore, the model performance is improved.

[0009] In a possible implementation of the first aspect, the data generation manner comprises: generating the first simulation data set according to the first data and the first model; or performing data augmentation on the first data to obtain the first simulation data set. This implementation provides two different data generation manners, enriching the implementation of the scheme. This facilitates the first device to generate more simulation data sets, so as to realize that the second device provides more data for model training.

[0010] In a possible implementation of the first aspect, the data generation manner comprises: generating the first simulation data set according to the first data and the first model; or performing data augmentation on the first data to obtain the first simulation data set. This implementation provides two different data generation manners, enriching the implementation of the scheme. This facilitates the first device to generate more simulation data sets, so as to realize that the second device provides more data for model training.

[0011] In a possible implementation of the first aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a similarity between each first simulation data set in the at least one first simulation data set and the first data. This shows a determination manner of the weight of the first simulation data set, which is beneficial to the implementation of the scheme. Further, the weight of the first simulation data set represents the importance of the first simulation data set. The similarity between the first simulation data set and the first data can better represent the importance of the first simulation data set. For example, the higher the similarity between the first simulation data set and the first data, the higher the importance of the first simulation data set, that is, the higher the weight of the first simulation data set.

[0012] In a possible implementation of the first aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a data distribution difference between each first simulation data set in the at least one first simulation data set and the first data. This shows another determination manner of the weight of the first simulation data set, which is beneficial to the implementation of the scheme. Further, the weight of the first simulation data set represents the importance of the first simulation data set. The data distribution difference between the first simulation data set and the first data can better represent the importance of the first simulation data set. For example, the higher the data difference between the first simulation data set and the first data, the lower the importance of the first simulation data set, that is, the lower the weight of the first simulation data set.

[0013] In a possible implementation of the first aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a signal-to-noise ratio difference between the first simulation data set and the first data. This is another way of determining the weight of the first simulation data set, which facilitates implementation of the solution. Further, the weight of the first simulation data set represents the importance of the first simulation data set. The signal-to-noise ratio difference between the first simulation data set and the first data can better represent the importance of the first simulation data set. For example, the higher the signal-to-noise ratio difference between the first simulation data set and the first data, the lower the importance of the first simulation data set, that is, the lower the weight of the first simulation data set.

[0014] In a possible implementation of the first aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a position difference between the first simulation data set and the first data. This is another way of determining the weight of the first simulation data set, which facilitates implementation of the solution. Further, the weight of the first simulation data set represents the importance of the first simulation data set. The position difference between the first simulation data set and the first data can better represent the importance of the first simulation data set. For example, the higher the position difference between the first simulation data set and the first data, the lower the importance of the first simulation data set, that is, the lower the weight of the first simulation data set.

[0015] In a possible implementation of the first aspect, the method further includes: receiving, by the first device, an expected weight of each first simulation data set in the at least one first simulation data set from the second device, the expected weight being used to indicate an expected importance of the first simulation data set; and generating, by the first device, the at least one first simulation data set according to the first data, including: generating, by the first device, the at least one first simulation data set according to the first data and the expected weight of each first simulation data set in the at least one first simulation data set. In this implementation, the expected weight of the first simulation data set affects the importance of the generated first simulation data set, which facilitates the first device to provide the second device with the first simulation data set that meets the corresponding importance.

[0016] In a possible implementation of the first aspect, the method further includes: receiving, by the first device, a first weight from the second device, the first weight being used to indicate an importance of data of the first device; and generating, by the first device, the at least one first simulation data set according to the first data, including: generating, by the first device, the at least one first simulation data set according to the first data and the first weight. In this implementation, the first weight affects the importance of the generated first simulation data set, which facilitates the first device to provide the second device with the first simulation data set that meets the corresponding importance.

[0017] In a possible implementation of the first aspect, the first indication information includes weights of the first simulation data sets in the at least one first simulation data set. The first device feeds back the weights of the first simulation data sets in the at least one first simulation data set to the second device.

[0018] In a possible implementation of the first aspect, the first indication information includes an identifier of each first simulation data set in the at least one first simulation data set, and the identifier of each first simulation data set is used to indicate the weight of the first simulation data set. In this way, the weight of each first simulation data set is indicated by the identifier of the first simulation data set. This is advantageous in reducing the indication overhead of the first device.

[0019] The second aspect of the present application provides a data transmission method, which can be performed by a second device. The second device can be a terminal device, a network device (for example, an access network device or a core network device), or a component (for example, a processor, a chip, or a chip system) in the terminal device or the network device, or a logic module or software capable of implementing all or part of the functions of the terminal device or the network device. The method includes the following steps. The second device receives first data, at least one first simulation data set, and first indication information from a first device. The first data is local data of the first device, the at least one first simulation data set is generated according to the first data, and the first indication information is used to indicate weights of the first simulation data sets in the at least one first simulation data set. The weight of each first simulation data set is used to indicate the importance of the first simulation data set. The second device performs model training according to the first data, the at least one first simulation data set, and the weights of the first simulation data sets in the at least one first simulation data set, and obtains one or more models.

[0020] According to the technical solution, the second device receives the first data, the at least one first simulation data set, and the first indication information from the first device. The second device performs model training according to the first data, the at least one first simulation data set, and the weight of each first simulation data set in the at least one first simulation data set, to obtain one or more models. In this way, the amount of data used for model training is increased, which is beneficial to improving the model performance and / or generalization performance of the trained model. Further, the first indication information is used to indicate the weight of each first simulation data set in the at least one first simulation data set, and the weight of each first simulation data set is used to indicate the importance of the first simulation data set. This is beneficial to the second device selecting appropriate data for corresponding model training based on the weight of each first simulation data set, so as to meet the requirements of different models. For example, for a model with high model performance requirements, the second device can preferentially select a first simulation data set with a higher weight for model training. For a model with high generalization performance requirements, the second device can preferentially select a first simulation data set with a lower weight for model training.

[0021] Based on the second aspect, in a possible implementation, the method further includes: the second device receiving a weight of the first data from the first device, the weight of the first data being used to indicate the importance of the first data; and the second device performing model training according to the first data, the at least one first simulation data set, and the weight of each first simulation data set in the at least one first simulation data set, to obtain one or more models, including: the second device performing model training according to the first data, the weight of the first data, the at least one first simulation data set, and the weight of each first simulation data set in the at least one first simulation data set, to obtain one or more models. In this way, the second device can obtain the weight of the first data, and consider whether to select the first data for model training in combination with the weight. Especially in the case where the first device provides multiple local data, the second device can select appropriate data for model training in combination with the weight of the multiple local data. For example, for a model with high model performance requirements, the second device can select local data with a higher weight for model training. In this way, the model performance is improved.

[0022] Based on the second aspect, in a possible implementation, the weight of each first simulation data set in the at least one first simulation data set is determined according to the similarity between each first simulation data set in the at least one first simulation data set and the first data. This shows a determination manner of the weight of the first simulation data set, which is beneficial to implementation of the scheme. Further, the weight of the first simulation data set represents the importance of the first simulation data set. The similarity between the first simulation data set and the first data can better represent the importance of the first simulation data set. For example, the higher the similarity between the first simulation data set and the first data, the higher the importance of the first simulation data set, that is, the higher the weight of the first simulation data set.

[0023] In a possible implementation of the second aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a data distribution difference between the each first simulation data set and the first data. Another way of determining the weight of the first simulation data set is shown, which facilitates implementation of the solution. Further, the weight of the first simulation data set represents the importance of the first simulation data set. The importance of the first simulation data set can be better represented by the data distribution difference between the first simulation data set and the first data. For example, the higher the data distribution difference between the first simulation data set and the first data, the lower the importance of the first simulation data set, i.e., the lower the weight of the first simulation data set.

[0024] In a possible implementation of the second aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a signal-to-noise ratio difference between the each first simulation data set and the first data. Another way of determining the weight of the first simulation data set is shown, which facilitates implementation of the solution. Further, the weight of the first simulation data set represents the importance of the first simulation data set. The importance of the first simulation data set can be better represented by the signal-to-noise ratio difference between the first simulation data set and the first data. For example, the higher the signal-to-noise ratio difference between the first simulation data set and the first data, the lower the importance of the first simulation data set, i.e., the lower the weight of the first simulation data set.

[0025] In a possible implementation of the second aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a position difference between the each first simulation data set and the first data. Another way of determining the weight of the first simulation data set is shown, which facilitates implementation of the solution. Further, the weight of the first simulation data set represents the importance of the first simulation data set. The importance of the first simulation data set can be better represented by the position difference between the first simulation data set and the first data. For example, the higher the position difference between the first simulation data set and the first data, the lower the importance of the first simulation data set, i.e., the lower the weight of the first simulation data set.

[0026] In a possible implementation of the second aspect, the method further includes: sending, by the second device, an expected weight of each first simulation data set in the at least one first simulation data set to the first device, the expected weight being used to indicate an expected importance of the first simulation data set. In this implementation, the expected weight of the first simulation data set affects the importance of the generated first simulation data set, which is conducive to the second device obtaining the first simulation data set meeting the corresponding importance from the first device.

[0027] In a possible implementation manner of the second aspect, the one or more models comprise a second model; and the second device performs model training according to the first data, the at least one first simulation data set, and the weight of each first simulation data set in the at least one first simulation data set, to obtain the one or more models, including: the second device selects target data from the first data and the at least one first simulation data set according to at least one of a model performance requirement and a generalization performance requirement of the second model and the weight of each first simulation data set in the at least one first simulation data set; and the second device performs model training by using the target data to obtain the second model. In this way, the second device selects appropriate data to perform model training, to obtain the second model, so that the second model obtained by training meets the corresponding performance requirement.

[0028] In a possible implementation manner of the second aspect, the method further includes: the second device receives second data, at least one second simulation data set, and second indication information from a third device, the second data is local data of the third device, the at least one second simulation data set is generated according to the second data, and the second indication information is used to indicate the weight of each second simulation data set in the at least one second simulation data set, the weight of each second simulation data set in the at least one second simulation data set is used to indicate the importance of the second simulation data set; and the second device performs model training according to the first data, the at least one first simulation data set, and the weight of each first simulation data set in the at least one first simulation data set, to obtain the one or more models, including: the second device performs model training according to the first data, the at least one first simulation data set, the second data, the at least one second simulation data set, the weight of each first simulation data set in the at least one first simulation data set, and the weight of each second simulation data set in the at least one second simulation data set, to obtain the one or more models. In this implementation manner, the second device can obtain local data, simulation data sets, and weights of simulation data sets provided by multiple devices. Then, the second device performs model training in combination with the local data, the simulation data sets, and the weights of the simulation data sets provided by the multiple devices. That is, the second device performs model training by fusing data of multiple manufacturers or multiple users. This is beneficial to improving the model performance and / or the generalization performance of the model obtained by training.

[0029] In a possible implementation manner of the second aspect, the second device performs model training according to the first data, the at least one first simulation data set, the second data, the at least one second simulation data set, the weight of each first simulation data set in the at least one first simulation data set, and the weight of each second simulation data set in the at least one second simulation data set, to obtain one or more models, including: the second device determines a target data set according to a first weight, the first data, the at least one first simulation data set, the weight of each first simulation data set in the at least one first simulation data set, a second weight, the second data, the at least one second simulation data set, and the weight of each second simulation data set in the at least one second simulation data set, the first weight being used to indicate the importance of data of the first device, and the second weight being used to indicate the importance of data of the third device; and the second device performs model training by using the target data set, to obtain one or more models. In this implementation manner, when the second device performs model training by using data provided by multiple devices, the second device further selects data according to the importance of data of different devices, and then performs model training. This is beneficial to the model trained by the second device to meet corresponding performance requirements. For example, for a model with high performance requirements, the second device can select data of a device with high importance to perform model training.

[0030] In a possible implementation manner of the second aspect, the method further includes: the second device sends the first weight to the first device. In this implementation manner, the first weight affects the importance of the generated first simulation data set, which is beneficial to the second device to obtain the first simulation data set meeting corresponding importance from the first device.

[0031] In a possible implementation manner of the second aspect, the method further includes: the second device sends the second weight to the third device. In this implementation manner, the second weight affects the importance of the generated second simulation data set, which is beneficial to the second device to obtain the first simulation data set meeting corresponding importance from the third device.

[0032] In a possible implementation manner of the second aspect, the first indication information includes the weight of each first simulation data set in the at least one first simulation data set. This is beneficial to the first device to feed back the weight of each first simulation data set in the at least one first simulation data set to the second device.

[0033] In a possible implementation manner of the second aspect, the first indication information includes the identity of each first simulation data set in the at least one first simulation data set, and the identity of each first simulation data set is used to indicate the weight of each first simulation data set in the at least one first simulation data set. In this way, the weight of each first simulation data set is indicated by the identity of each first simulation data set. This is beneficial to reducing the indication overhead of the first device.

[0034] The third aspect of the present application provides a data transmission method, which can be executed by a first device. The first device can be a terminal device, a network device (such as an access network device or a core network device), or a component (for example, a processor, a chip, or a chip system) in the terminal device or the network device, or a logic module or software capable of realizing all or part of the functions of the terminal device or the network device. The method comprises the following steps: determining, by the first device, first data and first indication information, the first indication information being used for indicating at least one first generation configuration, the at least one first generation configuration being used for generating at least one first simulation data set based on the first data, the first data being local data of the first device, each first simulation data set in the at least one first simulation data set having a corresponding weight, and the weight of each first simulation data set being used for indicating the importance of the first simulation data set; and sending, by the first device, the first data and the first indication information to a second device.

[0035] According to the above technical solution, the first device indicates the at least one first generation configuration to the second device. This facilitates the second device to generate the at least one first simulation data set based on the at least one first generation configuration. The second device can perform model training based on the at least one first simulation data set and the first data, thereby increasing the amount of data used for model training, which is conducive to improving the model performance and / or the generalization performance of the model obtained through training. Furthermore, each first simulation data set in the at least one first simulation data set has a corresponding weight, and the weight of each first simulation data set is used for indicating the importance of the first simulation data set. This facilitates the second device to select appropriate data for corresponding model training based on the weight of each first simulation data set, thereby meeting the needs of different models. For example, for a model with high model performance requirements, the second device can preferentially select a first simulation data set with a larger weight for model training. For a model with high generalization performance requirements, the second device can preferentially select a first simulation data set with a smaller weight for model training.

[0036] Based on the third aspect, in a possible implementation manner, each first generation configuration in the at least one first generation configuration corresponds to one or more first simulation data sets in the at least one first simulation data set, and each first generation configuration in the at least one first generation configuration comprises information or parameters used for assisting in generating the first simulation data set corresponding to the first generation configuration. A possible example of the content included in the at least one first generation configuration is provided, which facilitates the second device to generate the first simulation data set based on the first generation configuration.

[0037] In a possible implementation manner of the third aspect, the method further includes: the first device sending second indication information to the second device, the second indication information being used to indicate the data generation manner corresponding to each first simulation data set in the at least one first simulation data set; the first device receiving first confirmation information from the second device, the first confirmation information being used to confirm the data generation manner corresponding to each first simulation data set in the at least one first simulation data set; and the at least one first generation configuration being determined according to the data generation manner corresponding to each first simulation data set in the at least one first simulation data set. The data generation manner corresponding to each first simulation data set is aligned between the first device and the second device. The first device is facilitated to determine the at least one first generation configuration.

[0038] In a possible implementation manner of the third aspect, the method further includes: the first device receiving, from the second device, an expected weight of each first simulation data set in the at least one first simulation data set, the expected weight being used to indicate the importance of the expected first simulation data set; and the at least one first generation configuration being determined according to the expected weight of each first simulation data set in the at least one first simulation data set. In this implementation manner, the expected weight affects the setting of the first generation configuration, thereby affecting the importance of the first simulation data set. The first device is facilitated to determine the appropriate at least one first generation configuration in combination with the expected weight of each first simulation data set, and indicate the at least one first generation configuration to the second device.

[0039] In a possible implementation manner of the third aspect, the method further includes: the first device receiving, from the second device, a first weight, the first weight being used to indicate the importance of data of the first device; and the at least one first generation configuration being determined according to the first weight. In this implementation manner, the first weight affects the setting of the first generation configuration, thereby affecting the importance of the first simulation data set. The first device is facilitated to determine the appropriate at least one first generation configuration in combination with the first weight, and indicate the at least one first generation configuration to the second device.

[0040] The fourth aspect of the present application provides a data transmission method, which can be executed by a second device. The second device can be a terminal device, a network device (such as an access network device or a core network device), or a component (for example, a processor, a chip, or a chip system) in the terminal device or the network device, or a logic module or software capable of realizing all or part of the functions of the terminal device or the network device. The method comprises the following steps: the second device receives first data and first indication information from a first device, the first indication information is used to indicate at least one first generation configuration, and the first data is local data of the first device; and the second device generates at least one first simulation data set according to the first data and the at least one first generation configuration, each first simulation data set in the at least one first simulation data set has a corresponding weight, and the weight of each first simulation data set is used to indicate the importance of the first simulation data set.

[0041] According to the above technical solution, the second device receives first data and first indication information from the first device, and the first indication information is used to indicate at least one first generation configuration. The second device generates at least one first simulation data set according to the first data and the at least one first generation configuration. This facilitates the second device to perform model training according to the first data and the at least one first simulation data set, thereby increasing the amount of data used for model training, which is beneficial to improving the model performance and / or generalization performance of the trained model. Further, each first simulation data set in the at least one first simulation data set has a corresponding weight, and the weight of each first simulation data set is used to indicate the importance of each first simulation data set. This is beneficial to the second device to select appropriate data for corresponding model training based on the weight of each first simulation data set, thereby meeting the needs of different models. For example, for a model with high model performance requirements, the second device can preferentially select a first simulation data set with a larger weight for model training. For a model with high generalization performance requirements, the second device can preferentially select a first simulation data set with a smaller weight for model training.

[0042] Based on the fourth aspect, in a possible implementation, each first generation configuration in the at least one first generation configuration corresponds to one or more first simulation data sets in the at least one first simulation data set, and each first generation configuration in the at least one first generation configuration comprises information or parameters used to assist in generating the first simulation data set corresponding to the first generation configuration. A possible example of the content included in the at least one first generation configuration is provided. This facilitates the second device to generate the first simulation data set based on the first generation configuration.

[0043] In a possible implementation manner of the fourth aspect, the method further includes: receiving, by the second device, second indication information from the first device, the second indication information being used to indicate the data generation manner corresponding to each of the at least one first simulation data set; and sending, by the second device, first confirmation information to the first device, the first confirmation information being used to confirm the data generation manner corresponding to each of the at least one first simulation data set. The data generation manner corresponding to each of the at least one first simulation data set is aligned between the first device and the second device. This facilitates the first device to determine the at least one first generation configuration.

[0044] In a possible implementation manner of the fourth aspect, the second device generates the at least one first simulation data set according to the first data and the at least one first generation configuration, including: each of the at least one first simulation data set has a corresponding data generation manner; and the second device generates the first simulation data set according to the first data, the data generation manner, and the first generation configuration corresponding to the first simulation data set. This facilitates the second device to generate the first simulation data set. In the future, the second device can perform model training based on the second simulation data set, so as to improve the model performance and / or the generalization performance of the model.

[0045] In a possible implementation manner of the fourth aspect, the data generation manner includes: generating the first simulation data set according to the first data and the first model; or performing data enhancement on the first data to obtain the first simulation data set. This implementation manner provides two different data generation manners, and enriches the implementation of the scheme. This facilitates the second device to generate more simulation data sets.

[0046] In a possible implementation manner of the fourth aspect, the method further includes: performing, by the second device, model training according to the first data, the at least one first simulation data set, and the weight of each of the at least one first simulation data set, to obtain one or more models. This increases the amount of data used for model training, and is beneficial to improving the model performance and / or the generalization performance of the model obtained through training. Further, the second device performs model training in combination with the weight of each of the at least one first simulation data set. This is beneficial to the second device to select appropriate data for corresponding model training based on the weight of each of the at least one first simulation data set, so as to meet the requirements of different models. For example, for a model with a higher requirement on model performance, the second device can preferentially select a first simulation data set with a larger weight for model training. For a model with a higher requirement on generalization performance, the second device can preferentially select a first simulation data set with a smaller weight for model training.

[0047] In a possible implementation manner based on the fourth aspect, the one or more models comprise a first model; and the model training by the second device according to the first data, the at least one first simulation data set, and the weight of each first simulation data set in the at least one first simulation data set to obtain the one or more models comprises: selecting, by the second device, target data from the first data and the at least one first simulation data set according to at least one of a model performance requirement and a generalization performance requirement of the first model and the weight of each first simulation data set in the at least one first simulation data set; and performing, by the second device, model training by using the target data to obtain the first model. In this way, the second device can select appropriate data to perform model training to obtain the second model, so that the second model obtained by training meets the corresponding performance requirement.

[0048] In a possible implementation manner based on the fourth aspect, the method further comprises: receiving, by the second device, second data and at least one second generation configuration from a third device, the second data being local data of the third device; generating, by the second device, at least one second simulation data set according to the second data and the at least one second generation configuration, each second simulation data set in the at least one second simulation data set having a corresponding weight, the weight of each second simulation data set being used to indicate an importance of the second simulation data set; and performing, by the second device, model training according to the first data, the at least one first simulation data set, and the weight of each first simulation data set in the at least one first simulation data set to obtain the one or more models, comprising: performing, by the second device, model training according to the first data, the second data, the at least one first simulation data set, the at least one second simulation data set, the weight of each first simulation data set in the at least one first simulation data set, and the weight of each second simulation data set in the at least one second simulation data set to obtain the one or more models. In this implementation manner, the second device performs model training in combination with the local data of multiple devices, the simulation data sets, and the weights of the simulation data sets. That is, the second device performs model training by fusing data of multiple manufacturers or multiple users. This is beneficial to improving the model performance and / or the generalization performance of the model obtained by training.

[0049] In a possible implementation manner based on the fourth aspect, in the model training performed by the second device according to the first data, the second data, the at least one first simulation data set, the weight of each first simulation data set in the at least one first simulation data set, and the weight of each second simulation data set in the at least one second simulation data set, the second device determines a target data set according to a first weight, the first data, the at least one first simulation data set, the weight of each first simulation data set in the at least one first simulation data set, a second weight, the second data, the at least one second simulation data set, and the weight of each second simulation data set in the at least one second simulation data set, where the first weight is used to indicate the importance of the data of the first device, and the second weight is used to indicate the importance of the data of the third device, and the second device performs the model training by using the target data set to obtain the one or more models. In this implementation manner, when the second device performs the model training by using the local data and the simulation data set of the plurality of devices, the second device further selects the data according to the importance of the data of different devices, and then performs the model training. This is beneficial to the model trained by the second device to meet the corresponding performance requirement. For example, for a model with a high performance requirement, the second device can select the data of a device with a high importance to perform the model training.

[0050] In a possible implementation manner based on the fourth aspect, the method further includes that the second device sends the first weight to the first device. In this implementation manner, the first weight affects the setting of the at least one first generated configuration. This is beneficial to the second device to obtain the corresponding at least one first generated configuration from the first device.

[0051] In a possible implementation manner based on the fourth aspect, the method further includes that the second device sends the second weight to the third device. In this implementation manner, the second weight affects the setting of the at least one second generated configuration. This is beneficial to the second device to obtain the corresponding at least one second generated configuration from the third device.

[0052] In a possible implementation manner based on the fourth aspect, the method further includes that the second device sends the expected weight of each first simulation data set in the at least one first simulation data set to the first device, where the expected weight is used to indicate the expected weight of the first simulation data set. The expected weight of each first simulation data set in the at least one first simulation data set affects the setting of the at least one second generated configuration. This is beneficial to the second device to obtain the corresponding at least one first generated configuration from the first device.

[0053] In a possible implementation manner of the third aspect or the fourth aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a similarity between each first simulation data set in the at least one first simulation data set and the first data. A manner of determining the weight of the first simulation data set is shown, which facilitates implementation of the scheme. Further, the weight of the first simulation data set represents an importance of the first simulation data set. The importance of the first simulation data set can be better represented by the similarity between the first simulation data set and the first data. For example, the higher the similarity between the first simulation data set and the first data, the higher the importance of the first simulation data set, that is, the higher the weight of the first simulation data set.

[0054] In a possible implementation manner of the third aspect or the fourth aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a data distribution difference between each first simulation data set in the at least one first simulation data set and the first data. Another manner of determining the weight of the first simulation data set is shown, which facilitates implementation of the scheme. Further, the weight of the first simulation data set represents an importance of the first simulation data set. The importance of the first simulation data set can be better represented by the data distribution difference between the first simulation data set and the first data. For example, the higher the data distribution difference between the first simulation data set and the first data, the lower the importance of the first simulation data set, that is, the lower the weight of the first simulation data set.

[0055] In a possible implementation manner of the third aspect or the fourth aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a signal-to-noise ratio difference between each first simulation data set in the at least one first simulation data set and the first data. Another manner of determining the weight of the first simulation data set is shown, which facilitates implementation of the scheme. Further, the weight of the first simulation data set represents an importance of the first simulation data set. The importance of the first simulation data set can be better represented by the signal-to-noise ratio difference between the first simulation data set and the first data. For example, the higher the signal-to-noise ratio difference between the first simulation data set and the first data, the lower the importance of the first simulation data set, that is, the lower the weight of the first simulation data set.

[0056] In a possible implementation manner of the third aspect or the fourth aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a position difference between each first simulation data set in the at least one first simulation data set and the first data. Another determination manner of the weight of the first simulation data set is shown, which facilitates implementation of the scheme. Further, the weight of the first simulation data set represents the importance of the first simulation data set. The importance of the first simulation data set can be better represented by the position difference between the first simulation data set and the first data. For example, if the position difference between the first simulation data set and the first data is high, the importance of the first simulation data set is low, that is, the weight of the first simulation data set is low.

[0057] The fifth aspect of the present application provides a first device, comprising:

[0058] a processing module, configured to generate at least one first simulation data set according to first data, the first data being local data of the first device, each first simulation data set in the at least one first simulation data set having a corresponding weight, and the weight of each first simulation data set being used to indicate the importance of each first simulation data set;

[0059] a transceiver, configured to send the first data, the at least one first simulation data set and first indication information to a second device, the first indication information being used to indicate the weight of each first simulation data set in the at least one first simulation data set.

[0060] In a possible implementation manner of the fifth aspect, the transceiver is further configured to send the weight of the first data to the second device, the weight of the first data being used to indicate the importance of the first data.

[0061] In a possible implementation manner of the fifth aspect, each first simulation data set in the at least one first simulation data set has a corresponding data generation manner; and the processing module is specifically configured to generate the at least one first simulation data set according to the data generation manner corresponding to each first simulation data set in the at least one first simulation data set and the first data.

[0062] In a possible implementation manner of the fifth aspect, the data generation manner comprises: generating the first simulation data set according to the first data and a first model; or performing data enhancement on the first data to obtain the first simulation data set.

[0063] In a possible implementation manner of the fifth aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to the similarity between each first simulation data set in the at least one first simulation data set and the first data.

[0064] In a possible implementation manner of the fifth aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a difference in data distribution between each first simulation data set in the at least one first simulation data set and the first data.

[0065] In a possible implementation manner of the fifth aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a difference in signal-to-noise ratio between each first simulation data set in the at least one first simulation data set and the first data.

[0066] In a possible implementation manner of the fifth aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a difference in position between each first simulation data set in the at least one first simulation data set and the first data.

[0067] In a possible implementation manner of the fifth aspect, the transceiver is further configured to receive, from the second device, an expected weight of each first simulation data set in the at least one first simulation data set, the expected weight being used to indicate an importance of an expected first simulation data set; and the processor is specifically configured to generate the at least one first simulation data set according to the first data and the expected weight of each first simulation data set in the at least one first simulation data set.

[0068] In a possible implementation manner of the fifth aspect, the transceiver is further configured to receive, from the second device, a first weight, the first weight being used to indicate an importance of the first device; and the processor is specifically configured to generate the at least one first simulation data set according to the first data and the first weight.

[0069] In a possible implementation manner of the fifth aspect, the first indication information includes the weight of each first simulation data set in the at least one first simulation data set.

[0070] In a possible implementation manner of the fifth aspect, the first indication information includes an identifier of each first simulation data set in the at least one first simulation data set, the identifier of each first simulation data set being used to indicate the weight of each first simulation data set.

[0071] The sixth aspect of the present application provides a second device, comprising:

[0072] a transceiver configured to receive, from a first device, first data, at least one first simulation data set, and first indication information, the first data being local data of the first device, the at least one first simulation data set being generated according to the first data, and the first indication information being used to indicate a weight of each first simulation data set in the at least one first simulation data set, the weight of each first simulation data set being used to indicate an importance of each first simulation data set;

[0073] The processing module is configured to perform model training based on the first data, the at least one first simulation data set, and the weight of each first simulation data set in the at least one first simulation data set, to obtain one or more models.

[0074] In a possible implementation of the sixth aspect, the transceiver is further configured to receive the weight of the first data from the first device, the weight of the first data being used to indicate the importance of the first data; and the processing module is specifically configured to perform model training based on the first data, the weight of the first data, the at least one first simulation data set, and the weight of each first simulation data set in the at least one first simulation data set, to obtain the one or more models.

[0075] In a possible implementation of the sixth aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to the similarity between each first simulation data set in the at least one first simulation data set and the first data.

[0076] In a possible implementation of the sixth aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to the data distribution difference between each first simulation data set in the at least one first simulation data set and the first data.

[0077] In a possible implementation of the sixth aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to the signal-to-noise ratio difference between each first simulation data set in the at least one first simulation data set and the first data.

[0078] In a possible implementation of the sixth aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to the position difference between each first simulation data set in the at least one first simulation data set and the first data.

[0079] In a possible implementation of the sixth aspect, the transceiver is further configured to send, to the first device, the expected weight of each first simulation data set in the at least one first simulation data set, the expected weight being used to indicate the importance of the expected first simulation data set.

[0080] In a possible implementation of the sixth aspect, the one or more models include a second model; and the processing module is specifically configured to select target data from the first data and the at least one first simulation data set according to at least one of a model performance requirement and a generalization performance requirement of the second model and the weight of each first simulation data set in the at least one first simulation data set, and perform model training based on the target data to obtain the second model.

[0081] In a possible implementation manner based on the sixth aspect, the transceiving module is further configured to receive second data, at least one second simulation data set and second indication information from the third device, the second data being local data of the third device, the at least one second simulation data set being generated according to the second data, and the second indication information being used for indicating weights of respective second simulation data sets in the at least one second simulation data set, the weight of each second simulation data set in the at least one second simulation data set being used for indicating importance of the second simulation data set; and the processing module is specifically configured to perform model training according to the first data, the at least one first simulation data set, the second data, the at least one second simulation data set, the weight of each first simulation data set in the at least one first simulation data set and the weight of each second simulation data set in the at least one second simulation data set, to obtain one or more models.

[0082] In a possible implementation manner based on the sixth aspect, the processing module is specifically configured to determine a target data set according to the first weight, the first data, the at least one first simulation data set, the weight of each first simulation data set in the at least one first simulation data set, the second weight, the second data, the at least one second simulation data set and the weight of each second simulation data set in the at least one second simulation data set, the first weight being used for indicating importance of data of the first device, and the second weight being used for indicating importance of data of the third device; and perform model training through the target data set, to obtain one or more models.

[0083] In a possible implementation manner based on the sixth aspect, the transceiving module is further configured to send the first weight to the first device.

[0084] In a possible implementation manner based on the sixth aspect, the transceiving module is further configured to send the second weight to the third device.

[0085] In a possible implementation manner based on the sixth aspect, the first indication information includes the weight of each first simulation data set in the at least one first simulation data set.

[0086] In a possible implementation manner based on the sixth aspect, the first indication information includes an identifier of each first simulation data set in the at least one first simulation data set, the identifier of each first simulation data set being used for indicating the weight of each first simulation data set in the at least one first simulation data set.

[0087] The seventh aspect of the present application provides a first device, comprising:

[0088] The processing module is configured to determine first data and first indication information, the first indication information being used to indicate at least one first generation configuration, the at least one first generation configuration being used to generate at least one first simulation data set based on the first data, the first data being local data of the first device, each first simulation data set in the at least one first simulation data set having a corresponding weight, and the weight of each first simulation data set being used to indicate an importance of the first simulation data set;

[0089] The transceiver module is configured to send the first data and the first indication information to the second device.

[0090] According to a possible implementation of the seventh aspect, in each of the at least one first generation configuration, one or more first simulation data sets corresponding to the first generation configuration, and each of the at least one first generation configuration includes information or parameters used to assist in generating the first simulation data set corresponding to the first generation configuration.

[0091] According to a possible implementation of the seventh aspect, the transceiver module is further configured to: send, to the second device, second indication information used to indicate a data generation manner corresponding to each first simulation data set in the at least one first simulation data set; receive, from the second device, first confirmation information used to confirm the data generation manner corresponding to each first simulation data set in the at least one first simulation data set; and determine the at least one first generation configuration according to the data generation manner corresponding to each first simulation data set in the at least one first simulation data set.

[0092] According to a possible implementation of the seventh aspect, the transceiver module is further configured to: receive, from the second device, an expected weight of each first simulation data set in the at least one first simulation data set, the expected weight being used to indicate an expected importance of the first simulation data set; and determine the at least one first generation configuration according to the expected weight of each first simulation data set in the at least one first simulation data set.

[0093] According to a possible implementation of the seventh aspect, the transceiver module is further configured to: receive, from the second device, a first weight used to indicate an importance of data of the first device; and determine the at least one first generation configuration according to the first weight.

[0094] The eighth aspect of the present application provides a second device, comprising:

[0095] The transceiver module is configured to receive, from the first device, first data and first indication information, the first indication information being used to indicate at least one first generation configuration, and the first data being local data of the first device.

[0096] The processing module is configured to generate at least one first simulation data set according to the first data and at least one first generation configuration, each first simulation data set in the at least one first simulation data set having a corresponding weight, and the weight of each first simulation data set being used to indicate an importance of the first simulation data set.

[0097] Based on the eighth aspect, in a possible implementation, each first generation configuration in the at least one first generation configuration corresponds to one or more first simulation data sets in the at least one first simulation data set, and each first generation configuration in the at least one first generation configuration includes information or parameters used to assist in generating the first simulation data set corresponding to the first generation configuration.

[0098] Based on the eighth aspect, in a possible implementation, the transceiver module is further configured to: receive second indication information from the first device, the second indication information being used to indicate a data generation manner corresponding to each first simulation data set in the at least one first simulation data set; and send first confirmation information to the first device, the first confirmation information being used to confirm the data generation manner corresponding to each first simulation data set in the at least one first simulation data set.

[0099] Based on the eighth aspect, in a possible implementation, the processing module is specifically configured to: each first simulation data set in the at least one first simulation data set has a corresponding data generation manner; and generate the first simulation data set according to the first data, the data generation manner, and the first generation configuration corresponding to the first simulation data set.

[0100] Based on the eighth aspect, in a possible implementation, the data generation manner includes: generating the first simulation data set according to the first data and a first model; or performing data enhancement on the first data to obtain the first simulation data set.

[0101] Based on the eighth aspect, in a possible implementation, the processing module is further configured to: perform model training according to the first data, the at least one first simulation data set, and the weight of each first simulation data set in the at least one first simulation data set, to obtain one or more models.

[0102] Based on the eighth aspect, in a possible implementation, the one or more models include the first model; and the processing module is specifically configured to: select target data from the first data and the at least one first simulation data set according to at least one of a model performance requirement and a generalization performance requirement of the first model and the weight of each first simulation data set in the at least one first simulation data set; and perform model training by using the target data to obtain the first model.

[0103] In a possible implementation manner based on the eighth aspect, the transceiver is further configured to receive second data and at least one second generation configuration from the third device, the second data being local data of the third device; and the processor is further configured to generate at least one second simulation data set according to the second data and the at least one second generation configuration, each second simulation data set in the at least one second simulation data set having a corresponding weight, the weight of each second simulation data set being used to indicate an importance of the second simulation data set; and the processor is specifically configured to perform model training according to the first data, the second data, the at least one first simulation data set, the at least one second simulation data set, the weight of each first simulation data set in the at least one first simulation data set, and the weight of each second simulation data set in the at least one second simulation data set, to obtain one or more models.

[0104] In a possible implementation manner based on the eighth aspect, the processor is specifically configured to determine a target data set according to the first weight, the first data, the at least one first simulation data set and the weight of each first simulation data set in the at least one first simulation data set, the second weight, the second data, the at least one second simulation data set and the weight of each second simulation data set in the at least one second simulation data set, and perform model training on the target data set to obtain one or more models.

[0105] In a possible implementation manner based on the eighth aspect, the transceiver is further configured to send the first weight to the first device.

[0106] In a possible implementation manner based on the eighth aspect, the transceiver is further configured to send the second weight to the third device.

[0107] In a possible implementation manner based on the eighth aspect, the transceiver is further configured to send, to the first device, an expected weight of each first simulation data set in the at least one first simulation data set, the expected weight being used to indicate an expected weight of the first simulation data set.

[0108] In a possible implementation manner based on the seventh aspect or the eighth aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a similarity between each first simulation data set in the at least one first simulation data set and the first data.

[0109] In a possible implementation manner based on the seventh aspect or the eighth aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a data distribution difference between each first simulation data set in the at least one first simulation data set and the first data.

[0110] In a possible implementation manner of the seventh aspect or the eighth aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a signal-to-noise ratio difference between each first simulation data set in the at least one first simulation data set and the first data.

[0111] In a possible implementation manner of the seventh aspect or the eighth aspect, the weight of each first simulation data set in the at least one first simulation data set is determined according to a position difference between each first simulation data set in the at least one first simulation data set and the first data.

[0112] For the fifth aspect or the seventh aspect, the first apparatus can be a terminal device, a network device (for example, an access network device or a core network device), or a component (for example, a processor, a chip, or a chip system) in the terminal device or the network device, or logic modules or software capable of implementing all or part of the terminal device function, or logic modules or software capable of implementing all or part of the network device function. The transceiver module can be a transceiver, or an input / output interface; and the processing module can be a processor.

[0113] In an implementation manner, the first apparatus is a chip, a chip system, or a circuit configured in the terminal device or the network device. When the first apparatus is the chip, the chip system, or the circuit configured in the terminal device or the network device, the transceiver module can be an input / output interface, an interface circuit, an output circuit, an input circuit, a pin, or related circuit on the chip, the chip system, or the circuit; and the processing module can be a processor, a processing circuit, or a logic circuit.

[0114] For the sixth aspect or the eighth aspect, the second apparatus can be a terminal device, a network device (for example, an access network device or a core network device), or a component (for example, a processor, a chip, or a chip system) in the terminal device or the network device, or logic modules or software capable of implementing all or part of the terminal device function, or logic modules or software capable of implementing all or part of the network device function. The transceiver module can be a transceiver, or an input / output interface; and the processing module can be a processor.

[0115] In an implementation manner, the second apparatus is a chip, a chip system, or a circuit configured in the terminal device or the network device. When the second apparatus is the chip, the chip system, or the circuit configured in the terminal device or the network device, the transceiver module can be an input / output interface, an interface circuit, an output circuit, an input circuit, a pin, or related circuit on the chip, the chip system, or the circuit; and the processing module can be a processor, a processing circuit, or a logic circuit.

[0116] The first aspect of the present application provides a first device, comprising a processor and a memory. The memory stores a computer program or computer instructions. The processor is configured to invoke and execute the computer program or computer instructions stored in the memory, so that the processor implements any one of the implementation manners of the first aspect.

[0117] Optionally, the first device further comprises a transceiver, and the processor is configured to control the transceiver to transceive signals.

[0118] The tenth aspect of the present application provides a second device, comprising a processor and a memory. The memory stores a computer program or computer instructions. The processor is configured to invoke and execute the computer program or computer instructions stored in the memory, so that the processor implements any one of the implementation manners of the second aspect.

[0119] Optionally, the second device further comprises a transceiver, and the processor is configured to control the transceiver to transceive signals.

[0120] The eleventh aspect of the present application provides a first device, comprising a processor and an interface circuit. The processor is configured to communicate with other devices through the interface circuit and execute the method of the first aspect or the second aspect. The processor comprises one or more.

[0121] The twelfth aspect of the present application provides a second device, comprising a processor and an interface circuit. The processor is configured to communicate with other devices through the interface circuit and execute the method of the second aspect or the fourth aspect. The processor comprises one or more.

[0122] The thirteenth aspect of the present application provides a first device, comprising a processor. The processor is configured to be connected with a memory and invoke the program stored in the memory to execute the method of the first aspect or the second aspect. The memory can be located in the first device or outside the first device. The processor comprises one or more.

[0123] The fourteenth aspect of the present application provides a second device, comprising a processor. The processor is configured to be connected with a memory and invoke the program stored in the memory to execute the method of the second aspect or the fourth aspect. The memory can be located in the second device or outside the second device. The processor comprises one or more.

[0124] In an implementation manner, the first device of the first aspect, the third aspect, the fifth aspect and the seventh aspect can be a chip or a chip system. The second device of the second aspect, the fourth aspect, the sixth aspect and the eighth aspect can be a chip or a chip system.

[0125] The fifteenth aspect of the present application provides a computer program product comprising computer instructions, characterized by causing a computer to perform any of the implementation manners of any one of the first aspect to the fourth aspect when the instructions are run on the computer.

[0126] The sixteenth aspect of the present application provides a computer readable storage medium comprising computer instructions, characterized by causing a computer to perform any of the implementation manners of any one of the first aspect to the fourth aspect when the instructions are run on the computer.

[0127] The seventeenth aspect of the present application provides a chip device comprising a processor, configured to invoke a computer program or computer instructions in a memory, so as to cause the processor to perform any of the implementation manners of any one of the first aspect to the fourth aspect.

[0128] Optionally, the processor is coupled with the memory through an interface.

[0129] The eighteenth aspect of the present application provides a communication system, comprising the first device shown in the first aspect and the second device shown in the second aspect; or the communication system comprises the first device shown in the third aspect and the second device shown in the fourth aspect.

[0130] As described in the above technical solution, the first device generates at least one first simulation dataset based on first data, where the first data is local data of the first device. Each of the at least one first simulation datasets has a corresponding weight, which indicates the importance of each first simulation dataset. Then, the first device sends the first data, at least one first simulation dataset, and first indication information to the second device. The first indication information indicates the weight of each of the at least one first simulation dataset. Therefore, the first device generates at least one first simulation dataset based on the first data and sends the first data, at least one first simulation dataset, and first indication information to the second device. This means the first device provides more data to the second device; in addition to providing the first data, the first device also provides at least one first simulation dataset. This enables the second device to train a model based on the first data and at least one first simulation dataset. This increases the amount of data used for model training, which is beneficial for improving the model performance and / or generalization performance of the trained model. Furthermore, the first indication information indicates the weight of each of the at least one first simulation dataset, and the weight of each first simulation dataset indicates the importance of each first simulation dataset. This allows the second device to select appropriate data for model training based on the weights of each first simulation dataset, thus meeting the needs of different models. For example, for models with high performance requirements, the second device can prioritize training the first simulation dataset with larger weights. For models with high generalization performance requirements, the second device can prioritize training the first simulation dataset with smaller weights. Attached Figure Description

[0131] Figure 1 is a schematic diagram of a wireless AI framework according to an embodiment of this application;

[0132] Figure 2A is a schematic diagram of a communication system according to an embodiment of this application;

[0133] Figure 2B is another schematic diagram of the communication system according to an embodiment of this application;

[0134] Figure 3 is a schematic diagram of an embodiment of the data transmission method of this application;

[0135] Figure 4 is a schematic diagram of the identifier of the first data and the identifier of at least one first simulation dataset in an embodiment of this application;

[0136] Figure 5 is a schematic diagram of one or more models obtained by the second device in the embodiment of this application through model training;

[0137] Figure 6 is a schematic diagram of another embodiment of the data transmission method of this application;

[0138] FIG. 7 is a structural schematic diagram of a first device according to an embodiment of the present application;

[0139] FIG. 8 is a structural schematic diagram of a second device according to an embodiment of the present application;

[0140] FIG. 9 is a structural schematic diagram of a device according to an embodiment of the present application;

[0141] FIG. 10 is a structural schematic diagram of a terminal device according to an embodiment of the present application;

[0142] FIG. 11 is a structural schematic diagram of a network device according to an embodiment of the present application. DETAILED DESCRIPTION

[0143] The embodiments of the present application provide a data transmission method and related devices, which are used for a first device to generate at least one first simulation data set according to first data, and to send the first data, the at least one first simulation data set and first indication information to a second device. Therefore, the first device generates at least one first simulation data set according to the first data, and sends the first data, the at least one first simulation data set and the first indication information to the second device. Therefore, the first device provides more data to the second device, that is, the first device provides not only the first data but also the at least one first simulation data set. The second device performs model training based on the first data and the at least one first simulation data set. Therefore, the amount of data used for model training is increased, which is beneficial to improving the model performance and / or generalization performance of the model obtained by training. Further, the first indication information is used to indicate the weight of each first simulation data set in the at least one first simulation data set, and the weight of each first simulation data set is used to indicate the importance of the first simulation data set. This is beneficial to the second device to select appropriate data for corresponding model training based on the weight of each first simulation data set, so as to meet the needs of different models. For example, for a model with higher model performance requirement, the second device can preferentially select the first simulation data set with larger weight for model training. For a model with higher generalization performance requirement, the second device can preferentially select the first simulation data set with smaller weight for model training.

[0144] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person skilled in the art without creative work fall within the protection scope of the present application.

[0145] Reference within this application to "one embodiment" or "an embodiment" or "some embodiments" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment. The appearances of the phrase "in one embodiment" or "in some embodiments" in various places within specifications are not necessarily all referring to the same embodiment, although it can. The terms "including," "comprising," "having" and variations thereof are meant to encompass the items listed thereafter and equivalents thereof as well as additional items. Unless specified otherwise, "or" as used herein is

[0146] In the description of the application, unless otherwise specified, " / " means "or", for example, A / B can mean A or B. "And / or" in this article is only a description of the relationship between the associated objects, which means that there can be three relationships, for example, A and / or B, which can mean: A alone, A and B exist at the same time, and B alone. In addition, "at least one" means one or more, and "multiple" means two or more. "At least one of the following" or similar expressions means any combination of the items, including any combination of single or multiple items. For example, at least one of a, b, or c, can mean a, b, c; a and b; a and c; b and c; or a and b and c. Where a, b, c can be single or multiple.

[0147] It can be understood that in this application, "indication" can include direct indication, indirect indication, display indication, and implicit indication. When describing a certain indication information for indicating A, it can be understood that the indication information carries A, directly indicates A, or indirectly indicates A.

[0148] In this application, the information indicated by the indication information is called the to-be-indicated information. In the specific 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, or the to-be-indicated information can be indirectly indicated by indicating other information, wherein the other information and the to-be-indicated information have an association relationship. It can also only indicate part of the to-be-indicated information, 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 also be achieved by means of the arrangement order of each information agreed in advance (for example, the protocol stipulates), thereby reducing the indication overhead to a certain extent.

[0149] The to-be-indicated information can be sent as a whole or can be divided into multiple sub-information and sent separately, and the sending period and / or sending occasion of the sub-information can be the same or different. The specific sending method is not limited in the present application. The sending period and / or sending occasion of the sub-information can be predefined, for example, predefined according to a protocol, or configured by the transmitting end device through sending configuration information to the receiving end device.

[0150] It can be understood that “sending” and “receiving” in the present application 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 a chip interface, and “receiving” can also be understood as “input” of a chip interface.

[0151] In other words, sending and receiving can be between devices, for example, between network devices and terminal devices, or within a device, for example, between components, modules, chips, software modules or hardware modules in a device through a bus, wire or interface.

[0152] It can be understood that the information can be processed as necessary between the source and the destination of the information transmission, such as encoding and modulation, but the destination can understand the effective information from the source. Similar expressions in the present application can be understood similarly, and will not be repeated here.

[0153] The technical solutions of the present application can be applied to a third generation partnership project (3rd generation partnership project, 3GPP) related cellular communication system. For example, a fourth generation (4th generation, 4G) communication system, a fifth generation (5th generation, 5G) communication system, a future communication system after the fifth generation communication system. For example, the fourth generation communication system can include a long term evolution (long term evolution, LTE) communication system. The fifth generation communication system can include a new radio (new radio, NR) communication system. The technical solutions of the present application can also be applied to a wireless fidelity (wireless fidelity, WiFi) system, a communication system supporting multiple wireless technology integration, a device-to-device (device-to-device, D2D) system, or a vehicle-to-everything (vehicle to everything, V2X) communication system, etc.

[0154] The technical solution of the present application is applicable to a communication system including a first device and a second device. The first device and the second device can perform the technical solution of the present application.

[0155] In a possible implementation, the first device is a terminal device, or a chip, a chip system, or a processor in the terminal device, or a logic module or software for realizing part or all of the functions of the terminal device. The second device is an access network device, or a chip, a chip system, or a processor in the access network device, or a logic module or software for realizing part or all of the functions of the access network device. For example, as shown in FIG. 2A, the first device is a terminal device 201, and the second device is an access network device 202.

[0156] In another possible implementation, the first device is an access network device, or a chip, a chip system, or a processor in the access network device, or a logic module or software for realizing part or all of the functions of the access network device. The second device is a terminal device, or a chip, a chip system, or a processor in the terminal device, or a logic module or software for realizing part or all of the functions of the terminal device. For example, as shown in FIG. 2A, the first device is an access network device 202, and the second device is a terminal device 201.

[0157] In yet another possible implementation, the first device is an access network device, or a chip, a chip system, or a processor in the access network device, or a logic module or software for realizing part or all of the functions of the access network device. The second device is a core network device, or a chip, a chip system, or a processor in the core network device, or a logic module or software for realizing part or all of the functions of the core network device. For example, as shown in FIG. 2A, the first device is an access network device 202, and the second device is a core network device 203.

[0158] In yet another possible implementation, the first device is a core network device, or a chip, a chip system, or a processor in the core network device, or a logic module or software for realizing part or all of the functions of the core network device. The second device is an access network device, or a chip, a chip system, or a processor in the access network device, or a logic module or software for realizing part or all of the functions of the access network device. For example, as shown in FIG. 2A, the first device is a core network device 203, and the second device is an access network device 202.

[0159] In another possible implementation, the first device is a first terminal device, or a chip, chip system, or processor in the first terminal device, or a logic module or software for implementing part or whole function of the first terminal device. The second device is a second terminal device, or a chip, chip system, or processor in the second terminal device, or a logic module or software for implementing part or whole function of the second terminal device. For example, as shown in FIG. 2B, the first device is terminal device 1, and the second device is terminal device 2.

[0160] The first device and the second device can also be other forms of devices, which are not limited in the present application.

[0161] The terminal device, the access network device, and the core network device involved in the present application are described as follows.

[0162] The terminal device can be a wireless terminal device capable of receiving scheduling information and indication information of the access network device. The wireless terminal device can be a device providing voice and / or data connectivity to a user, or a handheld device with wireless connection function, or another processing device connected to a wireless modem.

[0163] The terminal device can communicate with one or more core networks or the Internet via an access network. The terminal device can be a mobile terminal device, such as a mobile phone (or called "cellular" phone, mobile phone), a computer and a data card, for example, can be a portable, pocket-sized, handheld, computer-built-in or vehicle-mounted mobile device, which exchanges voice and / or data with a wireless access network. For example, personal communication service (PCS) phones, cordless phones, session initiation protocol phones, wireless local loop (WLL) stations, personal digital assistants (PDAs), tablets (Pads), computers with wireless transceiver functions, and the like. The wireless terminal device can also be referred to as a system, a subscriber unit, a subscriber station, a mobile station, a mobile station (MS), a remote station, an access point (AP), a remote terminal, an access terminal, a user terminal, a user agent, a subscriber station (SS), customer premises equipment (CPE), a terminal, user equipment (UE), a mobile terminal (MT), and the like.

[0164] 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 wearable smart device or a smart wearable device, etc. It is a general term for devices that are designed and developed by applying wearable technology to daily wear. For example, glasses, gloves, watches, clothing, and shoes, etc. The wearable device is a portable device that is directly worn on the body or integrated into the user's clothes or accessories. The wearable device is not only a hardware device, but also a powerful function realized 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 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. for monitoring vital signs.

[0165] The terminal device can also be a drone, a robot, a terminal device in device-to-device (D2D) communication, a terminal device in vehicle to everything (V2X), a virtual reality (VR) terminal device, an augmented reality (AR) terminal device, a wireless terminal device in industrial control, a wireless terminal device in self driving, a wireless terminal device in remote medical treatment, a wireless terminal device in a smart grid, a wireless terminal device in transportation safety, a wireless terminal device in a smart city, or a wireless terminal device in a smart home, etc.

[0166] In addition, the terminal device can also be a terminal device in a future communication system evolved from a 5th generation (5G) communication system or a terminal device in a future evolved public land mobile network (PLMN), etc. For example, the future communication system can further expand the form and function of the 5G communication terminal, and the terminal in the future communication system includes but is not limited to a vehicle, a cellular network terminal (integrating satellite terminal function), a drone, or an internet of things (IoT) device.

[0167] In the embodiments of the present application, the terminal device has artificial intelligence (AI) capability. For example, the terminal device can obtain AI services provided by a network device or a server. The terminal device also has AI processing capability.

[0168] It should be noted that the terminal device can be a device or apparatus with a chip, or a device or apparatus integrated with a circuit, or a chip, module or control unit in the above-mentioned devices or apparatus, and the specific application is not limited.

[0169] The access network device can be a device in a wireless network. For example, the access network device can be an access network node (also referred to as a base station) that accesses a terminal device to a wireless network. Currently, some examples of the access network device are: a base station (gNodeB, gNB) in a 5G communication system, a transmission reception point (TRP), an evolved Node B (eNB), a radio network controller (RNC), a Node B (NB), a home base station (for example, a home evolved Node B, or a home Node B, HNB), a baseband unit (BBU), or a wireless fidelity (Wi-Fi) access point AP, and the like. In addition, in a network structure, the access network device can include a centralized unit (CU) node, a distributed unit (DU) node, a CU-control plane (CP), a CU-user plane (UP), or a radio unit (RU), or a RAN device including the CU node and the DU node. The CU and the DU can be separately arranged, or can be included in the same network element, such as a baseband unit (BBU). The RU can be included in a radio frequency device or a radio frequency unit, such as a remote radio unit (RRU), an active antenna unit (AAU), or a remote radio head (RRH). In different systems, the CU (or CU-CP and CU-UP), DU, or RU can also have different names, but those skilled in the art can understand their meanings. For example, in an open RAN (ORAN) system, the CU can also be referred to as an open CU (O-CU), the DU can also be referred to as an open DU (O-DU), the CU-CP can also be referred to as an open CU-CP (O-CU-CP), the CU-UP can also be referred to as an open CU-UP (O-CU-UP), and the RU can also be referred to as an open RU (O-RU). Any one of the CU (or CU-CP, CU-UP), DU, and RU can be implemented by a software module, a hardware module, or a combination of a software module and a hardware module.

[0170] The access network device can be other apparatuses that provide wireless communication functions for the terminal device. Embodiments of the present application do not limit the specific technology and specific device form adopted by the access network device. For the convenience of description, embodiments of the present application do not limit.

[0171] The core network device includes, 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 (P-GW), a network element such as an access and mobility management function (AMF), a user plane function (UPF), or a session management function (SMF) in a 5G network. In addition, the core network device can also include other core network devices in a 5G network and future networks of the 5G network. Optionally, the core network device is configured to be responsible for the mobile management of the terminal device and the like.

[0172] In embodiments of the present application, the access network device and the core network device can be network nodes with AI capabilities, and can provide AI services for terminal devices or other network devices. For example, the access network device and the core network device can be AI nodes, computing nodes, access network nodes with AI capabilities, or core network elements with AI capabilities on the network side (access network or core network).

[0173] It should be noted that the access network device and the core network device can be the devices or apparatuses shown above, or can be components (such as chips), modules, or units in the devices or apparatuses shown above. The specific embodiments of the present application are not limited.

[0174] In order to meet the vision of smart universal access in the future, intelligentization will further evolve at the wireless network architecture level, AI will be further deeply integrated with wireless networks to achieve network-born intelligence, and the intelligentization of terminal devices is also included.

[0175] The types of terminal devices are diversified, and the connection of terminal devices is more flexible and intelligent. Network-born intelligence refers to that network devices will further provide computing and AI services in addition to providing traditional communication connection services, thereby better supporting AI services that are inclusive, real-time and highly secure. As shown in FIG. 1, data collection is an important part of AI model life cycle management. However, the number of real collection data is limited, for example, time limit and / or data collection overhead limit result in less number of real collection data, and the model performance and / or generalization performance of the model trained by the device based on the real collection data is poor. The present application provides corresponding technical solutions, which can be specifically referred to the related introduction of the embodiments below.

[0176] The technical solutions of the present application will be described below in conjunction with specific embodiments.

[0177] FIG. 3 is a schematic diagram of one embodiment of the data transmission method of the present application. Please refer to FIG. 3, the method comprises:

[0178] 301. The first device generates at least one first simulation data set according to the first data.

[0179] The first data is the local data of the first device, i.e. the real collection data collected by the first device. For example, the first data can include channel information between the first device and the second device.

[0180] Optionally, each first simulation data set in the at least one first simulation data set has a corresponding data generation method. The above step 301 specifically comprises: the first device generates at least one first simulation data set according to the corresponding data generation method of each first simulation data set in the at least one first simulation data set and the first data.

[0181] Two possible implementation manners of the data generation method will be introduced below.

[0182] Implementation manner one: the data generation method comprises: generating a first simulation data set according to the first data and a first model.

[0183] The first model is a 3GPP standard channel model, a pre-trained wireless AI large model, or a generative AI model.

[0184] Optionally, the input parameters of the 3GPP standard channel model comprise at least one of: a channel type between the first device and the second device (e.g., tapped delay line-A (TDL-A), tapped delay line-B (TDL-B), tapped delay line-C (TDL-C), cluster delay line-A (CDL-A), cluster delay line-B (CDL-B), or cluster delay line-C (CDL-C)), location information of the first device, environment information, and a speed of the first device (e.g., 3 km / h or 30 km / h). The output parameters of the 3GPP standard channel model comprise multipath information. Optionally, the multipath information comprises at least one of: a direction of arrival (DOA), a direction of departure (DOD), a power, or a time delay.

[0185] Optionally, the input parameters of the pre-trained wireless AI large model comprise at least one of: environment information of the first device, or location information of the first device. The output parameters of the pre-trained wireless AI large model comprise multipath information. Optionally, the multipath information comprises at least one of: a DOA, a DOD, a power, or a time delay. The pre-trained wireless AI large model can be trained by real data.

[0186] Optionally, the input parameters of the generative AI model comprise at least one of: environment information of the first device, or location information of the first device. The output parameters of the generative AI model comprise multipath information, or channel information. Optionally, the multipath information comprises at least one of: a DOA, a DOD, a power, or a time delay. The channel information comprises at least one of: a channel quality between the first device and the second device. Optionally, the generative AI model is pre-trained.

[0187] Optionally, the environment information comprises a signal-to-noise ratio of a received signal. The location information comprises a coordinate position of the first device.

[0188] Implementation two: the data generation manner comprises: performing data augmentation on the first data to obtain a first simulation data set.

[0189] Optionally, the data augmentation comprises at least one of: repeating data, adding noise to data, adding phase to data, truncating data, or adding offset to data.

[0190] The following describes several possible implementation manners of the first device generating the first simulation data set by taking a first simulation data set as an example.

[0191] Implementation manner 1: The first simulation data set is generated by using a first model. Specifically, the first device extracts first feature information from the first data, and determines second feature information based on the first feature information. For example, the similarity between the first feature information and the second feature information is greater than a first threshold. The first device inputs the second feature information into the first model to obtain the first simulation data set output by the first model.

[0192] Optionally, the first device inputs the second feature information and data generation parameters corresponding to the first simulation data set into the first model to obtain the first simulation data set output by the first model. The data generation parameters corresponding to the first simulation data set include at least one of the following: a simulation position corresponding to the second simulation data set, a simulation speed corresponding to the first simulation data, or a simulation signal-to-noise ratio corresponding to the first simulation data. The simulation position refers to a position simulated by the first device to collect the second simulation data set. The simulation speed refers to a speed simulated by the first device to collect the second simulation data set. The simulation signal-to-noise ratio refers to a signal-to-noise ratio simulated by the first device to collect the first simulation data set.

[0193] Implementation manner 2: The first simulation data set is generated by using a first model. Specifically, the first device inputs part or all of the data in the first data and data generation parameters corresponding to the first simulation data set into the first model as input parameters of the first model to obtain the first simulation data set output by the first model. For the data generation parameters corresponding to the first simulation data set, refer to the related description in the foregoing.

[0194] Implementation manner 3: The first simulation data set is generated by using data enhancement. The first device determines data enhancement parameters corresponding to the first simulation data set. Then, the first device performs data enhancement on the first data by using the data enhancement parameters to obtain the first simulation data set.

[0195] Optionally, different first simulation data sets in the at least one first simulation data set can correspond to the same data generation manner, and the data generation parameters corresponding to different first simulation data sets are different. For example, the at least one first simulation data set includes a first simulation data set 1 and a first simulation data set 2. The first simulation data set 1 and the first simulation data set 2 are both generated by using a first model. The data generation parameters corresponding to the first simulation data set 1 include input parameters 1 of the first model, and the data generation parameters corresponding to the first simulation data set 2 include input parameters 2 of the first model. That is, the input parameters of the first model are different, and therefore the first simulation data sets output by the first model are different.

[0196] Optionally, different first simulation data sets in the at least one first simulation data set correspond to different data generation manners. Different first simulation data sets correspond to different data generation parameters. For example, the at least one first simulation data set includes a first simulation data set 1 and a first simulation data set 2. The first simulation data set 1 is generated in a manner of a first model, and the data generation parameter corresponding to the first simulation data set 1 includes an input parameter of the first model. The first device inputs the data generation parameter corresponding to the first simulation data set 1 and the first data into the first model to obtain the first simulation data set 1. The first simulation data set 2 is generated in a manner of data enhancement, and the data generation parameter corresponding to the first simulation data set 2 includes a data enhancement parameter. The first device performs data enhancement on the first data according to the data generation parameter corresponding to the first simulation data set 2 to obtain the first simulation data set 2.

[0197] Optionally, each first simulation data set in the at least one first simulation data set has a corresponding weight, and the weight of each first simulation data set is used to indicate the importance of the first simulation data set. The higher the weight of the first simulation data set is, the higher the importance of the first simulation data set is. The lower the weight of the first simulation data set is, the lower the importance of the first simulation data set is. The second device performs model training by using the first simulation data set with a higher weight, which is beneficial to improving the model performance. The second device performs model training by using the first simulation data set with a lower weight, which is beneficial to improving the generalization performance of the model. Therefore, for a model with a higher requirement for model performance, the second device can select the first simulation data set with a higher weight to perform model training. For a model with a higher requirement for generalization performance of the model, the second device can select the first simulation data set with a lower weight to perform model training.

[0198] The following introduces some possible determination manners of the weight of each first simulation data set in the at least one first simulation data set.

[0199] I. The weight of each first simulation data set in the at least one first simulation data set is determined according to the similarity between the first simulation data set and the first data.

[0200] For a first simulation data set, the higher the similarity between the first simulation data set and the first data is, the higher the weight of the first simulation data set is. The first simulation data set is more helpful for improving the model performance. For a first simulation data set, the lower the similarity between the first simulation data set and the first data is, the lower the weight of the first simulation data set is. The first simulation data set is more helpful for improving the generalization performance of the model.

[0201] In a possible implementation, the similarity of each first simulation data set in the at least one first simulation data set to the first data is represented by a normalized mean square error (NMSE) or a square generalized cosine similarity (SGCS) of each first simulation data set in the at least one first simulation data to the first data. For example, the greater the NMSE or the SGCS of each first simulation data set to the first data, the higher the similarity of the first simulation data set to the first data.

[0202] Specifically, the first device calculates the NMSE or the SGCS of each first simulation data set to the first data, and thereby determines the similarity of each first simulation data set to the first data.

[0203] In another possible implementation, the similarity of each first simulation data set in the at least one first simulation data set to the first data is represented by a similarity between a first feature and a second feature. The first feature is extracted from each first simulation data set in the at least one first simulation data set. The second feature is extracted from the first data.

[0204] Specifically, the first device extracts the first feature from each first simulation data set in the at least one first simulation data set. The first device extracts the second feature from the first data. The first device calculates the similarity between the first feature and the second feature. The first device determines the similarity of each first simulation data set to the first data by the similarity between the first feature and the second feature.

[0205] II. The weight of each first simulation data set in the at least one first simulation data set is determined according to a data distribution difference of each first simulation data set in the at least one first simulation data set to the first data.

[0206] For a first simulation data set, the greater the data distribution difference of the first simulation data set to the first data, the lower the weight of the first simulation data set. The first simulation data set is more helpful to the improvement of the generalization performance of the model. For a first simulation data set, the lower the data distribution difference of the first simulation data set to the first data, the higher the weight of the first simulation data set. The first simulation data set is more helpful to the improvement of the model performance.

[0207] For example, each of the at least one first simulation dataset has a corresponding data distribution. The first device determines a data distribution difference between the data distribution of each of the at least one first simulation dataset and the data distribution of the first data. The first device takes the data distribution difference between each of the at least one first simulation dataset and the first data as the weight of each of the at least one first simulation dataset. Alternatively, the first device adjusts the data distribution difference between each of the at least one first simulation dataset and the first data to obtain the weight of each of the at least one first simulation dataset.

[0208] III. The weight of each of the at least one first simulation dataset is determined according to the position difference between each of the at least one first simulation dataset and the first data.

[0209] For a first simulation dataset, the higher the position difference between the first simulation dataset and the first data, the lower the weight of the first simulation dataset. The first simulation dataset helps more in improving the generalization performance of the model. For a first simulation dataset, the lower the position difference between the first simulation dataset and the first data, the higher the weight of the first simulation dataset. The first simulation dataset helps more in improving the performance of the model.

[0210] For example, each of the at least one first simulation dataset has a corresponding simulation position. The simulation position refers to the position where the simulation is collected to obtain the first simulation dataset. The first device determines a position difference between the simulation position of each of the at least one first simulation dataset and the position of the first data. The first device takes the position difference between the simulation position of each of the at least one first simulation dataset and the position of the first data as the weight of each of the at least one first simulation dataset. Alternatively, the first device adjusts the position difference between the simulation position of each of the at least one first simulation dataset and the position of the first data to obtain the weight of each of the at least one first simulation dataset.

[0211] IV. The weight of each of the at least one first simulation dataset is determined according to the signal-to-noise ratio difference between each of the at least one first simulation dataset and the first data.

[0212] For a first simulation dataset, the higher the signal-to-noise ratio difference between the first simulation dataset and the first data, the lower the weight of the first simulation dataset. The first simulation dataset helps more in improving the generalization performance of the model. For a first simulation dataset, the lower the signal-to-noise ratio difference between the first simulation dataset and the first data, the higher the weight of the first simulation dataset. The first simulation dataset helps more in improving the performance of the model.

[0213] For example, each of the at least one first simulation data set has a corresponding simulation signal-to-noise ratio. The simulation signal-to-noise ratio refers to a signal-to-noise ratio used by the simulation to collect the first simulation data set. The first device determines a signal-to-noise ratio difference between the simulation signal-to-noise ratio of each of the at least one first simulation data set and the signal-to-noise ratio of the first data. The first device takes the signal-to-noise ratio difference between the simulation signal-to-noise ratio of each of the at least one first simulation data set and the signal-to-noise ratio of the first data as the weight of each of the at least one first simulation data set. Alternatively, the first device adjusts the signal-to-noise ratio difference between the simulation signal-to-noise ratio of each of the at least one first simulation data set and the signal-to-noise ratio of the first data to obtain the weight of each of the at least one first simulation data set.

[0214] Optionally, the embodiment shown in FIG. 3 further includes step 301a. Step 301a can be performed before step 301.

[0215] 301a. The second device sends the expected weight of each of the at least one first simulation data set to the first device. Correspondingly, the first device receives the expected weight of each of the at least one first simulation data set from the second device.

[0216] The expected weight of each of the at least one first simulation data set refers to the expected, desired, or intended weight of each of the at least one first simulation data set. The expected weight can also be referred to as a target weight, or a desired weight, etc., which is not limited in the present application.

[0217] Based on the above step 301a, the above step 301 specifically includes: the first device generates the at least one first simulation data set according to the first data and the expected weight of each of the at least one first simulation data set.

[0218] For example, for a first simulation dataset, the first simulation dataset is generated in a manner of a first model. Specifically, the first device extracts first feature information from the first data, and determines second feature information based on the first feature information and an expected weight of the first simulation dataset. For example, the greater the expected weight of the first simulation dataset, the higher the similarity between the first feature information and the second feature information. The smaller the expected weight of the first simulation dataset, the lower the similarity between the first feature information and the second feature information. The first device inputs the second feature information into the first model to obtain the first simulation dataset output by the first model. Optionally, the first device determines a data generation parameter corresponding to the first simulation dataset according to the expected weight of the first simulation dataset. The data generation parameter corresponding to the first simulation dataset includes at least one of the following: a simulation position corresponding to the first simulation dataset, a simulation speed corresponding to the second simulation data, or a simulation signal-to-noise ratio corresponding to the second simulation data. For example, the greater the expected weight of the first simulation dataset, the closer the simulation speed corresponding to the first simulation dataset to the speed when the first device collects the first data. The smaller the expected weight of the first simulation dataset, the farther the simulation speed corresponding to the first simulation dataset from the speed when the terminal device collects the first data. That is, the second device sets the data generation parameter corresponding to the first simulation dataset according to the expected weight of the first simulation dataset. The importance of the first simulation dataset is realized by the data generation parameter corresponding to the first simulation dataset. Then, the first device inputs the second feature information and the data generation parameter corresponding to the first simulation dataset into the first model to obtain the first simulation dataset output by the first model.

[0219] For example, for a first simulation dataset, the first simulation dataset is generated in a manner of a first model. Specifically, the first device extracts first feature information from the first data, and determines second feature information based on the first feature information and an expected weight of the first simulation dataset. For example, the greater the expected weight of the first simulation dataset, the higher the similarity between the first feature information and the second feature information. The smaller the expected weight of the first simulation dataset, the lower the similarity between the first feature information and the second feature information. The first device inputs the second feature information into the first model to obtain the first simulation dataset output by the first model. Optionally, the first device determines a data generation parameter corresponding to the first simulation dataset according to the expected weight of the first simulation dataset. The data generation parameter corresponding to the first simulation dataset includes at least one of the following: a simulation position corresponding to the first simulation dataset, a simulation speed corresponding to the second simulation data, or a simulation signal-to-noise ratio corresponding to the second simulation data. For example, the greater the expected weight of the first simulation dataset, the closer the simulation speed corresponding to the first simulation dataset to the speed when the first device collects the first data. The smaller the expected weight of the first simulation dataset, the farther the simulation speed corresponding to the first simulation dataset from the speed when the terminal device collects the first data. That is, the second device sets the data generation parameter corresponding to the first simulation dataset according to the expected weight of the first simulation dataset. The importance of the first simulation dataset is realized by the data generation parameter corresponding to the first simulation dataset. Then, the first device inputs the second feature information and the data generation parameter corresponding to the first simulation dataset into the first model to obtain the first simulation dataset output by the first model.

[0220] For example, for a first simulation dataset, the first simulation dataset is generated in a manner of a first model. Specifically, the first device extracts first feature information from the first data, and determines second feature information based on the first feature information and an expected weight of the first simulation dataset. For example, the greater the expected weight of the first simulation dataset, the higher the similarity between the first feature information and the second feature information. The smaller the expected weight of the first simulation dataset, the lower the similarity between the first feature information and the second feature information. The first device inputs the second feature information into the first model to obtain the first simulation dataset output by the first model. Optionally, the first device determines a data generation parameter corresponding to the first simulation dataset according to the expected weight of the first simulation dataset. The data generation parameter corresponding to the first simulation dataset includes at least one of the following: a simulation position corresponding to the first simulation dataset, a simulation speed corresponding to the second simulation data, or a simulation signal-to-noise ratio corresponding to the second simulation data. For example, the greater the expected weight of the first simulation dataset, the closer the simulation speed corresponding to the first simulation dataset to the speed when the first device collects the first data. The smaller the expected weight of the first simulation dataset, the farther the simulation speed corresponding to the first simulation dataset from the speed when the terminal device collects the first data. That is, the second device sets the data generation parameter corresponding to the first simulation dataset according to the expected weight of the first simulation dataset. The importance of the first simulation dataset is realized by the data generation parameter corresponding to the first simulation dataset. Then, the first device inputs the second feature information and the data generation parameter corresponding to the first simulation dataset into the first model to obtain the first simulation dataset output by the first model.

[0221] Optionally, the embodiment shown in FIG. 3 further includes step 301b. Step 301b can be performed before step 301.

[0222] 301b. The second device sends the first weight to the first device. Correspondingly, the first device receives the first weight from the second device.

[0223] The first weight is used to indicate the importance of the data of the first device. For example, the importance of the first device for model training is high, so the first weight can be high.

[0224] It should be noted that the second device can determine the first weight according to the data importance of the first device. The data importance of the first device can be determined according to at least one of the importance of the data of the specific region, the influence degree of the historical reported data of the first device on model training, and the capability information of the first device. The importance of the data of the specific region refers to the importance of the data of the users in the specific region. Generally, the importance of the data of the users in the specific region is high. The data of the users in the specific region includes the data of the central users, the data of the edge users, and the data of the users in the scene of user aggregation (e.g., concert, stadium), etc. For example, the data of the first device is the data of the users in the specific region, so the data importance of the first device is high. For another example, the influence degree of the first device on model training is high, so the data importance of the first device is high.

[0225] Based on the above step 301b, optionally, the above step 301 specifically includes: the first device generates at least one first simulation data set according to the first data and the first weight.

[0226] For example, for a first simulation data set, the first simulation data set is generated in a manner of a first model. Specifically, the first device extracts first feature information from the first data, and determines second feature information based on the first feature information and a first weight. For example, the greater the first weight, the higher the similarity between the first feature information and the second feature information. The smaller the first weight, the lower the similarity between the first feature information and the second feature information. The first device inputs the second feature information into the first model to obtain the first simulation data set output by the first model. Optionally, the first device determines data generation parameters corresponding to the first simulation data set according to the first weight. The data generation parameters corresponding to the first simulation data set include at least one of the following: a simulation position corresponding to the first simulation data set, a simulation speed corresponding to the second simulation data, or a simulation signal-to-noise ratio corresponding to the second simulation data. For example, the greater the expected weight of the first simulation data set, the closer the simulation speed corresponding to the first simulation data set to the speed when the first device collects the first data. The smaller the expected weight of the first simulation data set, the farther the simulation speed corresponding to the first simulation data set from the speed when the terminal device collects the first data. That is, the first device sets the data generation parameters corresponding to the first simulation data set according to the first weight. The importance of the first simulation data set is realized by the data generation parameters corresponding to the first simulation data set. Then, the first device inputs the second feature information and the data generation parameters corresponding to the first simulation data set into the first model to obtain the first simulation data set output by the first model.

[0227] If the embodiment shown in FIG. 3 includes step 301a and step 301b, optionally, the above-mentioned step 301 specifically includes: the first device generates at least one first simulation data set according to the first weight, the expected weight of the at least one first simulation data set, and the first data.

[0228] For example, the first device determines the data generation parameters corresponding to each first simulation data set in combination with the first weight and the expected weight of the at least one first simulation data set. Then, the first device generates at least one first simulation data set according to the first data and the data generation parameters corresponding to each first simulation data set.

[0229] In this implementation, there is no fixed execution order between step 301a and step 301b. Step 301a can be executed first, and then step 301b can be executed. Alternatively, step 301b can be executed first, and then step 301a can be executed. Alternatively, step 301a and step 301b can be executed simultaneously according to the situation, and the specific application is not limited.

[0230] 302、The first device sends the first data, at least one first simulation data set, and first indication information to the second device. Correspondingly, the second device receives the first data, at least one first simulation data set, and first indication information from the first device.

[0231] The first indication information is used to indicate the weight of each first simulation data set in the at least one first simulation data set. Two possible implementation manners of the first indication information indicating the weight of each first simulation data are introduced below.

[0232] Implementation manner 1: The first indication information includes the weight of each first simulation data set in the at least one first simulation data set.

[0233] Implementation manner 2: The first indication information includes the identification of each first simulation data set in the at least one first simulation data set. The identification of each first simulation data set is used to indicate the weight of each first simulation data set. For example, as shown in FIG. 4, the at least one first simulation data set includes first simulation data set 1, first simulation data set 2, and first simulation data set 3. The identification of the first simulation data set 1 is 11, the identification of the first simulation data set 2 is 10, and the identification of the first simulation data set 3 is 01. In a possible implementation manner, the mapping relationship between the identification of the first simulation data set and the weight of the first simulation data set is shown in Table 1:

[0234] Table 1

[0235] As shown in Table 1, the identification of the first simulation data set 1 indicates that the weight of the first simulation data set 1 is 0.8, the identification of the first simulation data set 2 indicates that the weight of the first simulation data set 2 is 0.6, and the identification of the first simulation data set 3 indicates that the weight of the first simulation data set 3 is 0.4.

[0236] In another possible implementation manner, the mapping relationship between the identification of the first simulation data set and the range to which the weight of the first simulation data set belongs is shown in Table 2:

[0237] Table 2

[0238] As shown in Table 2, the identification of the first simulation data set 1 indicates that the weight of the first simulation data set 1 belongs to the interval range of 0.8 to 1. The identification of the first simulation data set 2 indicates that the weight of the first simulation data set 2 belongs to the interval range of 0.6 to 0.8. The identification of the first simulation data set 3 indicates that the weight of the first simulation data set 3 belongs to the interval range of 0.4 to 0.6. Thus, the weight of each first simulation data set is indirectly indicated.

[0239] 303. The second device performs model training according to the first data, the at least one first simulation data set, and the weight of each first simulation data set in the at least one first simulation data set, to obtain one or more models.

[0240] Optionally, the one or more models include a second model. The step 302 includes steps 1 and 2.

[0241] Step 1: The second device selects target data from the first data and at least one first simulation data set according to at least one of a model performance requirement and a generalization performance requirement of the second model and a weight of each first simulation data set in the at least one first simulation data set.

[0242] Step 2: The second device performs model training on the target data to obtain the second model.

[0243] For example, as shown in FIG. 5, the second model is model 1, the model performance requirement of model 1 is high, and the generalization performance requirement of model 1 is low. The second device can select only the first data to perform model training to obtain model 1.

[0244] For another example, as shown in FIG. 5, the second model is model 2, the model performance requirement of model 2 is high, and the generalization performance requirement of model 2 is medium. The second device can select the first data and the first simulation data set 1 to perform model training to obtain model 2.

[0245] For another example, as shown in FIG. 5, the second model is model 3, the model performance requirement of model 3 is high, and the generalization performance requirement of model 3 is high. The second device can select the first data, the first simulation data set 1, and the first simulation data set 2 to perform model training to obtain model 3.

[0246] Optionally, the embodiment shown in FIG. 3 further includes step 302a.

[0247] 302a. The first device sends the weight of the first data to the second device. Correspondingly, the second device receives the weight of the first data from the first device.

[0248] The weight of the first data is used to indicate the importance of the first data. In this implementation, for the first device, each piece of local data of the first device has a corresponding weight. When the first device provides multiple pieces of local data to the second device, the second device can select corresponding local data for model training in combination with the weight of each piece of local data. For example, local data 1 collected by the first device in region 1 is of high importance for model training, and therefore the weight of local data 1 is high. Local data 2 collected by the first device in region 2 is of low importance for model training, and therefore the weight of local data 2 is low.

[0249] It should be noted that step 302a and steps 301 to 302 do not have a fixed execution order. Step 302a can be executed first, and then steps 301 to 302 can be executed. Alternatively, steps 301 to 302 can be executed first, and then step 302a can be executed. Alternatively, steps 302a and steps 301 to 302 can be executed simultaneously according to circumstances, and the specific application is not limited.

[0250] Based on the step 302a, the step 303 can further include: training the model according to the first data, the at least one first simulation data set, the weight of the first data, and the weight of each first simulation data set in the at least one first simulation data set, to obtain one or more models.

[0251] For example, the weight of the first data is high. The one or more models include a second model, and the model performance requirement of the second model is high, and the generalization performance requirement is medium. The second device can select a corresponding first simulation data set from the at least one first simulation data set, and can take the first data and the selected first simulation data set as target data, and train the model through the target data to obtain the second model.

[0252] Optionally, the embodiment shown in FIG. 3 further includes steps 303a to 303b.

[0253] 303a. The third device generates at least one second simulation data set according to the second data.

[0254] The step 303a is similar to the step 301, and details are referable to the description of the step 301.

[0255] Optionally, the embodiment shown in FIG. 3 further includes a step 303c, which can be performed before the step 303a.

[0256] 303c. The second device sends the expected weight of each second simulation data set in the at least one second simulation data set to the third device. Correspondingly, the third device receives the expected weight of each second simulation data set in the at least one second simulation data set from the second device.

[0257] The step 303c is similar to the step 301a, and details are referable to the description of the step 303c.

[0258] Based on the step 303c, the third device can generate at least one second simulation data set according to the second data and the expected weight of each second simulation data set in the at least one second simulation data set.

[0259] Optionally, the embodiment shown in FIG. 3 further includes a step 303d, which can be performed before the step 303a.

[0260] 303d. The second device sends the second weight to the third device. Correspondingly, the third device receives the second weight from the second device.

[0261] The second weight is used to indicate the importance of the data of the third device. The second weight is similar to the first weight, and details are referable to the description of the first weight.

[0262] Step 303d is similar to step 301b, and details can be referred to the foregoing description of step 301b, which will not be repeated here.

[0263] Based on step 303d, the third device can generate at least one second simulation data set according to the second data and the second weight.

[0264] If the embodiment shown in FIG. 3 further includes step 303c and step 303d, step 303a specifically includes: the third device generates at least one second simulation data set according to the second data, the second weight, and the expected weight of each second simulation data set in the at least one second simulation data set. In this implementation manner, there is no fixed execution order between step 303c and step 303d. Step 303c can be executed first, and then step 303d is executed. Alternatively, step 303d can be executed first, and then step 303c is executed. Alternatively, step 303c and step 303d can be executed simultaneously according to the situation, and the specific implementation is not limited in the present application.

[0265] 303b. The third device sends the second data, the at least one second simulation data set, and the second indication information to the second device. Correspondingly, the second device receives the second data, the at least one second simulation data set, and the second indication information from the third device.

[0266] The second data is local data of the third device, that is, real data collected by the third device locally. The second indication information is used to indicate the weight of each second simulation data set in the at least one second simulation data set. The indication manner of the weight of each second simulation data set indicated by the second indication information is similar to the indication manner of the weight of each first simulation data set indicated by the first indication information in step 302, and details can be referred to the foregoing description of the first indication information, which will not be repeated here.

[0267] It should be noted that there is no fixed execution order between step 301b to step 302a and step 303a to step 303d. Step 301b to step 302a can be executed first, and then step 303a to step 303d is executed. Alternatively, step 303a to step 303d can be executed first, and then step 301b to step 302a is executed. Alternatively, step 301b to step 302a and step 303a to step 303d can be executed simultaneously according to the situation, and the specific implementation is not limited in the present application.

[0268] Based on the step 303b, the step 303 can further include that the second device trains the model according to the first data, the at least one first simulation data set, the weight of each first simulation data set in the at least one first simulation data set, the second data, the at least one second simulation data set, and the weight of each second simulation data set in the at least one second simulation data set, to obtain one or more models.

[0269] In a possible implementation, the second device combines the at least one first simulation data set and the at least one second simulation data set with the same or similar weight to obtain one or more target simulation data sets. Each target simulation data set has a corresponding weight. Then, the second device trains the model according to the first data, the second data, the one or more target simulation data sets, and the weight of each target simulation data set in the one or more target simulation data sets, to obtain one or more models. For example, the one or more target simulation data sets include a target simulation data set 1 and a target simulation data set 2. The weight of the target simulation data set 1 is 0.8, and the weight of the target simulation data set 2 is 0.6. The one or more models include a second model, and the model performance requirement of the second model is high and the generalization performance requirement is medium. The second device can select the target simulation data set 1, and train the model according to the first data, the second data, and the target simulation data set 1 to obtain the second model.

[0270] Optionally, the step 303 can further include that the second device trains the model according to the first data, the at least one first simulation data set, the weight of each first simulation data set in the at least one first simulation data set, the first weight, the second data, the at least one second simulation data set, the weight of each second simulation data set in the at least one second simulation data set, and the second weight, to obtain one or more models.

[0271] In a possible implementation, the second device determines a target data set according to the first weight, the first data, the at least one first simulation data set, the weight of each first simulation data set in the at least one first simulation data set, the second weight, the second data, the at least one second simulation data set, and the weight of each second simulation data set in the at least one second simulation data set. Then, the second device trains the model by using the target data set to obtain one or more models.

[0272] For example, the one or more models comprise a second model. The second device performs data selection from the at least one first simulation dataset and the first data according to at least one of a model performance requirement and a generalization performance requirement of the second model and a weight of each first simulation dataset in the at least one first simulation dataset, to obtain a dataset A. The second device performs data selection from the at least one second simulation dataset and the second data according to at least one of a model performance requirement and a generalization performance requirement of the second model and a weight of each second simulation dataset in the at least one second simulation dataset, to obtain a dataset B. Then, the second device determines a target data set according to the first weight, the second weight, the dataset A and the dataset B. For example, the target data set Set C = SetA * a + SetB * b. Wherein, SetA is the dataset A, SetB is the dataset B, a is the first weight, and b is the second weight. It should be noted that, optionally, the dataset A multiplied by the first weight can represent selecting a certain amount of data from the dataset A. For example, the first weight is 0.6, then the second device can select 60% of the data from the dataset A. The dataset B multiplied by the second weight can represent selecting a certain amount of data from the dataset B. For example, the second weight is 0.8, then the second device can select 80% of the data from the dataset B. That is, the second device represents the data importance of each device by the amount of selected data.

[0273] In the embodiments of this application, the first device generates at least one first simulation data set according to the first data, and the first data is local data of the first device. Each first simulation data set in the at least one first simulation data set has a corresponding weight, and the weight of each first simulation data set is used to indicate the importance of the first simulation data set. Then, the first device sends the first data, the at least one first simulation data set and the first indication information to the second device. The first indication information is used to indicate the weight of each first simulation data set in the at least one first simulation data set. Therefore, the first device generates at least one first simulation data set according to the first data, and sends the first data, the at least one first simulation data set and the first indication information to the second device. Therefore, the first device provides more data to the second device, that is, the first device provides not only the first data but also the at least one first simulation data set. The second device performs model training based on the first data and the at least one first simulation data set. Therefore, the amount of data used for model training is increased, which is beneficial to improve the model performance and / or generalization performance of the trained model. Further, the first indication information is used to indicate the weight of each first simulation data set in the at least one first simulation data set, and the weight of each first simulation data set is used to indicate the importance of the first simulation data set. This is beneficial to the second device to select appropriate data for corresponding model training based on the weight of each first simulation data set, so as to meet the needs of different models. For example, for a model with high model performance requirement, the second device can preferentially select a first simulation data set with a larger weight for model training. For a model with high generalization performance requirement, the second device can preferentially select a first simulation data set with a smaller weight for model training.

[0274] FIG. 6 is another embodiment of the data transmission method of the present application. Please refer to FIG. 6, the method comprises:

[0275] 601, the first device determines the first data and the first indication information.

[0276] The first data is local data of the first device, that is, the actual data collected by the first device.

[0277] The first indication information is used to indicate at least one first generation configuration. The at least one first generation configuration is used to generate at least one first simulation dataset based on the first data. For example, each of the at least one first generation configuration corresponds to one or more first simulation datasets. Each of the at least one first generation configuration is used to assist in generating information or parameters of the first simulation dataset corresponding to the first generation configuration. For example, one of the at least one first simulation dataset is generated in a manner of a first model, and the first generation configuration corresponding to the first simulation dataset can include a reference input parameter of the first model. For details of the first model, please refer to the related description in the foregoing embodiment shown in FIG. 3. For another example, one of the at least one first simulation dataset is generated in a data generation manner of data enhancement, and the first generation configuration corresponding to the first simulation dataset can include a reference data enhancement parameter. Hereinafter, the technical solution of the present application is mainly introduced by taking an example of each of the first generation configuration corresponding to one first simulation dataset and different first generation configurations corresponding to different first simulation datasets.

[0278] Two possible implementation manners of the first indication information indicating the at least one first generation configuration are introduced below.

[0279] Implementation manner one: the first indication information includes the at least one first generation configuration.

[0280] Implementation manner two: the first indication information includes an identifier or an index of each of the at least one first generation configuration.

[0281] In this implementation manner, the at least one first generation configuration can be preconfigured in the first device and the second device. Then, the first device can indicate the at least one first generation configuration to the second device through the first indication information. Thus, the indication overhead is reduced.

[0282] Optionally, the embodiment shown in FIG. 6 further includes steps 601a to 601b. The steps 601a to 601b can be executed before the step 601.

[0283] 601a. The first device sends second indication information to the second device. Correspondingly, the second device receives the second indication information from the first device.

[0284] The second indication information is used to indicate a data generation manner of each of the at least one first simulation dataset. For details of the data generation manner, please refer to the related description in the foregoing embodiment shown in FIG. 3.

[0285] 601b. The second device sends first confirmation information to the first device. Correspondingly, the first device receives the first confirmation information from the second device.

[0286] The first confirmation information is used to confirm the data generation manner corresponding to each first simulation data set in the at least one first simulation data set. Thus, the first device and the second device align the data generation manner corresponding to each first simulation data set.

[0287] Optionally, the at least one first generation configuration in the step 601 is generated according to the data generation manner corresponding to each first simulation data set in the at least one first simulation data set. For example, for a first simulation data set, the first simulation data set is generated by using a first model, and the first generation configuration corresponding to the first simulation data set can include a reference input parameter of the first model. For another example, for a first simulation data set, the first simulation data set is generated by using data enhancement, and the first generation configuration corresponding to the first simulation data set can include a reference data enhancement parameter.

[0288] Optionally, the embodiment shown in FIG. 6 further includes a step 601c. The step 601c can be performed before the step 601.

[0289] 601c. The second device sends, to the first device, an expected weight of each first simulation data set in the at least one first simulation data set. Correspondingly, the first device receives the expected weight of each first simulation data set in the at least one first simulation data set from the second device.

[0290] The expected weight of each first simulation data set in the at least one first simulation data set is used to indicate an expected, desired, or intended importance of the first simulation data set. The expected weight can also be referred to as a target weight, a desired weight, or the like, which is not limited in the present application.

[0291] Optionally, the at least one first generation configuration is determined according to an expected weight of each first simulation data set in the at least one first simulation data set. For example, the at least one first simulation data set includes a first simulation data set 1 and a first simulation data set 2. The first simulation data set 1 and the first simulation data set 2 are both generated in a manner of the first model. The first device is a terminal device, and the input parameters of the first model include a simulation speed and a simulation position of the terminal device. The expected weight of the first simulation data set 1 is 0.8, and the expected weight of the first simulation data set 2 is 0.4. The first generation configuration 1 corresponding to the first simulation data set 1 includes a simulation speed 1 and a simulation position 1 of the terminal device. The simulation speed 1 is close to the speed of the terminal device when collecting the first data, and the simulation position 1 is close to the position of the terminal device when collecting the first data. The first generation configuration 2 corresponding to the first simulation data set 2 includes a simulation speed 2 and a simulation position 2. The simulation speed 2 is far greater than the speed of the terminal device when collecting the first data, and the simulation position 2 is far away from the position of the terminal device when collecting the first data. Thus, the first generation configuration corresponding to the first simulation data set is set through the expected weight of the first simulation data set, and the importance of the first simulation data set is affected through the first generation configuration corresponding to the first simulation data set.

[0292] It should be noted that, if the embodiment shown in FIG. 6 includes steps 601a to 601b and step 601c, the at least one first generation configuration is determined according to the data generation manner of each first simulation data set in the at least one first simulation data set and the expected weight of the each first simulation data set.

[0293] In this implementation manner, there is no fixed execution order between steps 601a to 601b and step 601c. Steps 601a to 601b can be executed first, and then step 601c is executed; or step 601c is executed first, and then steps 601a to 601b are executed; or steps 601a to 601b and step 601c are executed at the same time according to the situation, which is not limited in the present application.

[0294] Optionally, the embodiment shown in FIG. 6 further includes step 601d. Step 601d can be executed before step 601.

[0295] 601d, the second device sends the first weight to the first device. Correspondingly, the first device receives the first weight from the second device.

[0296] The first weight is used to indicate the importance of the data of the first device. Optionally, the at least one first generation configuration is determined according to the first weight. For example, the first device is a terminal device, and the first weight is relatively large, indicating that the terminal device has high importance of data. The at least one first simulation data set is generated by using a first model. The at least one first generation configuration includes an input parameter of the first model. For example, the input parameter of the first model can include a simulation speed of the terminal device, a simulation position, and the like. The simulation speed can be close to the speed of the terminal device when collecting the first data. The simulation position can be close to the position of the terminal device when using the first data. Thus, the importance of the at least one first simulation data set is affected by the at least one first generation configuration and the first weight.

[0297] For the determination of the first weight, refer to the related description in step 301b in the embodiment shown in FIG. 3, which will not be repeated here.

[0298] Optionally, if the embodiment shown in FIG. 6 includes steps 601a to 601b and step 601d, the at least one first generation configuration is determined according to the data generation manner of each first simulation data set in the at least one first simulation data set and the first weight.

[0299] Optionally, if the embodiment shown in FIG. 6 includes step 601c and step 601d, the at least one first generation configuration is determined according to the first weight and the expected weight of each first simulation data set in the at least one first simulation data set.

[0300] Optionally, if the embodiment shown in FIG. 6 includes steps 601a to 601b, step 601c, and step 601d, the at least one first generation configuration is determined according to the first weight, the expected weight of each first simulation data set in the at least one first simulation data set, and the data generation manner of each first simulation data set in the at least one first simulation data set.

[0301] 602. The first device sends the first data and the first indication information to the second device. Correspondingly, the second device receives the first data and the first indication information from the first device.

[0302] 603. The second device generates at least one first simulation data set according to the first data and the at least one first generation configuration.

[0303] Optionally, each first simulation data set in the at least one first simulation data set has a corresponding data generation manner. The second device generates each first simulation data set according to the first data, the corresponding data generation manner of each first simulation data set, and the corresponding first generation configuration of each first simulation data set.

[0304] For example, for a first simulation data set, the first simulation data set is generated in a manner of a first model. The second device determines data generation parameters corresponding to the first simulation data set according to the first data and a first generation configuration corresponding to the first simulation data set. For example, the second device extracts first feature information from the first data, and takes the first feature information and information included in the first generation configuration as the data generation parameters corresponding to the first simulation data set. Alternatively, the second device takes part or all of the data in the first data and information included in the first generation configuration as the data generation parameters corresponding to the first simulation data set. Then, the second device inputs the data generation parameters corresponding to the first simulation data set into the first model, and obtains the first simulation data set output by the first model. For example, the first generation configuration corresponding to the first simulation data set includes at least one of the following: a simulation speed corresponding to the first simulation data set, or a simulation position.

[0305] For another example, for a first simulation data set, the first simulation data set is generated in a manner of data augmentation. The first generation configuration corresponding to the first simulation data set includes data augmentation parameters. The second device performs data augmentation on the first data through the data augmentation parameters, and obtains the first simulation data set.

[0306] Optionally, each of the at least one first simulation data set has a corresponding weight. The weight of each of the at least one first simulation data set is used to indicate the importance of each of the at least one first simulation data set. For some possible ways of determining the weight of each of the at least one first simulation data set, please refer to the related description in the foregoing embodiment shown in FIG. 3, which will not be repeated here.

[0307] Optionally, the embodiment shown in FIG. 6 further includes step 604. Step 604 can be performed after step 603.

[0308] 604. The second device performs model training according to the first data, the at least one first simulation data set, and the weight of each of the at least one first simulation data set, and obtains one or more models.

[0309] Step 604 is similar to step 303 in the foregoing embodiment shown in FIG. 3, and for details, please refer to the related description of step 303 in the foregoing embodiment shown in FIG. 3, which will not be repeated here.

[0310] Optionally, the embodiment shown in FIG. 6 further includes steps 603a to 603c.

[0311] 603a. The third device determines the second data and third indication information.

[0312] The second data is local data of the second device, i.e., real data collected by the second device. The third indication information is used for indicating at least one second generation configuration. The at least one second generation configuration is used for generating at least one second simulation data set. The at least one second generation configuration is similar to the at least one first generation configuration, and details can be referred to the foregoing description of the at least one first generation configuration, which will not be repeated here.

[0313] Optionally, the third indication information includes the at least one second generation configuration, or the third indication information includes an identifier or an index of the at least one second generation configuration.

[0314] Optionally, the embodiment shown in FIG. 6 further includes steps 603d to 603e. The steps 603d to 603e can be performed before the step 603a.

[0315] 603d. The third device sends fourth indication information to the second device. Correspondingly, the second device receives the fourth indication information from the third device.

[0316] The fourth indication information is used for indicating a data generation manner corresponding to each second simulation data set in the at least one second simulation data set. The data generation manner can be referred to the foregoing description of the embodiment shown in FIG. 3.

[0317] 603e. The second device sends second confirmation information to the third device. Correspondingly, the third device receives the second confirmation information from the second device.

[0318] The second confirmation information is used for confirming the data generation manner corresponding to each second simulation data set in the at least one second simulation data set. Thus, the second device and the third device align the data generation manner corresponding to each second simulation data set.

[0319] Optionally, the at least one second generation configuration is generated according to the data generation manner corresponding to each second simulation data set in the at least one second simulation data set. For example, for a second simulation data set, the second simulation data set is generated by using a first model, and the second generation configuration corresponding to the second simulation data set can include a reference input parameter of the first model. For another example, for a second simulation data set, the second simulation data set is generated by using data enhancement, and the second generation configuration corresponding to the second simulation data set can include a reference data enhancement parameter.

[0320] Optionally, the embodiment shown in FIG. 6 further includes a step 603f. The step 603f can be performed before the step 603a.

[0321] 603f, the second device sends, to the third device, an expected weight of each second simulation data set in the at least one second simulation data set. Correspondingly, the third device receives the expected weight of each second simulation data set in the at least one second simulation data set from the second device.

[0322] The expected weight of each second simulation data set in the at least one second simulation data set is used to indicate an importance of the expected, desired, or intended each second simulation data set. The expected weight can also be referred to as a target weight, or a desired weight, etc., which is not limited in the present application.

[0323] Optionally, the at least one second generation configuration is determined according to the expected weight of each second simulation data set in the at least one second simulation data set. The related examples of this implementation mode can be referred to the related examples in the aforementioned step 601c.

[0324] It should be noted that, if the embodiment shown in FIG. 6 includes the steps 603d to 603e and the step 603f, the at least one second generation configuration is determined according to the data generation mode of each second simulation data set in the at least one second simulation data set and the expected weight of the each second simulation data set.

[0325] In this implementation mode, there is no fixed execution order between the steps 603d to 603e and the step 603f. The step 603d to 603e can be executed first, and then the step 603f is executed; or the step 603f is executed first, and then the step 603d to 603e is executed; or the step 603d to 603e and the step 603f are executed at the same time according to the situation, which is not limited in the present application.

[0326] Optionally, the embodiment shown in FIG. 6 further includes the step 603g. The step 603g can be executed before the step 603a.

[0327] 603g, the second device sends, to the third device, a second weight. Correspondingly, the third device receives the second weight from the second device.

[0328] The second weight is used to indicate an importance of data of the second device. Optionally, the at least one second generation configuration is determined according to the second weight. The step 603g is similar to the step 601d, and the related introduction of the step 601d can be referred to.

[0329] The determination mode of the second weight is similar to the determination mode of the first weight, and the related introduction of the step 301b in the embodiment shown in FIG. 3 can be referred to, which is not described herein again.

[0330] Optionally, if the embodiment shown in FIG. 6 includes steps 603d-603e and step 603g, the at least one second generation configuration is determined according to the data generation manner of each second simulation data set in the at least one second simulation data set and the second weight. In this implementation manner, there is no fixed execution order between steps 603d-603e and step 603g. Steps 603d-603e can be executed first, and then step 603g is executed; or step 603g is executed first, and then steps 603d-603e are executed; or steps 603d-603e and step 603g are executed simultaneously according to the situation, which is not limited in the present application.

[0331] Optionally, if the embodiment shown in FIG. 6 includes step 603f and step 603g, the at least one second generation configuration is determined according to the second weight and the expected weight of each second simulation data set in the at least one second simulation data set. In this implementation manner, there is no fixed execution order between step 603f and step 603g. Step 603f can be executed first, and then step 603g is executed; or step 603g is executed first, and then step 603f is executed; or steps 603f and step 603g are executed simultaneously according to the situation, which is not limited in the present application.

[0332] Optionally, if the embodiment shown in FIG. 6 includes steps 603d-603e, step 603f and step 603g, the at least one second generation configuration is determined according to the data generation manner of each second simulation data set in the at least one second simulation data set, the second weight and the expected weight of each second simulation data set in the at least one second simulation data set. In this implementation manner, there is no fixed execution order between steps 603d-603e, step 603f and step 603g, which is not limited in the present application. For example, steps 603d-603e are executed first, then step 603f is executed, and finally step 603g is executed.

[0333] 603b, the third device sends second data and third indication information to the second device. Correspondingly, the second device receives the second data and the third indication information from the third device.

[0334] 603c, the second device generates at least one second simulation data set according to the second data and the at least one second generation configuration.

[0335] Step 603c is similar to the aforementioned step 603, and details can be referred to the related description of the aforementioned step 603, which is not described here.

[0336] The steps 603a to 603c do not have a fixed execution order with the aforementioned steps 601 to 603. The steps 601 to 603 can be executed first, and then the steps 603a to 603c are executed; or the steps 603a to 603c are executed first, and then the steps 601 to 603 are executed; or the steps 603a to 603c and the steps 601 to 603 are executed simultaneously according to the situation, and the specific application is not limited.

[0337] Optionally, the step 604 specifically includes that the second device performs model training according to the first data, the at least one first simulation data set, the weight of each first simulation data set in the at least one first simulation data set, the second data, the at least one second simulation data set, and the weight of each second simulation data set in the at least one second simulation data set, to obtain one or more models. This implementation manner can refer to the related introduction in the step 303b in the embodiment shown in FIG. 3, and details are not described herein again.

[0338] Optionally, the step 604 specifically includes that the second device performs model training according to the first data, the at least one first simulation data set, the weight of each first simulation data set in the at least one first simulation data set, the first weight, the second data, the at least one second simulation data set, the weight of each second simulation data set in the at least one second simulation data set, and the second weight, to obtain one or more models. This implementation manner can refer to the related introduction in the step 303b in the embodiment shown in FIG. 3, and details are not described herein again.

[0339] In the embodiments of this application, the first device determines first data and first indication information. The first indication information is used to indicate at least one first generation configuration. The at least one first generation configuration is used to generate at least one first simulation data set based on the first data. The first data is local data of the first device. Each first simulation data set in the at least one first simulation data set has a corresponding weight. The weight of each first simulation data set is used to indicate the importance of each first simulation data set. The first device sends the first data and the first indication information to the second device. Therefore, the first device indicates the at least one first generation configuration to the second device. This facilitates the second device to generate the at least one first simulation data set based on the at least one first generation configuration. This enables the second device to perform model training based on the at least one first simulation data set and the first data. Therefore, the amount of data used for model training is increased, which is beneficial to improve the model performance and / or generalization performance of the trained model. Further, each first simulation data set in the at least one first simulation data set has a corresponding weight. The weight of each first simulation data set is used to indicate the importance of each first simulation data set. This is beneficial to the second device to select appropriate data for corresponding model training based on the weight of each first simulation data set, so as to meet the needs of different models. For example, for a model with high model performance requirements, the second device can preferentially select a first simulation data set with a larger weight for model training. For a model with high generalization performance requirements, the second device can preferentially select a first simulation data set with a smaller weight for model training.

[0340] It should be noted that in the embodiments shown in FIG. 6, the second device generates at least one first simulation data set based on the first data and the at least one first generation configuration. In actual application, the second device can generate at least one first simulation data set based on the at least one first generation configuration. Each of the at least one first generation configuration includes part or all of the feature information of the first data and / or the data generation parameter corresponding to the first generation configuration. For the data generation parameter corresponding to the first generation configuration, please refer to the related description in the foregoing.

[0341] FIG. 7 is a structural schematic diagram of a first device according to an embodiment of the present application. Please refer to FIG. 7. The first device can be used to execute the processes performed by the first device in the embodiments shown in FIG. 3 and FIG. 6. For details, please refer to the related description in the foregoing method embodiments.

[0342] The first device 700 includes a transceiver module 701 and a processing module 702.

[0343] The processing module 702 is used for data processing. The transceiver module 701 can realize corresponding communication functions. The transceiver module 701 can also be called a communication interface or a communication module.

[0344] Optionally, the first device 700 further includes a storage module, which can be configured to store instructions and / or data. The processing module 802 can read the instructions and / or data in the storage module, so that the first device implements the foregoing method embodiments.

[0345] The first device 700 can be configured to perform the actions performed by the first device in the foregoing method embodiments. The first device 700 can be the first device or a component configurable to the first device. The processing module 702 is configured to perform the processing-related operations of the first device side in the foregoing method embodiments. The transceiver module 701 is configured to perform the receiving-related operations of the first device side in the foregoing method embodiments.

[0346] Optionally, the transceiver module 701 can include a sending module and a receiving module. The sending module is configured to perform the sending operations in the foregoing method embodiments. The receiving module is configured to perform the receiving operations in the foregoing method embodiments.

[0347] It should be noted that the first device 700 can include the sending module and not include the receiving module. Alternatively, the first device 700 can include the receiving module and not include the sending module. Specifically, whether the first device 700 includes the sending module and the receiving module can depend on whether the first device 700 performs the sending operations and the receiving operations in the foregoing schemes.

[0348] Optionally, the first device 700 is configured to perform the actions performed by the first device in the embodiment shown in FIG. 3. For details, refer to the related description in the embodiment shown in FIG. 3, which will not be repeated here. For example, the first device 700 is configured to perform the following scheme:

[0349] The processing module 702 is configured to generate at least one first simulation data set according to the first data. The first data is local data of the first device 700. Each first simulation data set in the at least one first simulation data set has a corresponding weight. The weight of each first simulation data set is used to indicate the importance of each first simulation data set.

[0350] The transceiver module 701 is configured to send the first data, the at least one first simulation data set, and first indication information to the second device. The first indication information is used to indicate the weight of each first simulation data set in the at least one first simulation data set.

[0351] In one possible implementation, the transceiver module 701 is further configured to send the weight of the first data to the second device. The weight of the first data is used to indicate the importance of the first data.

[0352] In a possible implementation, the data generation manner comprises: generating the first simulation data set according to the first data and the first model; or performing data augmentation on the first data to obtain the first simulation data set.

[0353] In a possible implementation, the data generation manner comprises: generating the first simulation data set according to the first data and the first model; or performing data augmentation on the first data to obtain the first simulation data set.

[0354] In a possible implementation, the weight of each first simulation data set in the at least one first simulation data set is determined according to a similarity between each first simulation data set in the at least one first simulation data set and the first data.

[0355] In a possible implementation, the weight of each first simulation data set in the at least one first simulation data set is determined according to a data distribution difference between each first simulation data set in the at least one first simulation data set and the first data.

[0356] In a possible implementation, the weight of each first simulation data set in the at least one first simulation data set is determined according to a signal-to-noise ratio difference between each first simulation data set in the at least one first simulation data set and the first data.

[0357] In a possible implementation, the weight of each first simulation data set in the at least one first simulation data set is determined according to a position difference between each first simulation data set in the at least one first simulation data set and the first data.

[0358] In a possible implementation, the transceiver 701 is further configured to receive, from the second device, an expected weight of each first simulation data set in the at least one first simulation data set, the expected weight being used to indicate an importance of an expected first simulation data set; and the processor 702 is specifically configured to generate the at least one first simulation data set according to the first data and the expected weight of each first simulation data set in the at least one first simulation data set.

[0359] In a possible implementation, the transceiver 701 is further configured to receive, from the second device, a first weight, the first weight being used to indicate an importance of the first device 700; and the processor 702 is specifically configured to generate the at least one first simulation data set according to the first data and the first weight.

[0360] In a possible implementation, the first indication information comprises a weight of each first simulation data set in the at least one first simulation data set.

[0361] In another possible implementation, the first indication information includes an identifier of each first simulation data set in the at least one first simulation data set, and the identifier of each first simulation data set is used to indicate a weight of the first simulation data set.

[0362] Optionally, the first device 700 is configured to perform the actions performed by the first device in the embodiment shown in FIG. 6. For details, refer to the related description in the embodiment shown in FIG. 6, which will not be repeated here. For example, the first device 700 is configured to perform the following scheme:

[0363] The processing module 702 is configured to determine first data and first indication information, the first indication information being used to indicate at least one first generation configuration, the at least one first generation configuration being used to generate at least one first simulation data set based on the first data, the first data being local data of the first device 700, each first simulation data set in the at least one first simulation data set having a corresponding weight, and the weight of each first simulation data set being used to indicate an importance of the first simulation data set.

[0364] The transceiver module 701 is configured to send the first data and the first indication information to a second device.

[0365] In a possible implementation, each first generation configuration in the at least one first generation configuration corresponds to one or more first simulation data sets in the at least one first simulation data set, and each first generation configuration in the at least one first generation configuration includes information or parameters used to assist in generating the first simulation data set corresponding to the first generation configuration.

[0366] In another possible implementation, the transceiver module 701 is further configured to: send second indication information to the second device, the second indication information being used to indicate a data generation manner corresponding to each first simulation data set in the at least one first simulation data set; receive first confirmation information from the second device, the first confirmation information being used to confirm the data generation manner corresponding to each first simulation data set in the at least one first simulation data set; and determine the at least one first generation configuration according to the data generation manner corresponding to each first simulation data set in the at least one first simulation data set.

[0367] In another possible implementation, the transceiver module 701 is further configured to: receive expected weights of each first simulation data set in the at least one first simulation data set from the second device, the expected weights being used to indicate expected importance of the first simulation data set; and determine the at least one first generation configuration according to the expected weights of each first simulation data set in the at least one first simulation data set.

[0368] In another possible implementation, the transceiver module 701 is further configured to receive the first weight from the second device, the first weight being used to indicate the importance of data of the first device 700, and the at least one first generated configuration is determined according to the first weight.

[0369] It should be understood that the specific processes in which the modules perform the corresponding processes described above have been described in detail in the method embodiments described above, and thus will not be described again here for the sake of brevity.

[0370] The processing module 702 in the above embodiments can be implemented by at least one processor or processor-related circuit. The transceiver module 701 can be implemented by a transceiver or transceiver-related circuit. The transceiver module 701 can also be referred to as a communication module or a communication interface. The storage module can be implemented by at least one memory.

[0371] FIG. 8 is a structural schematic diagram of a second device according to an embodiment of the present application. Referring to FIG. 8, the second device can be configured to perform the processes performed by the second device in the embodiments shown in FIGS. 3 and 6. For details, please refer to the related description in the method embodiments described above.

[0372] The second device 800 includes a transceiver module 801 and a processing module 802.

[0373] The processing module 802 is configured to perform data processing. The transceiver module 801 can implement corresponding communication functions. The transceiver module 801 can also be referred to as a communication interface or a communication module.

[0374] Optionally, the second device 800 can further include a storage module, which can be configured to store instructions and / or data. The processing module 802 can read the instructions and / or data in the storage module, so that the second device implements the method embodiments described above.

[0375] The second device 800 can be configured to perform the actions performed by the second device in the method embodiments described above. The second device 800 can be the second device or a component configurable to the second device. The processing module 802 is configured to perform the processing-related operations of the second device side in the method embodiments described above. The transceiver module 801 is configured to perform the receiving-related operations of the second device side in the method embodiments described above.

[0376] Optionally, the transceiver module 801 can include a sending module and a receiving module. The sending module is configured to perform the sending operations in the method embodiments described above. The receiving module is configured to perform the receiving operations in the method embodiments described above.

[0377] It should be noted that the second device 800 can include a sending module but not a receiving module. Alternatively, the second device 800 can include a receiving module but not a sending module. The specific implementation can depend on whether the above-described scheme executed by the second device 800 includes sending actions and receiving actions.

[0378] Optionally, the second device 800 is configured to perform the actions performed by the second device in the embodiment shown in FIG. 3. For details, refer to the related description in the embodiment shown in FIG. 3, which will not be repeated here. For example, the second device 800 is configured to perform the following scheme:

[0379] The transceiver module 801 is configured to receive first data, at least one first simulation data set, and first indication information from the first device, the first data being local data of the first device, the at least one first simulation data set being generated according to the first data, and the first indication information being used to indicate the weight of each first simulation data set in the at least one first simulation data set, the weight of each first simulation data set being used to indicate the importance of each first simulation data set.

[0380] The processing module 802 is configured to perform model training according to the first data, the at least one first simulation data set, and the weight of each first simulation data set in the at least one first simulation data set, to obtain one or more models.

[0381] In a possible implementation, the transceiver module 801 is further configured to receive the weight of the first data from the first device, the weight of the first data being used to indicate the importance of the first data; and the processing module 802 is specifically configured to perform model training according to the first data, the weight of the first data, the at least one first simulation data set, and the weight of each first simulation data set in the at least one first simulation data set, to obtain one or more models.

[0382] In another possible implementation, the weight of each first simulation data set in the at least one first simulation data set is determined according to the similarity between each first simulation data set in the at least one first simulation data set and the first data.

[0383] In another possible implementation, the weight of each first simulation data set in the at least one first simulation data set is determined according to the data distribution difference between each first simulation data set in the at least one first simulation data set and the first data.

[0384] In another possible implementation, the weight of each first simulation data set in the at least one first simulation data set is determined according to the signal-to-noise ratio difference between each first simulation data set in the at least one first simulation data set and the first data.

[0385] In another possible implementation, the weight of each first simulation data set in the at least one first simulation data set is determined according to the position difference between each first simulation data set in the at least one first simulation data set and the first data.

[0386] In another possible implementation, the transceiving module 801 is further configured to send, to the first device, an expected weight of each first simulation data set in the at least one first simulation data set, the expected weight being used to indicate an importance of the expected first simulation data set.

[0387] In another possible implementation, the one or more models comprise a second model; and the processing module 802 is specifically configured to: select, from the first data and the at least one first simulation data set, target data according to at least one of a model performance requirement and a generalization performance requirement of the second model and the weight of each first simulation data set in the at least one first simulation data set; and perform model training by using the target data to obtain the second model.

[0388] In another possible implementation, the transceiving module 801 is further configured to: receive, from a third device, second data, at least one second simulation data set, and second indication information, the second data being local data of the third device, the at least one second simulation data set being generated according to the second data, and the second indication information being used to indicate a weight of each second simulation data set in the at least one second simulation data set, the weight of each second simulation data set in the at least one second simulation data set being used to indicate an importance of the second simulation data set; and the processing module 802 is specifically configured to: perform model training according to the first data, the at least one first simulation data set, the second data, the at least one second simulation data set, the weight of each first simulation data set in the at least one first simulation data set, and the weight of each second simulation data set in the at least one second simulation data set to obtain the one or more models.

[0389] In another possible implementation, the processing module 802 is specifically configured to: determine a target data set according to a first weight, the first data, the at least one first simulation data set, the weight of each first simulation data set in the at least one first simulation data set, a second weight, the second data, the at least one second simulation data set, and the weight of each second simulation data set in the at least one second simulation data set, the first weight being used to indicate an importance of the data of the first device, and the second weight being used to indicate an importance of the data of the third device; and perform model training by using the target data set to obtain the one or more models.

[0390] In another possible implementation, the transceiving module 801 is further configured to: send, to the first device, the first weight.

[0391] In another possible implementation, the transceiving module 801 is further configured to: send, to the third device, the second weight.

[0392] In another possible implementation, the first indication information comprises the weight of each first simulation data set in the at least one first simulation data set.

[0393] In another possible implementation, the first indication information includes an identifier of each first simulation data set in the at least one first simulation data set, and the identifier of each first simulation data set is used to indicate a weight of each first simulation data set in the at least one first simulation data set.

[0394] Optionally, the second device 800 is configured to perform the actions performed by the second device in the embodiment shown in FIG. 6. For details, refer to the related description in the embodiment shown in FIG. 6, which will not be repeated here. For example, the second device 800 is configured to perform the following scheme:

[0395] The transceiver module 801 is configured to receive first data and first indication information from the first device, the first indication information is used to indicate at least one first generation configuration, and the first data is local data of the first device.

[0396] The processing module 802 is configured to generate at least one first simulation data set according to the first data and the at least one first generation configuration, each first simulation data set in the at least one first simulation data set has a corresponding weight, and the weight of each first simulation data set is used to indicate an importance of the first simulation data set.

[0397] In a possible implementation, each first generation configuration in the at least one first generation configuration corresponds to one or more first simulation data sets in the at least one first simulation data set, and each first generation configuration in the at least one first generation configuration includes information or parameters used to assist in generating the first simulation data set corresponding to the first generation configuration.

[0398] In another possible implementation, the transceiver module 801 is further configured to: receive second indication information from the first device, the second indication information is used to indicate a data generation manner corresponding to each first simulation data set in the at least one first simulation data set; and send first confirmation information to the first device, the first confirmation information is used to confirm the data generation manner corresponding to each first simulation data set in the at least one first simulation data set.

[0399] In another possible implementation, the processing module 802 is specifically configured to: each first simulation data set in the at least one first simulation data set has a corresponding data generation manner; and generate the first simulation data set according to the first data, the data generation manner, and the first generation configuration corresponding to the first simulation data set.

[0400] In another possible implementation, the data generation manner includes: generating the first simulation data set according to the first data and a first model; or performing data enhancement on the first data to obtain the first simulation data set.

[0401] In another possible implementation, the processing module 802 is further configured to perform model training according to the first data, the at least one first simulation data set, and the weight of each first simulation data set in the at least one first simulation data set, to obtain the one or more models.

[0402] In another possible implementation, the one or more models include a first model, and the processing module 802 is specifically configured to select target data from the first data and the at least one first simulation data set according to at least one of a model performance requirement and a generalization performance requirement of the first model and the weight of each first simulation data set in the at least one first simulation data set, and perform model training on the target data to obtain the first model.

[0403] In another possible implementation, the transceiver 801 is further configured to receive second data and at least one second generation configuration from a third device, the second data being local data of the third device, and the processing module 802 is further configured to generate at least one second simulation data set according to the second data and the at least one second generation configuration, each second simulation data set in the at least one second simulation data set having a corresponding weight, and the weight of each second simulation data set being used to indicate an importance of the second simulation data set, and the processing module is specifically configured to perform model training according to the first data, the second data, the at least one first simulation data set, the at least one second simulation data set, the weight of each first simulation data set in the at least one first simulation data set, and the weight of each second simulation data set in the at least one second simulation data set, to obtain the one or more models.

[0404] In another possible implementation, the processing module 802 is specifically configured to determine a target data set according to the first weight, the first data, the at least one first simulation data set and the weight of each first simulation data set in the at least one first simulation data set, the second weight, the second data, the at least one second simulation data set, and the weight of each second simulation data set in the at least one second simulation data set, wherein the first weight is used to indicate an importance of data of the first device, and the second weight is used to indicate an importance of data of the third device, and perform model training on the target data set to obtain the one or more models.

[0405] In another possible implementation, the transceiver 801 is further configured to send the first weight to the first device.

[0406] In another possible implementation, the transceiver 801 is further configured to send the second weight to the third device.

[0407] In another possible implementation, the transceiver 801 is further configured to send, to the first device, an expected weight of each first simulation data set in the at least one first simulation data set, the expected weight being used to indicate an expected weight of the first simulation data set.

[0408] It should be understood that the specific processes in which the modules perform the respective processes described above have been described in detail in the method embodiments described above, and for the sake of brevity, will not be repeated here.

[0409] The processing module 802 in the above embodiments can be implemented by at least one processor or processor-related circuit. The transceiver module 801 can be implemented by a transceiver or transceiver-related circuit. The transceiver module 801 can also be referred to as a communication module or a communication interface. The storage module can be implemented by at least one memory.

[0410] Embodiments of the present application also provide a device 900. Please refer to FIG. 9, the device 900 includes a processor 910, and the processor 910 is coupled with a memory 920, the memory 920 is used to store computer programs or instructions and / or data, and the processor 910 is used to execute the computer programs or instructions and / or data stored in the memory 920, so that the method in the above method embodiments is executed. The device 900 is used to implement the operations performed by the first device or the second device in the above method embodiments.

[0411] Optionally, the processor 910 included in the device 900 is one or more.

[0412] Optionally, as shown in FIG. 9, the device 900 can also include a memory 920.

[0413] Optionally, the memory 920 included in the device 900 can be one or more.

[0414] Optionally, the memory 920 can be integrated with the processor 910 or separately arranged.

[0415] Optionally, as shown in FIG. 9, the device 900 can also include a transceiver 930, which is used for receiving and / or sending signals. For example, the processor 910 is used to control the transceiver 930 to receive and / or send signals.

[0416] The present application also provides a device 1000, which can be a terminal device, a processor in a terminal device, or a chip. The device 1000 can be used to execute the operations performed by the first device or the second device in the above method embodiments.

[0417] When the device 1000 is a terminal device, FIG. 10 shows a simplified structural schematic diagram of a terminal device. As shown in FIG. 10, the terminal device includes a processor, a memory, and a transceiver. The memory can store computer program codes, and the transceiver includes a transmitter 1031, a receiver 1032, a radio frequency circuit (not shown in the figure), an antenna 1033, and an input and output device (not shown in the figure).

[0418] The processor is mainly used for processing communication protocol and communication data, controlling terminal device, executing software program and processing data of software program, etc.

[0419] The memory is mainly used for storing software program and data.

[0420] The radio frequency circuit is mainly used for converting baseband signal and radio frequency signal and processing radio frequency signal.

[0421] The antenna is mainly used for receiving and sending radio frequency signal in the form of electromagnetic wave.

[0422] The input and output device can include touch screen, display screen, keyboard, etc. The input and output device is mainly used for receiving user input data and outputting data to user. It should be noted that some kinds of terminal device can not have input and output device.

[0423] When data needs to be sent, the processor performs baseband processing on the data to be sent, and outputs the baseband signal to the radio frequency circuit. Then, the radio frequency circuit performs radio frequency processing on the baseband signal, and sends the radio frequency signal in the form of electromagnetic wave through the antenna. When data is sent to the terminal device, the radio frequency circuit receives the radio frequency signal through the antenna. The radio frequency circuit converts the radio frequency signal into baseband signal, and outputs the baseband signal to the processor. The processor converts the baseband signal into data and processes the data. For the convenience of description, only one memory, processor and transceiver are shown in FIG. 10. In actual terminal device product, one or more processors and one or more memories can exist. The memory can also be referred to as storage medium or storage device, etc. The memory can be set independently of the processor, or can be integrated with the processor. The embodiments of the present application do not limit this.

[0424] In the embodiments of the present application, the antenna and the radio frequency circuit with transceiving function can be regarded as the transceiving module of the terminal device, and the processor with processing function can be regarded as the processing module of the terminal device.

[0425] As shown in FIG. 10, the terminal device includes processor 1010, memory 1020 and transceiver 1030. The processor 1010 can also be referred to as processing unit, processing board, processing module, or processing device, etc. The transceiver 1030 can also be referred to as transceiving unit, transceiver, or transceiving device, etc.

[0426] Optionally, the device for implementing the receiving function in the transceiver 1030 is regarded as a receiving module, and the device for implementing the sending function in the transceiver 1030 is regarded as a sending module, that is, the transceiver 1030 includes a receiver and a transmitter. The transceiver can also be referred to as a transceiver, a transceiver module, or a transceiver circuit, etc. The receiver can also be referred to as a receiver, a receiving module, or a receiving circuit, etc. The transmitter can also be referred to as a transmitter, a transmitting module, or a transmitting circuit, etc.

[0427] The processor 1010 is configured to perform the processing operation of the first device or the second device in the embodiments shown in FIG. 3 and FIG. 6. The transceiver 1030 is configured to perform the transceiving operation of the first device in the embodiments shown in FIG. 3 and FIG. 6.

[0428] It should be understood that FIG. 10 is merely an example and not limiting, and the terminal device including the transceiver module and the processing module described above can not depend on the structure shown in FIG. 7, FIG. 8, or FIG. 10.

[0429] When the device 1000 is a chip, the chip includes a processor, a memory, and a transceiver. The transceiver can be an input / output circuit or a communication interface. The processor can be a processing module integrated on the chip or a microprocessor or an integrated circuit. The sending operation of the first device or the second device in the method embodiments can be understood as the output of the chip, and the receiving operation of the first device or the second device in the method embodiments can be understood as the input of the chip.

[0430] The present application also provides a device 1100, which can be a network device or a chip. The device 1100 can be configured to perform the operations performed by the first device or the second device in the embodiments shown in FIG. 3 and FIG. 6.

[0431] When the device 1100 is a network device, for example, a base station. FIG. 11 shows a simplified structure diagram of a base station. The base station includes a 1110 part, a 1120 part, and a 1130 part.

[0432] The 1110 part is mainly used for baseband processing, controlling the base station, etc. The 1110 part is usually the control center of the base station and can be referred to as a processor, which is configured to control the base station to perform the processing operation of the first device or the second device in the method embodiments.

[0433] The 1120 part is mainly used for storing computer program codes and data.

[0434] 1130 is mainly configured to transceive radio frequency signals and convert radio frequency signals and baseband signals. 1130 can be referred to as a transceiver module, a transceiver, a transceiving circuit, or a transceiver device. The transceiver module of 1130 can also be referred to as a transceiver or a transceiver device, and includes an antenna 1133 and a radio frequency circuit (not shown in the figure), where the radio frequency circuit is mainly configured to perform radio frequency processing. Optionally, the devices in 1130 that are configured to implement receiving functions can be regarded as a receiver, and the devices in 1130 that are configured to implement transmitting functions can be regarded as a transmitter, that is, 1130 includes a receiver 1132 and a transmitter 1131. The receiver can also be referred to as a receiving module, a receiver, or a receiving circuit, and the transmitter can be referred to as a transmitting module, a transmitter, or a transmitting circuit.

[0435] 1110 and 1120 can include one or more single boards, and each single board can include one or more processors and one or more memories. The processors are configured to read and execute programs in the memories to implement baseband processing functions and control the base station. If there are multiple single boards, the single boards can be interconnected to enhance processing capability. As an optional implementation, multiple single boards can share one or more processors, or multiple single boards can share one or more memories, or multiple single boards can share one or more processors at the same time.

[0436] For example, in an implementation, the transceiver module of 1130 is configured to perform the transceiving-related processes performed by the first device or the second device in the embodiments shown in FIG. 3 and FIG. 6. The processor of 1110 is configured to perform the processing-related processes performed by the first device or the second device in the embodiments shown in FIG. 3 and FIG. 6.

[0437] It should be understood that FIG. 11 is merely an example and not limiting, and the network device including the processor, the memory, and the transceiver described above can not depend on the structure shown in FIG. 7, FIG. 8, or FIG. 11.

[0438] When the device 1100 is a chip, the chip includes a transceiver, a memory, and a processor. The transceiver can be an input / output circuit, a communication interface; the processor is a processor integrated on the chip, or a microprocessor, or an integrated circuit. The transmitting operation of the first device or the second device in the method embodiments described above can be understood as the output of the chip, and the receiving operation of the first device or the second device in the method embodiments described above can be understood as the input of the chip.

[0439] The embodiments of the present application also provide a computer readable storage medium, which stores computer instructions for implementing the method performed by the first device or the second device in the method embodiments described above.

[0440] For example, the computer program, when executed by a computer, enables the computer to implement the method performed by the first device or the second device in the above method embodiments.

[0441] The embodiments of the present application further provide a computer program product containing instructions, which, when executed by a computer, enable the computer to implement the method performed by the first device or the second device in the above method embodiments.

[0442] The embodiments of the present application further provide a communication system, which comprises the first device in the above embodiments and the second device in the above embodiments. The first device is configured to perform part or all of the operations performed by the first device in the above method embodiments, and the second device is configured to perform part or all of the operations performed by the second device in the above method embodiments.

[0443] The embodiments of the present application further provide a chip device, which comprises a processor configured to invoke computer degrees or computer instructions stored in a memory, so as to enable the processor to perform the method provided by the embodiments shown in FIG. 3 and FIG. 6.

[0444] In a possible implementation, the input of the chip device corresponds to the receiving operation in any one of the embodiments shown in FIG. 3 and FIG. 6, and the output of the chip device corresponds to the sending operation in any one of the embodiments shown in FIG. 3 and FIG. 6.

[0445] Optionally, the processor is coupled to the memory through an interface.

[0446] Optionally, the chip device further comprises a memory, and the memory stores computer degrees or computer instructions.

[0447] The processor mentioned in any one of the above embodiments can be a general central processing unit, a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits for controlling the execution of the program of the method provided by any one of the embodiments shown in FIG. 3 and FIG. 6. The memory mentioned in any one of the above embodiments can be a read-only memory (ROM) or other type of static storage device that can store static information and instructions, a random access memory (RAM), and the like.

[0448] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the explanation and beneficial effects of the related content in any one of the above provided devices can refer to the corresponding method embodiments provided above, and will not be repeated here.

[0449] 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 described device embodiments are merely schematic. The division of the units is merely logical function division. There can be other division manners in 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 different units, can be indirect couplings or communication connections through some interfaces, devices or units, and can be in electrical, mechanical or other forms.

[0450] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e., can be located in one place, or can be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0451] In addition, each functional unit in the various embodiments of the present application can be integrated in one processing unit, or each unit can be physically present separately, 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.

[0452] The integrated unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application essentially make contributions or the entire 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 number 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 methods described in the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, ROM, RAM, magnetic disk or optical disk, and various program codes that can be stored in the medium.

[0453] The above embodiments are merely used to describe the technical solutions of the present application, rather than limit them. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced by equivalent ones. Such modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A data transmission method, characterized in that, The method includes: The first device generates at least one first simulation dataset based on the first data, wherein the first data is local data of the first device, and each first simulation dataset in the at least one first simulation dataset has a corresponding weight, and the weight of each first simulation dataset is used to indicate the importance of the first simulation dataset. The first device sends the first data, the at least one first simulation dataset, and first indication information to the second device. The first indication information is used to indicate the weight of each first simulation dataset in the at least one first simulation dataset.

2. The method according to claim 1, characterized in that, The method further includes: The first device sends the weight of the first data to the second device, and the weight of the first data is used to indicate the importance of the first data.

3. The method according to claim 1 or 2, characterized in that, Each of the at least one first simulation datasets has a corresponding data generation method; The first device generates at least one first simulation dataset based on the first data, including: The first device generates the at least one first simulation dataset according to the data generation method corresponding to each first simulation dataset in the at least one first simulation dataset and the first data.

4. The method according to claim 3, characterized in that, The data generation methods include: The first simulation dataset is generated based on the first data and the first model; or... The first data is augmented to obtain the first simulation dataset.

5. The method according to any one of claims 1 to 4, characterized in that, The method further includes: The first device receives the expected weights of each first simulation dataset in the at least one first simulation dataset from the second device, the expected weights being used to indicate the expected importance of the first simulation dataset; The first device generates at least one first simulation dataset based on the first data, including: The first device generates the at least one first simulation dataset based on the first data and the expected weights of each of the first simulation datasets in at least one first simulation dataset.

6. The method according to any one of claims 1 to 5, characterized in that, The method further includes: The first device receives a first weight from the second device, the first weight being used to indicate the importance of the data in the first device; The first device generates at least one first simulation dataset based on the first data, including: The first device generates the at least one first simulation dataset based on the first data and the first weight.

7. The method according to any one of claims 1 to 6, characterized in that, The first indication information includes the weights of each of the at least one first simulation datasets.

8. The method according to any one of claims 1 to 6, characterized in that, The first indication information includes the identifier of each first simulation dataset in the at least one first simulation dataset, and the identifier of each first simulation dataset is used to indicate the weight of each first simulation dataset.

9. A data transmission method, characterized in that, The method includes: The second device receives first data, at least one first simulation dataset, and first indication information from the first device. The first data is local data of the first device. The at least one first simulation dataset is generated based on the first data. The first indication information is used to indicate the weight of each first simulation dataset in the at least one first simulation dataset. The weight of each first simulation dataset is used to indicate the importance of each first simulation dataset. The second device trains a model based on the first data, the at least one first simulation dataset, and the weights of each first simulation dataset in the at least one first simulation dataset, to obtain one or more models.

10. The method according to claim 9, characterized in that, The method further includes: The second device receives the weight of the first data from the first device, the weight of the first data being used to indicate the importance of the first data; The second device trains a model based on the first data, the at least one first simulation dataset, and the weights of each first simulation dataset in the at least one first simulation dataset, to obtain one or more models, including: The second device trains a model based on the first data, the weights of the first data, the at least one first simulation dataset, and the weights of each first simulation dataset in the at least one first simulation dataset, to obtain the one or more models.

11. The method according to claim 9 or 10, characterized in that, The method further includes: The second device sends the expected weights of each of the at least one first simulation datasets to the first device, the expected weights being used to indicate the expected importance of the first simulation dataset.

12. The method according to any one of claims 9 to 11, characterized in that, The one or more models include a second model; the second device trains the model based on the first data, the at least one first simulation dataset, and the weights of each first simulation dataset in the at least one first simulation dataset to obtain one or more models, including: The second device selects target data from the first data and the at least one first simulation dataset based on at least one of the model performance requirements and generalization performance requirements of the second model and the weights of each first simulation dataset in the at least one first simulation dataset; The second device trains the model using the target data to obtain the second model.

13. The method according to any one of claims 9 to 12, characterized in that, The method further includes: The second device receives second data, at least one second simulation dataset, and second indication information from the third device. The second data is local data of the third device. The at least one second simulation dataset is generated based on the second data. The second indication information is used to indicate the weight of each second simulation dataset in the at least one second simulation dataset. The weight of each second simulation dataset in the at least one second simulation dataset is used to indicate the importance of the second simulation dataset. The second device trains a model based on the first data, the at least one first simulation dataset, and the weights of each first simulation dataset in the at least one first simulation dataset, to obtain one or more models, including: The second device trains a model based on the first data, the at least one first simulation dataset, the second data, the at least one second simulation dataset, the weights of each first simulation dataset in the at least one first simulation dataset, and the weights of each second simulation dataset in the at least one second simulation dataset, to obtain the one or more models.

14. The method according to claim 13, characterized in that, The second device trains a model based on the first data, the at least one first simulation dataset, the second data, the at least one second simulation dataset, the weights of each first simulation dataset in the at least one first simulation dataset, and the weights of each second simulation dataset in the at least one second simulation dataset, to obtain the one or more models, including: The second device determines the target data set based on a first weight, the first data, the at least one first simulation dataset, the weight of each first simulation dataset in the at least one first simulation dataset, a second weight, the second data, the at least one second simulation dataset, and the weight of each second simulation dataset in the at least one second simulation dataset. The first weight is used to indicate the importance of the data in the first device, and the second weight is used to indicate the importance of the data in the third device. The second device trains a model using the target dataset to obtain one or more models.

15. The method according to any one of claims 9 to 14, characterized in that, The method further includes: The second device sends a first weight to the first device, the first weight being used to indicate the importance of the data in the first device.

16. The method according to any one of claims 9 to 15, characterized in that, The method further includes: The second device sends a second weight to the third device, the second weight being used to indicate the importance of the data in the third device.

17. The method according to any one of claims 9 to 16, characterized in that, The first indication information includes the weights of each of the at least one first simulation datasets.

18. The method according to any one of claims 9 to 16, characterized in that, The first indication information includes the identifier of each first simulation dataset in the at least one first simulation dataset, and the identifier of each first simulation dataset is used to indicate the weight of each first simulation dataset in the at least one first simulation dataset.

19. The method according to any one of claims 9 to 18, characterized in that, The weight of each first simulation dataset in the at least one first simulation dataset is determined based on the similarity between each first simulation dataset in the at least one first simulation dataset and the first data.

20. The method according to any one of claims 9 to 18, characterized in that, The weights of each first simulation dataset in the at least one first simulation dataset are determined based on the differences in data distribution between each first simulation dataset in the at least one first simulation dataset and the first data.

21. The method according to any one of claims 9 to 18, characterized in that, The weights of each first simulation dataset in the at least one first simulation dataset are determined based on the signal-to-noise ratio difference between each first simulation dataset in the at least one first simulation dataset and the first data.

22. The method according to any one of claims 9 to 18, characterized in that, The weight of each first simulation dataset in the at least one first simulation dataset is determined based on the positional difference between each first simulation dataset and the first data.

23. A data transmission method, characterized in that, The method includes: The first device determines first data and first indication information. The first indication information is used to indicate at least one first generation configuration. The at least one first generation configuration is used to generate at least one first simulation dataset based on the first data. The first data is local data of the first device. Each first simulation dataset in the at least one first simulation dataset has a corresponding weight. The weight of each first simulation dataset is used to indicate the importance of the first simulation dataset. The first device sends the first data and the first instruction information to the second device.

24. The method according to claim 23, characterized in that, Each of the at least one first generation configurations corresponds to one or more first simulation datasets in the at least one first simulation dataset. Each of the at least one first generation configurations includes information or parameters for assisting in the generation of the first simulation dataset corresponding to the first generation configuration.

25. The method according to claim 23 or 24, characterized in that, The method further includes: The first device sends a second instruction to the second device, the second instruction being used to indicate the data generation method corresponding to each of the at least one first simulation datasets; The first device receives first confirmation information from the second device, the first confirmation information being used to confirm the data generation method corresponding to each of the at least one first simulation datasets; The at least one first generation configuration is determined based on the data generation method corresponding to each first simulation dataset in the at least one first simulation dataset.

26. The method according to any one of claims 23 to 25, characterized in that, The method further includes: The first device receives the expected weights of each first simulation dataset in the at least one first simulation dataset from the second device, the expected weights being used to indicate the expected importance of the first simulation dataset; The at least one first generation configuration is determined based on the expected weights of each of the at least one first simulation datasets.

27. The method according to any one of claims 23 to 26, characterized in that, The method further includes: The first device receives a first weight from the second device, the first weight being used to indicate the importance of the data in the first device; The at least one first generation configuration is determined based on the first weight.

28. A data transmission method, characterized in that, The method includes: The second device receives first data and first indication information from the first device, wherein the first indication information is used to indicate at least one first generation configuration, and the first data is local data of the first device. The second device generates at least one first simulation dataset based on the first data and the at least one first generation configuration. Each first simulation dataset in the at least one first simulation dataset has a corresponding weight, and the weight of each first simulation dataset is used to indicate the importance of the first simulation dataset.

29. The method according to claim 28, characterized in that, Each of the at least one first generation configurations corresponds to one or more first simulation datasets in the at least one first simulation dataset. Each of the at least one first generation configurations includes information or parameters for assisting in the generation of the first simulation dataset corresponding to the first generation configuration.

30. The method according to claim 28 or 29, characterized in that, The method further includes: The second device receives second instruction information from the first device, the second instruction information being used to indicate the data generation method corresponding to each of the at least one first simulation datasets; The second device sends a first confirmation message to the first device, the first confirmation message being used to confirm the data generation method corresponding to each of the at least one first simulation datasets.

31. The method according to any one of claims 28 to 30, characterized in that, The second device generates at least one first simulation dataset based on the first data and the at least one first generation configuration, including: Each of the at least one first simulation datasets has a corresponding data generation method; The first device generates the first simulation dataset based on the first data, the data generation method, and the first generation configuration corresponding to the first simulation dataset.

32. The method according to claim 31, characterized in that, The data generation methods include: The first simulation dataset is generated based on the first data and the first model; or, The first simulation dataset is obtained by performing data augmentation on the first data.

33. The method according to any one of claims 28 to 32, characterized in that, The method further includes: The second device trains a model based on the first data, the at least one first simulation dataset, and the weights of each first simulation dataset in the at least one first simulation dataset, to obtain one or more models.

34. The method according to claim 33, characterized in that, The one or more models include a first model; the second device trains the model based on the first data, the at least one first simulation dataset, and the weights of each first simulation dataset in the at least one first simulation dataset to obtain one or more models, including: The second device selects target data from the first data and the at least one first simulation dataset based on at least one of the model performance requirements and generalization performance requirements of the first model and the weights of each first simulation dataset in the at least one first simulation dataset; The second device trains the model using the target data to obtain the first model.

35. The method according to claim 33 or 34, characterized in that, The method further includes: The second device receives second data and at least one second generated configuration from the third device, wherein the second data is local data of the third device; The second device generates at least one second simulation dataset based on the second data and the at least one second generation configuration. Each second simulation dataset in the at least one second simulation dataset has a corresponding weight, and the weight of each second simulation dataset is used to indicate the importance of the second simulation dataset. The second device trains a model based on the first data, the at least one first simulation dataset, and the weights of each first simulation dataset in the at least one first simulation dataset, to obtain one or more models, including: The second device trains a model based on the first data, the second data, the at least one first simulation dataset, the at least one second simulation dataset, the weights of each first simulation dataset in the at least one first simulation dataset, and the weights of each second simulation dataset in the at least one second simulation dataset, to obtain the one or more models.

36. The method according to claim 35, characterized in that, The second device trains a model based on the first data, the second data, the at least one first simulation dataset, the at least one second simulation dataset, the weights of each first simulation dataset in the at least one first simulation dataset, and the weights of each second simulation dataset in the at least one second simulation dataset, to obtain the one or more models, including: The second device determines the target data set based on a first weight, the first data, the at least one first simulation dataset and the weights of each first simulation dataset in the at least one first simulation dataset, a second weight, the second data, the at least one second simulation dataset and the weights of each second simulation dataset in the at least one second simulation dataset, wherein the first weight is used to indicate the importance of the data of the first device, and the second weight is used to indicate the importance of the data of the third device; The second device trains a model using the target dataset to obtain one or more models.

37. The method according to any one of claims 28 to 36, characterized in that, The method further includes: The second device sends a first weight to the first device, the first weight being used to indicate the importance of the data in the first device.

38. The method according to any one of claims 28 to 37, characterized in that, The method further includes: The second device sends the expected weights of each of the at least one first simulation datasets to the first device, the expected weights being used to indicate the expected weights of the first simulation datasets.

39. An apparatus, characterized in that, The device includes a transceiver module and a processing module; The transceiver module is used to perform the transceiver operation of the method as described in any one of claims 1 to 8, and the processing module is used to perform the processing operation of the method as described in any one of claims 1 to 8; or, The transceiver module is used to perform the transceiver operation of the method as described in any one of claims 9 to 22, and the processing module is used to perform the processing operation of the method as described in any one of claims 9 to 22; or, The transceiver module is used to perform the transceiver operation of the method as described in any one of claims 23 to 27, and the processing module is used to perform the processing operation of the method as described in any one of claims 23 to 27; or, The transceiver module is used to perform the transceiver operation of the method as described in any one of claims 28 to 38, and the processing module is used to perform the processing operation of the method as described in any one of claims 28 to 38.

40. An apparatus, characterized in that, The apparatus includes a processor for executing a computer program or computer instructions in a memory to perform the method as described in any one of claims 1 to 38.

41. A computer-readable storage medium, characterized in that, It stores a computer program thereon, which, when executed by the device, causes the device to perform the method as described in any one of claims 1 to 38.

42. A computer program product, characterized in that, When the computer program product is run on a computer, it causes the computer to perform the method as described in any one of claims 1 to 38.

Images