Data transmission method and data transmission device

Abstract

The embodiment of the invention provides a data transmission method and a data transmission device, and relates to the technical field of communication. According to the method, the data of the first type can be received by sending the first information used for indicating to report the data of the first type. Wherein the data of the first category is one of a plurality of categories, and the data of each category in the plurality of categories is used for AI model training. For example, data of each of the plurality of categories is included in a training set. The training set is used for AI model training. Therefore, under the condition that the first category is the category with insufficient data volume in the training set, the data of the category with insufficient data volume can be collected in a targeted manner, so that the data volume of each category in the training set is sufficient, that is, the balance of the data features of the training set is relatively good, and the training efficiency is improved. And therefore, the generalization of the AI model trained by adopting the training set is relatively good.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of communication, and particularly relates to a data transmission method and a data transmission device. BACKGROUND

[0002] In a communication network, with the diversification of service requirements and the enhancement of network functions, service implementation, network planning, configuration and resource scheduling also become increasingly complex. For example, service implementation, network planning, configuration and resource scheduling may involve modulation, coding, transmitters, receivers, multi-antenna technology, or positioning technology in a communication system. Among them, the network device can implement schemes such as modulation and demodulation of signals, information encoding and decoding, channel state information (CSI) feedback, beam management (BM) or mobility management by performing related operations.

[0003] At present, in the process of implementing the above schemes through a communication network, the network device needs to manage the operations involved in the above schemes to configure corresponding resources. However, in actual application, the management scheme in the related art is difficult to ensure the normal execution of the related operations or the actual execution effect is very poor. SUMMARY

[0004] In order to solve the above technical problems, the embodiments of the present application provide a data transmission method and a data transmission device, which can ensure the normal use of functions or models and improve the actual use effect of functions or models.

[0005] In a first aspect, a data transmission method is provided, applied to a first device, and the method comprises: sending first information, the first information being used to indicate reporting data of a first category, the first category being one of multiple categories, and data of each category in the multiple categories being used for model training; and receiving data of the first category.

[0006] The specific implementation principles of the embodiments of the present application can refer to the specific implementation principles of the embodiments shown in S101-S102. The model in the embodiments of the present application can be an AI model. The specific implementation principles of the embodiments of the present application can also refer to the specific implementation principles of the embodiments shown in S706-S709 or the embodiments shown in S808-S811.

[0007] According to the above scheme, the first category in the plurality of categories is targeted for data collection. The amount of data of each category in the plurality of categories for model (such as AI model) training can be sufficient, and the probability of poor generalization of the trained model due to insufficient data amount or poor data feature balance of the training set in the training set can be reduced. The targeted data collection of the first category can be understood as targeted collection of data of the first category. The training set can be a training set for model (such as AI model) training. The training set can include data of each category in the plurality of categories. For example, in the case that the terminal device uses an AI model for beam management (such as optimal beam prediction), the probability of poor signal quality of the optimal beam predicted by the terminal device using the AI model can be reduced due to poor generalization of the model, and thus the probability of poor signal quality of the optimal beam fed back by the terminal device to the network device can be reduced.

[0008] In combination with the first aspect, in some implementations of the first aspect, the first device collects a data set, and the data of the first category is data with insufficient amount in the data set, and the data set includes data of the plurality of categories.

[0009] According to the above scheme, before sending the first information, the first device can collect a data set to select a category with insufficient data amount from a plurality of categories included in the data set, so as to collect data of the category with insufficient data amount, so that the amount of data of each category in the training set for model (such as AI model) training is sufficient, that is, the data feature balance of the training set is good. In turn, the generalization of the AI model trained by the training set can be good. The specific implementation principles and technical effects of the first device collecting the data set before sending the first information can be referred to the specific implementation principles and technical effects of S701-S704, or can be referred to the specific implementation principles and technical effects of S801-S804.

[0010] In combination with the first aspect, in some implementations of the first aspect, the data of the first category satisfies at least one of the following: the amount of data of the first category in the data set is less than a first threshold. The proportion of the amount of data of the first category in the data set is less than a second threshold. The ratio of the amount of data of the first category to the amount of data of a target category is less than a third threshold, the target category is different from the first category, and the data of the target category is included in the data set. The absolute value of the difference between the proportion of the first category and the proportion of the target category is greater than a fourth threshold, the proportion of the first category is the proportion of the amount of data of the first category in the data set, and the proportion of the target category is the proportion of the amount of data of the target category in the data set.

[0011] According to the scheme, the data set can be classified, and the data amount of each category in the data set can be obtained, so that the first category with insufficient data amount can be selected from the data set, and targeted data collection can be performed on the first category.

[0012] With reference to the first aspect, in some implementations of the first aspect, the data set further includes second category data to Nth category data, where N is an integer greater than 2.

[0013] According to the scheme, the data set further includes second category data to Nth category data, that is, the data set can include first category data to Nth category data. Before the first information is sent, the first device can collect the first category data to the Nth category data.

[0014] With reference to the first aspect, in some implementations of the first aspect, the categories in the data set are obtained based on one or more of the following classifications: preset N categories of data features, preset N categories of beam sets, preset N categories of cell sets, preset N categories of resource configurations, preset N categories of speeds, preset N categories of time ranges, or preset N categories of latitude and longitude ranges.

[0015] According to the scheme, the data set can be classified, and the data amount of each category in the data set can be obtained, so that the first category with insufficient data amount can be selected from the data set, and targeted data collection can be performed on the first category.

[0016] With reference to the first aspect, in some implementations of the first aspect, the resource configuration can include a radio resource control (RRC) configuration, a media access control-control element (MAC-CE) configuration, a downlink control information (DCI) configuration, and / or a beam configuration. The beam configuration includes a configuration of a first type of beam in a beam set A and a configuration of a second type of beam in a beam set B, and the first type of beam and the second type of beam are both transmitted by the network device.

[0017] For example, the configuration of the first type of beam in the beam set A can include the number of each first type of beam in the beam set A, the transmission order of each first type of beam, and / or the orientation of each first type of beam, etc. The configuration of the second type of beam in the beam set B can include the number of each second type of beam in the beam set B, the transmission order of each second type of beam, and / or the orientation of each second type of beam, etc.

[0018] According to the scheme, the N categories of resource configurations can include N categories of beam configurations, and the data in the data set can be classified according to the N categories of beam configurations.

[0019] In some implementations of the first aspect, the data in the data set comprises channel information, and / or, beam information.

[0020] According to the above scheme, when the data in the data set is channel information, the training set containing the data set can be used for training of an AI model implementing a channel state information feedback enhancement function. When the data in the data set is beam information, the training set containing the data set can be used for training of an AI model implementing a beam management function.

[0021] In some implementations of the first aspect, the channel information comprises signal quality of a beam, and the signal quality of the beam comprises one or more of: reference signal received power (RSRP), signal to interference plus noise ratio (SINR), and reference signal received quality (RSRQ). The RSRP comprises: physical layer reference signal received power (L1-RSRP), and / or, layer three reference signal received power (L3-RSRP).

[0022] According to the above scheme, the data in the training set for AI model training can be signal quality of a beam.

[0023] In some implementations of the first aspect, before the first information is sent, the method further comprises: sending second information, the second information being used to indicate reporting of data for model training; and receiving the data for model training.

[0024] According to the above scheme, before the first information is sent, data in each of a plurality of categories for model training can be indiscriminately collected, so that a category with insufficient data is selected from the plurality of categories that have been collected, so that targeted data collection is performed on the category with insufficient data, so that the amount of data in each category in the training set for model (such as AI model) training is sufficient, i.e., the balance of data features in the training set is good. The specific implementation principles and technical effects of the embodiments of the present application can be found in the specific implementation principles and technical effects of S701-S704.

[0025] In some implementations of the first aspect, the first information is further used to indicate reporting of data of an (N+1)th category, and the data set comprises data of a first category to data of an Nth category, N being an integer greater than 2.

[0026] Exemplarily, the data of the (N+1)th category can be new data generated by the communication system due to changes in the communication environment. The (N+1)th category can be different from the first category to the Nth category. The data of the first category to the data of the Nth category can all be data of known categories generated in the communication system before the changes in the communication environment.

[0027] According to the scheme, the data of the N+1th category can be collected, so that the amount of data of the N+1th category in the training set containing the data of the N+1th category is sufficient. In the case that the data of the N+1th category is new data caused by the change of the communication environment, the AI model is trained by using the training set containing sufficient data of the N+1th category, so that the generalization of the AI model is better, for example, the performance of the AI model in the changed communication environment is better. The data of the N+1th category can be understood as being collected.

[0028] In combination with the first aspect, in some implementations of the first aspect, the first information includes an identifier of the N+1th category.

[0029] According to the scheme, the data of the N+1th category corresponding to the identifier of the N+1th category reported can be indicated.

[0030] In combination with the first aspect, in some implementations of the first aspect, the data of the N+1th category is received.

[0031] According to the scheme, the data of the N+1th category can be collected.

[0032] In combination with the first aspect, in some implementations of the first aspect, the data of the N+1th category is determined based on the ith feature, the ith feature is included in the features of the jth category data in the N categories of the data set, 1≤i and i are integers, 1≤j≤N and j are integers.

[0033] According to the scheme, the data of the N+1th category is determined based on the ith feature, that is, the data features of the N+1th category can be determined based on the ith feature. The data of the N+1th category can be identified or determined.

[0034] In combination with the first aspect, in some implementations of the first aspect, the first value determined based on the value associated with the kth feature of the data of the N+1th category and the value associated with the ith feature of the data of the jth category is less than the fifth threshold value, the first value is used to indicate the correlation between the value associated with the kth feature and the value associated with the ith feature, 1≤k and k are integers.

[0035] According to the scheme, the data of the N+1th category can be identified or determined.

[0036] In combination with the first aspect, in some implementations of the first aspect, the jth category includes the signal quality of each beam of a plurality of beams, the ith feature is the maximum signal quality in the jth category, the N+1th category includes the signal quality of the beam, and the absolute value of the difference between the signal quality of the N+1th category and the maximum signal quality is greater than the tenth threshold value.

[0037] According to the above scheme, the identification or determination of the N+1th category of data can be implemented.

[0038] In combination with the first aspect, in some implementations of the first aspect, the N+1th category includes signal quality of each of the plurality of beams, and an absolute value of a difference between two largest signal qualities in the N+1th category is greater than a sixth threshold. Alternatively, the N+1th category includes a signal to interference plus noise ratio (SINR), and an absolute value of a difference between the SINR of the N+1th category and a second value of a first preset range is greater than a seventh threshold, the second value being a minimum value or a maximum value of the first preset range.

[0039] According to the above scheme, the identification or determination of the N+1th category of data can be implemented.

[0040] In combination with the first aspect, in some implementations of the first aspect, before the first information is sent, the method further includes: sending third information, the third information being used to indicate data of the N+1th category and data of at least one of the first category to the Nth category of the data set, the N+1th category being different from the first category to the Nth category, N being an integer greater than 2. The data of the N+1th category and the data of at least one of the first category to the Nth category are received.

[0041] According to the above scheme, before the first information is sent, the data of each of the first category to the Nth category can be collected without distinction, so as to select a category (e.g., the first category) with insufficient data amount from the collected plurality of categories, and the new data caused by the change of the communication environment can also be collected. The new data caused by the change of the communication environment is, for example, the data of the N+1th category. In the case where the category with insufficient data amount is selected and the new data caused by the change of the communication environment is determined, the data of the category with insufficient data amount and the N+1th category can be collected in a targeted manner. The targeted collection of the N+1th category can be understood as the targeted collection of the new data. This makes the data amount of each category in the training set including the data of the first category to the data of the N+1th category sufficient, i.e., the balance of the data features of the training set is good. This can further make the generalization of the AI model trained by using the training set good. In the case where the data of the N+1th category is the new data caused by the change of the communication environment, the AI model trained by using the training set can perform well when applied in the changed communication environment. The specific implementation principles and technical effects of the embodiments of the present application can be referred to the specific implementation principles and technical effects of S801-S807.

[0042] In combination with the first aspect, in some implementations of the first aspect, the third information includes an identifier of each of the first category to the N+1th category.

[0043] According to the above scheme, data of each of the first category to the N+1th category can be reported.

[0044] With reference to the first aspect, in some implementations of the first aspect, the first information comprises an identifier of the first category.

[0045] According to the above scheme, data of the first category corresponding to the identifier of the first category can be reported.

[0046] In a second aspect, a data transmission method is provided, applied to a second device, the method comprising: receiving first information, the first information being used to indicate reporting of data of a first category, the first category being one of a plurality of categories, data of each of the plurality of categories being used for model training; and sending data of the first category.

[0047] The specific implementation principles of the embodiments of the present application can refer to the specific implementation principles of the embodiments shown in S101-S102. The model in the embodiments of the present application can be an AI model. The specific implementation principles of the embodiments of the present application can also refer to the specific implementation principles of the embodiments shown in S706-S709 or the embodiments shown in S808-S811.

[0048] According to the above scheme, data of the first category in the plurality of categories is collected in a targeted manner. The amount of data of each of the plurality of categories used for model (such as AI model) training can be sufficient, which can reduce the probability of poor generalization of the trained model due to insufficient data amount of a category in the training set or poor balance of data characteristics of the training set. For example, in the case where the terminal device uses an AI model for beam management (such as optimal beam prediction), the probability of poor signal quality of the optimal beam predicted by the terminal device using the AI model due to poor generalization of the model can be reduced, and thus the probability of the terminal device feeding back the signal quality of the optimal beam to the network device can be reduced.

[0049] With reference to the second aspect, in some implementations of the second aspect, before receiving the first information, the method further comprises: receiving second information, the second information being used to indicate reporting of data used for model training; and sending data used for model training.

[0050] According to the above scheme, the data of each category in the plurality of categories for model training can be indiscriminately collected before the data of the category with insufficient data is collected, so as to select the category with insufficient data from the plurality of categories, so as to perform targeted data collection on the category with insufficient data, so that the data of each category in the training set for model (such as AI model) training is sufficient, that is, the balance of the data features of the training set is better. The specific implementation principles and technical effects of the embodiments of the present application can be referred to the specific implementation principles and technical effects of S701-S704.

[0051] With reference to the second aspect, in some implementations of the second aspect, the first information is further used to indicate that the data of the (N+1)th category is reported, the data set includes the data of the first category to the data of the Nth category, N is an integer greater than 2, the data set is used for model training, and the data in the data set is reported before the first information is received. The method further includes: sending the data of the (N+1)th category.

[0052] According to the above scheme, the data of the (N+1)th category can be collected in a targeted manner, so that the amount of data of the (N+1)th category in the training set containing the data of the (N+1)th category is sufficient. In the case that the data of the (N+1)th category is new data caused by changes in the communication environment, using the training set containing sufficient data of the (N+1)th category to train the AI model can make the generalization of the AI model better, for example, make the AI model perform better when applied in the changed communication environment.

[0053] With reference to the second aspect, in some implementations of the second aspect, the first information is further used to indicate a data reporting condition. The sending of the data of the (N+1)th category includes: sending the data of the (N+1)th category in the case that the data reporting condition is met.

[0054] According to the above scheme, the data of the (N+1)th category in the training set can be collected.

[0055] With reference to the second aspect, in some implementations of the second aspect, the data reporting condition can include at least one of the following: obtaining the data of the (N+1)th category, then reporting the data of the (N+1)th category; obtaining the data of the (N+1)th category more than or equal to an eighth threshold value, then reporting the data of the (N+1)th category; obtaining the data amount of the data of the (N+1)th category is greater than or equal to a ninth threshold value, then reporting the data of the (N+1)th category.

[0056] According to the above scheme, the collection of the data of the N+1th category in the training set can be implemented. In this way, when the number of times of obtaining the data of the N+1th category is greater than or equal to the eighth threshold value, the data of the N+1th category is reported. Alternatively, when the data amount of the data of the N+1th category is greater than or equal to the ninth threshold value, the data of the N+1th category is reported. In this way, the reporting frequency of the second device can be reduced, and the interaction frequency can be reduced.

[0057] With reference to the second aspect, in some implementations of the second aspect, before receiving the first information, the method further includes: receiving third information, the third information being used to indicate that the data of the N+1th category and the data of at least one category of the first category to the Nth category in the data set are reported, and the N+1th category is different from the first category to the Nth category. The data of the N+1th category and the data of at least one category of the first category to the Nth category are transmitted.

[0058] According to the above scheme, before the data of the first category and the data of the N+1th category are collected in a targeted manner, the data of each category of the first category to the Nth category can be collected indiscriminately, so as to select a category (for example, the first category) with insufficient data amount from the collected multiple categories, and the new data generated due to the change of the communication environment can also be collected. In this way, when the category with insufficient data amount is selected and it is determined that the new data is generated due to the change of the communication environment, the data of the first category and the N+1th category can be collected in a targeted manner. In this way, the data amount of each category of the data of the first category to the data of the N+1th category in the training set is sufficient, that is, the balance of the data characteristics of the training set is good. In this way, the generalization of the AI model trained by using the training set is good. When the data of the N+1th category is the new data generated due to the change of the communication environment, the AI model trained by using the training set can perform well when applied in the changed communication environment. The specific implementation principles and technical effects of the embodiments of the present application can be referred to the specific implementation principles and technical effects of S801-S807.

[0059] With reference to the second aspect, in some implementations of the second aspect, the third information is further used to indicate a data reporting condition. The data of the N+1th category and the data of at least one category of the first category to the Nth category are transmitted, including: the data of the N+1th category is transmitted when the data reporting condition is met.

[0060] According to the above scheme, the collection of the data of the N+1th category in the training set can be implemented.

[0061] With reference to the second aspect, in some implementations of the second aspect, the data reporting condition can include at least one of the following: the N+1th category of data is obtained, and the N+1th category of data is reported; the number of times that the N+1th category of data is obtained is greater than or equal to an eighth threshold value, and the N+1th category of data is reported; and the amount of data of the N+1th category of data obtained is greater than or equal to a ninth threshold value, and the N+1th category of data is reported.

[0062] According to the above scheme, the collection of the N+1th category of data in the training set can be implemented. The number of reporting times of the second device reporting data can be reduced, and the number of interactions can be reduced.

[0063] In a third aspect, a data transmission method is provided, applied to a third device, and the method includes: receiving first information from a first device, the first information being used to indicate reporting of a first category of data, the first category being one of a plurality of categories, and data of each category in the plurality of categories being used for model training; sending the first information to a second device; receiving the first category of data from the second device; and sending the first category of data to the first device.

[0064] The specific implementation principles of the embodiments of the present application can refer to the specific implementation principles of the embodiments shown in S101-S102. The model in the embodiments of the present application can be an AI model.

[0065] According to the above scheme, the first category of data in the plurality of categories is collected. The amount of data of each category in the plurality of categories used for AI model training can be sufficient, and the probability of the trained model having poor generalization can be reduced due to the presence of a category with insufficient data in the training set or poor data feature balance of the training set.

[0066] In a fourth aspect, a first device is provided, and the first device includes one or more processors and a memory. The memory is coupled to the one or more processors, and the memory is used to store computer program code including computer instructions. The one or more processors invoke the computer instructions to cause the first device to perform the method of the above first aspect and any possible implementation manner of the first aspect.

[0067] In a fifth aspect, a second device is provided, and the second device includes one or more processors and a memory. The memory is coupled to the one or more processors, and the memory is used to store computer program code including computer instructions. The one or more processors invoke the computer instructions to cause the second device to perform the method of the above second aspect and any possible implementation manner of the second aspect.

[0068] In a sixth aspect, a third device is provided, the third device comprising one or more processors and a memory. The memory is coupled to the one or more processors, and the memory is configured to store computer program codes comprising computer instructions. The one or more processors are configured to invoke the computer instructions to cause the third device to perform the method in the third aspect.

[0069] In a seventh aspect, a chip system is provided, the chip system being applied to the first device, the second device or the third device. The chip system comprises one or more processors. The one or more processors are configured to invoke computer instructions to cause the first device to perform the method in the first aspect and any possible implementation manner of the first aspect. Or, the one or more processors are configured to invoke the computer instructions to cause the second device to perform the method in the second aspect and any possible implementation manner of the second aspect. Or, the one or more processors are configured to invoke the computer instructions to cause the third device to perform the method in the third aspect.

[0070] In an eighth aspect, a computer readable storage medium is provided, the computer readable storage medium comprising computer instructions. When the computer instructions are run on the first device, the computer instructions cause the first device to perform the method in the first aspect and any possible implementation manner of the first aspect. Or, when the computer instructions are run on the second device, the computer instructions cause the second device to perform the method in the second aspect and any possible implementation manner of the second aspect. Or, when the computer instructions are run on the third device, the computer instructions cause the third device to perform the method in the third aspect.

[0071] In a ninth aspect, a computer program product is provided, the computer program product comprising computer program codes. When the computer program codes are run on the first device, the computer program codes cause the first device to perform the method in the first aspect and any possible implementation manner of the first aspect. Or, when the computer program codes are run on the second device, the computer program codes cause the second device to perform the method in the second aspect and any possible implementation manner of the second aspect. Or, when the computer program codes are run on the third device, the computer program codes cause the third device to perform the method in the third aspect.

[0072] It should be understood that the fourth aspect, the seventh aspect to the ninth aspect of the present application correspond to the technical solution of the first aspect of the present application, the fifth aspect, the seventh aspect to the ninth aspect of the present application correspond to the technical solution of the second aspect of the present application, and the sixth aspect to the ninth aspect of the present application correspond to the technical solution of the third aspect of the present application. The beneficial effects achieved by each aspect and the corresponding possible implementation manners are similar, and will not be described again. BRIEF DESCRIPTION OF DRAWINGS

[0073] Figure 1 A schematic diagram of a communication system provided by an embodiment of the present application;

[0074] Figure 2 Another schematic diagram of a communication system provided by an embodiment of the present application is shown in FIG. 3.

[0075] Figure 3 Another schematic diagram of an interaction process of a data transmission method provided by an embodiment of the present application is shown in FIG. 6.

[0076] Figure 4 Another schematic diagram of a spatial relationship between beam set A and beam set B provided by an embodiment of the present application is shown in FIG. 7.

[0077] Figure 5 Another schematic diagram of a spatial relationship between beam set A and beam set B provided by an embodiment of the present application is shown in FIG. 7.

[0078] Figure 6 A schematic diagram of a data collection scenario provided by an embodiment of the present application is shown in FIG. 8.

[0079] Figure 7 Another schematic diagram of an interaction process of a data transmission method provided by an embodiment of the present application is shown in FIG. 6.

[0080] Figure 8 Another schematic diagram of an interaction process of a data transmission method provided by an embodiment of the present application is shown in FIG. 6.

[0081] Figure 9 Another schematic diagram of an interaction process of a data transmission method provided by an embodiment of the present application is shown in FIG. 6. DETAILED DESCRIPTION

[0082] The technical solutions in the embodiments of the present application will be described below with reference to the drawings.

[0083] In the embodiments of the present application, the same items or similar items with basically the same functions and effects are distinguished by using "first", "second", etc. For example, the first chip and the second chip are only used to distinguish different chips, and do not limit the sequence. Those skilled in the art can understand that "first", "second", etc. do not limit the quantity and execution sequence, and "first", "second", etc. also do not necessarily mean different.

[0084] It should be noted that in the embodiments of the present application, the words "exemplary" or "for example" are used to mean serving as an example, instance, or illustration. Any embodiment or design scheme described as "exemplary" or "for example" in the present application should not be interpreted as being more preferred or having more advantages than other embodiments or design schemes. Rather, the words "exemplary" or "for example" are used in the sense of presenting a specific way of realizing the relevant concept.

[0085] In the embodiments of this application, "at least one" means one or more, "multiple" can be understood as "at least two", "multiple items" can be understood as "at least two items". "And / or" describes the association relationship of the associated objects, which means that there can be three kinds of relationships, for example, A and / or B can represent the following three cases: A exists alone, A and B exist together, and B exists alone, where A and B can be singular or plural. The character " / " generally represents an "or" relationship between the associated objects before and after it. "At least one of the following" or similar expressions means any combination of these items, including any combination of single item or multiple items. For example, at least one of a, b, or c can represent a, b, c, a-b, a-c, b-c, or a-b-c, where a, b, and c can be single or multiple.

[0086] The technical solutions of the embodiments of this application can be applied to a communication system. The communication system includes but is not limited to the following systems, for example: a 5th generation (5G) or new radio (NR) system, a long term evolution (LTE) system, an LTE frequency division duplex (FDD) system, an LTE time division duplex (TDD) system, a wireless local area network (WLAN) system, a satellite communication system, a future communication system such as a 6th generation (6G) mobile communication system, or a fusion system of multiple systems, etc. The technical solutions provided in this application can also be applied to device-to-device (D2D) communication, vehicle-to-everything (V2X) communication, machine-to-machine (M2M) communication, machine type communication (MTC), and internet of things (IoT) communication system or other communication systems, etc.

[0087] A network element in a communication system can send a signal to another network element or receive a signal from another network element. The signal can include information, signaling, or data, etc. The network element can also be replaced by an entity, a network entity, a device, a communication device, a communication module, a node, a communication node, etc. In the embodiments of this application, the network element can be taken as an example for description. For example, the communication system can include at least one terminal device and at least one network device. The network device can send a downlink signal to the terminal device, and / or the terminal device can send an uplink signal to the network device.

[0088] In a wireless communication network, e.g., in a mobile communication network, the services supported by the network are more and more diverse, and thus the requirements to be met are more and more diverse. For example, the network needs to be able to support high rates, ultra-low latency, and / or ultra-large connectivity. This feature makes network planning, network configuration, and / or resource scheduling more and more complex. In addition, as the functions of the network are more and more powerful, e.g., supporting higher and higher spectrum, supporting high-order multiple input multiple output (MIMO) technology, supporting beamforming, and / or supporting new technologies such as beam management, etc., network energy saving has become a hot research topic. These new requirements, new scenarios, and new features bring unprecedented challenges to network planning, operation and maintenance, and efficient operation. In order to meet this challenge, artificial intelligence (AI) technology can be introduced into wireless communication, so as to realize network intelligence and realize, for example, modulation and demodulation of signals, information encoding and decoding, channel state information feedback, beam management, or mobility management, etc. In order to support AI technology in the wireless network, an AI node can also be introduced into the network.

[0089] Referring to Figure 1 , Figure 1 A schematic diagram of a communication system suitable for the communication method provided by the embodiments of the present application is shown. As Figure 1 shown, the communication system 100 can include at least one network device, e.g., the network device 110 shown in Figure 1 ; the communication system 100 can also include at least one terminal device, e.g., the terminal device 120 shown in Figure 1 . The network device 110 and the terminal device 120 can communicate through a wireless link. The communication devices in the communication system, e.g., the network device 110 and the terminal device 120, can communicate through multi-antenna technology.

[0090] Referring to Figure 2 , Figure 2 is a schematic diagram of another communication system suitable for the communication method provided by the embodiments of the present application. As Figure 2 shown, compared with the communication system 100 shown in Figure 1 , the communication system 200 shown in Figure 2 also includes an AI network element 210. The AI network element 210 is used to perform AI-related operations, e.g., constructing a training data set or training an AI model, etc.

[0091] In a possible implementation, the network device 110 can send data related to training of the AI model to the AI network element 210, the AI network element 210 constructs a training data set and trains the AI model. For example, the data related to training of the AI model can include data reported by the terminal device. The AI network element 210 can send a result of an operation related to the AI model to the network device 110, and forward the result to the terminal device through the network device 110. For example, the result of the operation related to the AI model can include at least one of the following: a trained AI model, an evaluation result or a test result of the model, and the like. For example, part of the trained AI model can be deployed on the network device 110, and the other part can be deployed on the terminal device. Alternatively, the trained AI model can be deployed on the network device 110, or the trained AI model can be deployed on the terminal device.

[0092] It should be understood that, Figure 2 Only the case that the AI network element 210 is directly connected to the network device 110 is described, and in other scenarios, the AI network element 210 can also be connected to the terminal device. Alternatively, the AI network element 210 can be connected to both the network device 110 and the terminal device. Alternatively, the AI network element 210 can also be connected to the network device 110 through a third-party network element. The connection relationship between the AI network element and other network elements is not limited in the embodiments of the present application.

[0093] The AI network element 210 can also be arranged as a module in the network device and / or the terminal device, for example, in the network device 110 or the terminal device 120 as shown. Figure 1 It should be understood that,

[0094] It should be noted that, Figure 1 and Figure 2 The simplified schematic diagram is only an example for understanding, for example, the communication system can also include other devices, such as a wireless relay device and / or a wireless backhaul device, and the like, Figure 1 and Figure 2 are not shown in the figure. In actual application, the communication system can include multiple network devices, and can also include multiple terminal devices. The number of network devices and terminal devices included in the communication system is not limited in the embodiments of the present application.

[0095] In the embodiments of the present application, the terminal device can also be referred to as a user equipment (UE), an access terminal, a user unit, a user station, a mobile station, a mobile station, a remote station, a remote terminal, a mobile device, a user terminal, a terminal, a wireless communication device, a user agent or a user device.

[0096] The terminal device can be a device providing voice / data, for example, a handheld device with wireless connection function, a vehicle-mounted device, etc. At present, some examples of terminals are: mobile phone, pad, notebook computer, palm computer, mobile internet device (MID), wearable device, virtual reality (VR) device, augmented reality (AR) device, wireless terminal in industrial control, wireless terminal in self driving, wireless terminal in remote medical surgery, wireless terminal in smart grid, wireless terminal in transportation safety, wireless terminal in smart city, wireless terminal in smart home, cellular phone, cordless phone, session initiation protocol (SIP) phone, wireless local loop (WLL) station, personal digital assistant (PDA), handheld device with wireless communication function, computing device or other processing device connected to a wireless modem, wearable device, terminal device in a 5G network, or terminal device in a future evolved public land mobile network (PLMN), etc. The embodiments of the present application are not limited thereto.

[0097] By way of example and not limitation, in the 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, which is a general term for devices that are designed and developed by applying wearable technology to daily wear, such as glasses, gloves, watches, clothing, and shoes. 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 device that realizes powerful functions through software support and data interaction and cloud interaction. The general wearable smart device includes a device with full functions and large size, which can realize complete or partial functions without relying on a smart phone, such as a smart watch or smart glasses, and a device that focuses on a certain application function and needs to cooperate with other devices such as a smart phone, such as various smart wristbands and smart jewelry for monitoring vital signs.

[0098] In the embodiments of the present application, the apparatus for implementing the function of the terminal device can be a terminal device, or can be an apparatus capable of supporting the terminal device to implement the function, for example, a chip system, which can be installed in the terminal device or used in matching with the terminal device. In the embodiments of the present application, the chip system can be composed of a chip, or can include the chip and other discrete devices. In the embodiments of the present application, only the apparatus for implementing the function of the terminal device is taken as an example for description, and the scheme of the embodiments of the present application is not limited in this way.

[0099] The network device in the embodiments of the present application can be a device for communicating with a terminal device, and the network device can also be an access network device or a radio access network device, for example, the network device can be a base station. The network device in the embodiments of the present application can be a radio access network (RAN) node (or device) for accessing a terminal device to a wireless network. The base station can broadly cover various names in the following or be replaced by the following names, such as: Node B (NodeB), evolved Node B (eNB), next generation Node B (gNB), relay station, access point, transmitting and receiving point (TRP), transmitting point (TP), primary station, secondary station, motor slide retainer (MSR) node, home base station, network controller, access node, wireless node, access point (AP), transmission node, transceiver node, baseband unit (BBU), remote radio unit (RRU), active antenna unit (AAU), remote radio head (RRH), central unit (CU), distributed unit (DU), radio unit (RU), or positioning node, etc. The base station can be a macro base station, a micro base station, a relay node, a donor node, or the like, or a combination thereof. The base station can also refer to a communication module, modem, or chip used in the aforementioned device or apparatus. The base station can also be a mobile switching center and a device assuming a base station function in D2D, V2X, M2M communication, a network side device in a 6G network, a device assuming a base station function in a future communication system, etc. The base station can support networks of the same or different access technologies. Optionally, the RAN node can also be a server, a wearable device, a vehicle or a vehicle-mounted device, etc. For example, the access network device in vehicle to everything (V2X) technology can be a road side unit (RSU). The embodiments of the present application do not limit the specific technology and specific device form of the network device.

[0100] The base station can be fixed or mobile. For example, a helicopter or a drone can be configured to act as a mobile base station, and one or more cells can move according to the location of the mobile base station. In other examples, a helicopter or a drone can be configured to act as a device communicating with another base station.

[0101] In some deployments, the network devices mentioned by embodiments of the present application, for example, access network devices, radio access network devices, RAN nodes or RAN devices. The network device can be a device including a CU, or a DU, or a device including a CU and a DU, or a control plane CU node (central unit-control plane, CU-CP) and a user plane CU node (central unit-user plane, CU-UP) and a DU node. For example, the network device can include a gNB-CU-CP, a gNB-CU-UP and a gNB-DU. The control plane CU node can also be referred to as a central unit control plane. The user plane CU node can also be referred to as a central unit user plane.

[0102] In some deployments, a plurality of RAN nodes cooperate to assist terminals to implement wireless access, and different RAN nodes respectively implement part of the functions of a base station. For example, the RAN node can be a CU, a DU, a CU-CP, a CU-UP, or an RU, etc. The CU and the DU can be separately arranged, or can also be included in the same network element, for example, in a BBU. The RU can be included in a radio frequency device or a radio frequency unit, for example, included in an RRU, an AAU or an RRH.

[0103] The RAN node can support one or more types of fronthaul interfaces, different fronthaul interfaces respectively corresponding to DUs and RUs having different functions. If the fronthaul interface between the DU and the RU is a common public radio interface (CPRI), the DU is configured to implement one or more of baseband functions, and the RU is configured to implement one or more of radio frequency functions. If the fronthaul interface between the DU and the RU is another interface, relative to the CPRI, one or more of the partial baseband functions of the downlink and / or uplink, such as, for the downlink, one or more of precoding, digital beamforming (BF), or inverse fast Fourier transform (IFFT) / add cyclic prefix (CP), are moved from the DU to the RU for implementation, and for the uplink, one or more of digital beamforming (BF), or fast Fourier transform (FFT) / remove cyclic prefix (CP) are moved from the DU to the RU for implementation. In a possible implementation, the interface can be an enhanced common public radio interface (eCPRI). Under the eCPRI architecture, the splitting manner between the DU and the RU is different, corresponding to different categories (Cat) of eCPRI, such as eCPRI Cat A, B, C, D, E, F.

[0104] Taking eCPRI Cat A as an example, for downlink transmission, with layer mapping as the cut, the DU is configured to implement layer mapping and one or more functions (i.e., one or more of encoding, rate matching, scrambling, modulation, layer mapping) before layer mapping, while other functions (e.g., one or more of resource element (RE) mapping, digital beamforming (BF), or inverse fast Fourier transform (IFFT) / adding cyclic prefix (CP)) after layer mapping are implemented in the RU. For uplink transmission, with de-RE mapping as the cut, the DU is configured to implement de-mapping and one or more functions (i.e., one or more of decoding, de-rate matching, de-scrambling, de-modulation, inverse discrete Fourier transform (IDFT), channel equalization, de-RE mapping) before de-mapping, while other functions (e.g., one or more of digital BF or fast Fourier transform (FFT) / CP removal) after de-mapping are implemented in the RU. It can be understood that the function description of the DU and the RU corresponding to various types of eCPRI can refer to the eCPRI protocol, which is not described here.

[0105] In a possible design, the processing unit in the BBU for implementing baseband functions is referred to as a baseband high (BBH) unit, and the processing unit in the RRU / AAU / RRH for implementing baseband functions is referred to as a baseband low (BBL) unit.

[0106] 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 the ORAN system, the CU can also be referred to as an O-CU (open CU), the DU can also be referred to as an O-DU, the CU-CP can also be referred to as an O-CU-CP, the CU-UP can also be referred to as an O-CU-UP, and the RU can also be referred to as an O-RU. Any of the CU (or CU-CP, CU-UP), DU and RU in this application can be implemented by a software module, a hardware module, or a combination of a software module and a hardware module.

[0107] In the embodiments of the present application, the apparatus for implementing the function of the network device can be the network device, or can be an apparatus capable of supporting the network device to implement the function, such as a chip system, a hardware circuit, a software module, or a hardware circuit plus a software module. The apparatus can be installed in the network device or used in combination with the network device. In the embodiments of the present application, only the apparatus for implementing the function of the network device is taken as an example for description, and the scheme of the embodiments of the present application is not limited in this way.

[0108] The network device and / or the terminal device can be deployed on land, including indoors or outdoors, handheld or vehicle-mounted; can also be deployed on water; and can also be deployed on airplanes, balloons and satellites in the air. The scenarios in which the network device and the terminal device are located are not limited in the embodiments of the present application. In addition, the terminal device and the network device can be hardware devices, or can be software functions running on special hardware, software functions running on general hardware, such as virtualized functions instantiated on a platform (for example, a cloud platform), or entities including special or general hardware devices and software functions. The specific forms of the terminal device and the network device are not limited in the present application.

[0109] The communication system to which the technical scheme of the embodiments of the present application is applicable can also include other network elements or entities. As an example, the communication system to which the technical scheme of the embodiments of the present application is applicable can also include a core network, which can include one or more of the following entities: an access and mobility management function (AMF) entity, a session management function (SMF) entity, a unified data management (UDM) network element, or a user plane function (UPF) entity, etc. The network elements or entities in the core network can be referred to as core network devices. The network elements or entities in the core network can have other names, which are not limited in the present application. The communication system to which the technical scheme of the embodiments of the present application is applicable can also include one or more devices in operation administration and maintenance (OAM). The devices in the OAM can be referred to as OAM devices.

[0110] Optionally, the AI node can be deployed in one or more of the following positions in the communication system: an access network device, a terminal device, or a core network device, etc., or the AI node can also be deployed separately, for example, in a position other than any of the above devices, such as a host or a cloud server of an over the top (OTT) system. The AI node can communicate with other devices in the communication system, which can be one or more of the following: an access network device, a terminal device, or a network element of a core network, etc.

[0111] It can be understood that the number of AI nodes is not limited in the present application. For example, when there are multiple AI nodes, the multiple AI nodes can be divided based on functions, such as different AI nodes being responsible for different functions.

[0112] It can also be understood that the AI node can be a separate device, can be integrated into the same device to implement different functions, or can be a network element in a hardware device, or can be a software function running on a dedicated hardware, or a virtualized function instantiated on a platform (e.g., a cloud platform), and the specific form of the AI node is not limited in the present application.

[0113] The AI node can be an AI network element or an AI module.

[0114] Taking the AI node as an AI module as an example, the AI module is used to implement a corresponding AI function. The AI modules deployed in different network elements can be the same or different. The model of the AI module can implement different functions according to different parameter configurations. The model of the AI module can be configured based on one or more of the following parameters: a structure parameter (such as at least one of the number of neural network layers, the width of the neural network, the connection relationship between layers, the weight of neurons, the activation function of neurons, or the bias in the activation function), an input parameter (such as the type of input parameter and / or the dimension of input parameter), or an output parameter (such as the type of output parameter and / or the dimension of output parameter). The bias in the activation function can also be referred to as the bias of the neural network.

[0115] One AI module can have one or more models. One model can infer an output, which includes one parameter or multiple parameters. The learning process, the training process, or the inference process of different models can be deployed in different nodes or devices, or can be deployed in the same node or device.

[0116] The network device can be a network device provided with one or more AI modules. The network device can be a core network device, an access network device (such as a RAN node), or an OAM device.

[0117] Optionally, in the case that the AI node is a standalone device, the AI node can be a cloud server. The cloud server can be a server on the terminal device side, for example, a server on the UE side. The server on the UE side can be a server of a chip manufacturer, a server of a terminal manufacturer, or a server of a network device manufacturer.

[0118] It can be understood that the AI node in the embodiments of the present application can be replaced by a first network element or a first device, the terminal device can be replaced by a second network element or a second device, and the network device can be replaced by a third network element or a third device, and the three perform corresponding methods in the embodiments of the present application.

[0119] It can be understood that the AI node in the embodiments of the present application can be replaced by a first network element or a first device, the terminal device can be replaced by a second network element or a second device, and the network device can be replaced by a third network element or a third device, and the three perform corresponding methods in the embodiments of the present application.

[0120] 1. AI model

[0121] The AI model can be understood as a function model that maps a certain dimension of input to a certain dimension of output, and the model parameters are obtained through machine learning training. For example, f(x) = ax2 + b is a quadratic function model, which can be regarded as an AI model, and a and b correspond to the parameters of the model, which can be obtained through machine learning training.

[0122] 2. AI application case (use case)

[0123] The AI application case can be referred to as an AI application scenario or an AI function. The AI application case includes but is not limited to: channel status information (CSI) feedback enhancement, beam management enhancement, positioning accuracy enhancement, network energy saving, load balancing, and mobility optimization. The following will be described respectively. Exemplarily, one AI function can include multiple AI sub-functions.

[0124] In the embodiments of the present application, the AI application scenario can be understood as an application scenario of an AI model. The AI function can be understood as a function of the AI model. The AI function can be one of CSI feedback, beam management, positioning enhancement, mobility management, modulation, demodulation, encoding, decoding, channel equalization, channel estimation, pilot generation, precoding, resource mapping, resource demapping, interference suppression, interference estimation, interference prediction, receiver or transmitter, or a combination of multiple functions, and will not be listed one by one.

[0125] In some embodiments of the present application, CSI feedback can also be referred to as CSI feedback enhancement. Beam management can also be referred to as beam management enhancement.

[0126] The AI application case can be referred to as AI use case, for example, the AI use case can include CSI use case, beam management (BM) use case, positioning use case (such as positioning enhancement) and the like. The CSI use case, for example, CSI feedback enhancement or CSI feedback. The BM use case, for example, beam management enhancement or beam management. The positioning use case, for example, positioning enhancement.

[0127] 3. CSI feedback enhancement

[0128] CSI is the channel property of the communication link, and is the channel quality information reported by the terminal device to the network device. The terminal device reports the channel quality information to the network device, so as to select a suitable modulation and coding scheme (MCS) for the terminal device, so as to adapt to the changing wireless channel. For example, the terminal device performs channel estimation according to the received channel state information-reference signal (CSI-RS), and then feeds back the channel quality information to the network device. The information is used as the input of the model of the network device, so that the network device can realize AI model training. By applying AI to CSI feedback enhancement, the overhead can be reduced, the accuracy can be improved, and prediction can be realized.

[0129] The CSI-RS feedback enhancement can include at least one sub-function, such as CSI compression, CSI prediction, and CSI-RS configuration signaling reduction respectively. Among them, the CSI compression can be divided into CSI compression in at least one of the spatial domain, the time domain and the frequency domain.

[0130] 4. Beam management enhancement

[0131] BM is mainly to find the strongest transmit / receive beam pair. Based on AI-based sparse beam prediction, accuracy can be improved. According to AI training and inference, it can be divided into network-side AI sparse beam prediction and terminal device-side AI sparse beam prediction. Taking the terminal device-side AI sparse beam prediction as an example, the pre-trained AI model of the terminal device side can be issued by the network side or pre-stored by the terminal device side. In the training phase, the network device scans all possible beams, and then the network reports the transmit beam pattern to the terminal device. When the model training is completed, the network only needs to scan a small part of the beam, and then the terminal device feeds back the inference result to the network. Based on AI-based beam management, beam prediction in time and / or spatial domain can be realized to reduce overhead and delay and improve beam selection accuracy.

[0132] Beam management enhancement can include at least one sub-function, such as beam scanning matrix prediction and optimal beam prediction, respectively.

[0133] 5. Positioning enhancement

[0134] In line of sight (LOS) or non-line of sight (NLOS) scenarios, AI-based positioning can improve positioning accuracy with a smaller number of TRP antennas. Positioning enhancement can include at least one sub-function, such as access network device-based positioning enhancement, positioning management function network element-based positioning enhancement, and terminal device-based positioning enhancement.

[0135] 6. Network energy saving

[0136] Network energy saving can be achieved through cell activation / deactivation, load reduction, improved coverage, or other RAN setting adjustments. AI technology can be used to optimize energy saving decisions by utilizing data collected in the RAN network. AI algorithms can predict the energy efficiency and load status of the next period, which can be used to assist in decision-making for cell activation / deactivation to save energy. Based on the predicted load, the system can dynamically configure energy saving strategies to maintain a balance between system performance and energy efficiency and reduce energy consumption.

[0137] 7. Load balancing

[0138] Load balancing can make the load evenly distributed among cells and among areas within a cell, or divert part of the traffic from congested cells, or split users among cells, carriers or access technologies to improve network performance. AI model based load balancing can provide higher quality user experience and improve system capacity.

[0139] 8. Mobility management

[0140] Mobility management is a solution to ensure service continuity during terminal device movement by minimizing dropped calls, radio link failure (RLF), unnecessary handover and ping-pong effect. AI based mobility management can enhance, for example, reduce the probability of unexpected events, terminal device location / mobility / performance prediction and traffic steering.

[0141] It should be understood that the above definitions of various technical terms are only examples. For example, as technology continues to evolve, the scope of the above definitions can also change, and the embodiments of the present application are not limited.

[0142] From the above description of AI application cases, it can be seen that AI can be widely used in CSI feedback enhancement, beam management, positioning accuracy enhancement, energy saving, mobility enhancement and load balancing to improve network performance and / or terminal performance. The terminal performance can include the performance of the terminal device, and / or the performance of an entity associated with the terminal. The entity associated with the terminal can include a server on the terminal side, a computing and / or processing node on the terminal side, a computing and / or processing entity on the terminal side, and a computing and / or processing unit on the terminal side. The server on the terminal side, for example, an over-the-top (OTT) server, etc. OTT can be understood as an entity that provides various OTT services to users based on operator networks by third parties other than network operators. Exemplarily, OTT services can include OTT voice communication services, OTT multimedia services and / or OTT data processing services, etc. The terminal device can interact with relevant information (e.g. data) through communication with the entity associated with the terminal. For example, the entity associated with the terminal and the terminal device belong to the same manufacturer. Due to model training, model selection, etc., it can not be performed on the terminal, but on the OTT server on the terminal side, therefore, the terminal or terminal device in the embodiment can also include the OTT server on the terminal side.

[0143] It should be understood that the terminal in the embodiment can also be referred to as "terminal side" (UE side) or "terminal part" (UE part).

[0144] The AI model can be deployed on the network side and / or the terminal device side. The performance of the AI model depends on the training of the AI model. The training of the AI model depends on the collection of training data.

[0145] The training data set for AI model training can be referred to as a training set. The training set can include data of multiple categories. Since the training data in the training set mainly comes from the measurement and feedback of the terminal device. The differences in performance, configuration or deployment between terminal devices can cause some categories in the multiple categories of the training set to have more (or sufficient) data, and some categories to have less (or insufficient) data.

[0146] The sufficient data in each category of the multiple categories of the training set can represent that the balance of the data characteristics of the training set is good. The insufficient data in at least one category of the multiple categories of the training set can represent that the balance of the data characteristics of the training set is poor.

[0147] The better the balance of the data characteristics of the training set, the better the generalization of the AI model trained by the training set. The poorer the balance of the data characteristics of the training set, the poorer the generalization of the AI model trained by the training set.

[0148] Taking the UE using the AI model for beam management as an example, the poorer the generalization of the AI model, the more likely it is that the UE feeds back the optimal beam to the network device with poor signal quality, thereby affecting the data transmission between the UE and the network device.

[0149] Therefore, the embodiments of the present application provide a data transmission method applied to an AI node, which can receive data of a first category by sending first information indicating the reporting of the data of the first category. The data of the first category is one of multiple categories, and the data of each category of the multiple categories is used for AI model training. For example, the data of each category of the multiple categories is included in a training set. The training set is used for AI model training. In this way, in the case that the first category is a category with insufficient data in the training set, the data transmission method provided by the embodiments of the present application can collect data of the category with insufficient data, so that the data of each category in the training set is sufficient, that is, the balance of the data characteristics of the training set is good, and the generalization of the AI model trained by the training set is good.

[0150] Still taking the AI model for beam management as an example, the better the generalization of the AI model, the lower the probability that the optimal beam predicted by the terminal device (such as a UE) using the AI model has poor signal quality, that is, the lower the probability that the optimal beam fed back by the terminal device to the network device has poor signal quality, thereby reducing the probability of affecting the data transmission between the terminal device and the network device.

[0151] The data in the embodiments of the present application can include data obtained in a communication network, for example, signal processing information, channel information, and radio frequency information, etc.

[0152] The signal processing information can include information generated in a baseband signal processing process, for example, information generated in a signal sampling, modulation, demodulation, coding, decoding, precoding, resource mapping, and / or digital filtering process.

[0153] The channel information includes information corresponding to a channel environment, for example, at least one of power information, amplitude information, phase information, time delay information, multipath information, signal propagation time information, distance information, speed information, large-scale channel information, small-scale channel information, channel scattering information, LOS information, or NLOS information, etc.

[0154] Exemplarily, the data of the CSI use case and the BM use case can both be embodied as channel information CSI. The data in the positioning use case can be embodied as channel information and / or position information. The radio frequency information can include information generated in an analog processing process, for example, information generated in a digital-to-analog conversion, analog-to-digital conversion, digital pre-distortion, frequency conversion, radio frequency modulation, radio frequency demodulation, power amplification, low-noise amplification, analog filtering, and / or duplex processing.

[0155] Exemplarily, taking an AI node or a cloud server as the first device, a terminal device as the second device, and a network device as the third device as an example, Figure 3 A kind of interaction flow schematic diagram of the data transmission method provided by the embodiments of the present application is shown.

[0156] As Figure 3 The data transmission method provided by the embodiments of the present application can include S101-S102 as shown in the figure:

[0157] S101, the first device can send first information to the second device. Correspondingly, the second device can receive the first information from the first device. The first information is used to indicate the reporting of the first category of data. The first category is one of a plurality of categories. The data of each category in the plurality of categories is used for model training.

[0158] It should be understood that the model training is, for example: AI model training. The model can be an AI model.

[0159] Exemplarily, in the case that the first device is independent of the third device, the first device can send the first information to the third device. Correspondingly, the third device can receive the first information from the first device. The first device independent of the third device can be understood as that the first device does not have the function of the third device (or network device).

[0160] The third device can send the first information to the second device. Correspondingly, the second device can receive the first information sent by the third device.

[0161] In this way, the first information is forwarded by the third device, so that the second device can receive the first information from the first device.

[0162] It can be understood that, in the case that the first device has the function of a network device, the first information does not need to be forwarded by the third device. That is, the first device can send the first information to the second device without the third device. Correspondingly, the second device can receive the first information from the first device without the third device. The first device has the function of a network device, for example, an AI module is arranged on a network device, that is, the first device and the third device are the same device.

[0163] Exemplarily, the data of each category in the plurality of categories can be obtained by the first device before S101 is performed. For example, the first device obtains a data set before S101 is performed, and the data set can include data of each category in the plurality of categories. It should be understood that the data in the data set can be obtained by the first device from other devices performing model training, or can be collected by the first device. The other devices performing model training can not include the first device, the second device and the third device.

[0164] The first category can be a category with insufficient data in the plurality of categories. The data of the first category is a category of data used for AI model training.

[0165] Taking the plurality of categories as N categories and N being an integer greater than 2 as an example, the data of each category in the N categories is used for AI model training. The N categories can be classified according to one or more of the following: N preset data features, N preset beam sets, N preset cell sets, N preset resource configurations, N preset speeds, N preset time ranges, N preset latitude and longitude ranges, N preset beam numbers, N preset cell coverage ranges, or N preset identifiers of beams used for data transmission.

[0166] For example, in the case that the N categories are classified according to N preset data features, the data of the first category can be data corresponding to a data feature with insufficient data in the N data features.

[0167] In the case that the N categories are classified according to N preset beam sets, the data of the first category can be data corresponding to a beam set with insufficient data in the N beam sets.

[0168] In the case that the N categories are classified according to N preset beam numbers, the data of the first category can be data corresponding to a beam number with insufficient data in the N beam numbers.

[0169] In a case that the N categories are preset N cell identifications, the data of the first category can be data corresponding to a cell identification with insufficient data amount in the N cell identifications. One cell identification can be used to indicate a cell coverage. The cell identification can also be referred to as an identification of a cell.

[0170] For ease of understanding, the preset N categories of data features, the preset N categories of beam sets, the preset N categories of cell sets, the preset N categories of resource configurations, the preset N categories of speeds, the preset N categories of time ranges, the preset N categories of latitude and longitude ranges, the preset N beam numbers, the preset N cell coverage ranges, or the preset N identifications of beams for data transmission are described below.

[0171] In S102, the second device can send the first category of data to the first device. Correspondingly, the first device can receive the first category of data reported by the second device.

[0172] Exemplarily, in a case that the first device is independent of the third device, the second device can send the first category of data to the third device. Correspondingly, the third device can receive the first category of data reported by the second device.

[0173] The third device can send the first category of data to the first device. Correspondingly, the first device can receive the first category of data sent by the third device.

[0174] In this way, by forwarding the first category of data through the third device, the first device can receive the first category of data reported by the second device.

[0175] It can be understood that, in a case that the first device has the function of a network device, the first category of data does not need to be forwarded through the third device. That is, the second device can send the first category of data to the first device without the third device. Correspondingly, the first device can receive the first category of data reported by the second device without the third device.

[0176] In this way, the first device can collect the first category of data in a targeted manner.

[0177] The data transmission method provided by the embodiments of the present application can make the data quantity of each category in the training set for AI model training sufficient, i.e., the balance of the data features in the training set is good. In this way, the AI model trained by using the training set can have good generalization. For example, in the case where the AI model is used for beam management, the probability of the poor signal quality of the optimal beam predicted by the second device by using the AI model can be reduced, i.e., the probability of the poor signal quality of the optimal beam fed back by the second device to the third device affecting the data transmission between the second device and the third device can be reduced.

[0178] The S101 is described in detail below. Illustratively, before the first device sends the first information to the second device, the first device can collect the data set.

[0179] The data set can include data of multiple categories.

[0180] Illustratively, the data set can include data of the first category to data of the Nth category. N is an integer greater than 2.

[0181] Illustratively, the categories in the data set can be obtained based on one or more of the following classifications: preset N categories of data features, preset N categories of beam sets, preset N categories of cell sets, preset N categories of resource configurations, preset N categories of speeds, preset N categories of time ranges, or preset N categories of latitude and longitude ranges.

[0182] Among them, the N categories of data features can be: N categories of communication conditions (condition), N categories of communication situations (situation), N categories of communication environments (environment or circumstance), N categories of communication types (type), N categories of communication states (status), N categories of communication data (data), or N categories of communication data distribution (data distribution). The data feature can be referred to as a feature or a data classification feature. It can be understood that other feature-related terms in the embodiments of the present application, such as "feature" in xx feature, reference feature, physical feature, or feature of the xth data, can be replaced similarly with condition, situation, environment, type, status, data, or data distribution.

[0183] The preset N types of data features can be associated with preset N data feature identifiers. One data feature identifier can be used to indicate one type of data feature. The data feature identifier can be referred to as an associated ID (associated ID), a data ID (data ID), a dataset ID (dataset ID), a data categorization ID (data categorization ID), or a property ID (property ID). The data feature identifier can also be used by the terminal device to classify the obtained data or measurement quantity. The measurement quantity is used to determine the model input, or to determine the training data (for example, the model output or label) used for model training. The measurement quantity can also be used for operations related to the model inference stage. For example, the terminal device can determine that the corresponding beams in the downlink beam set or beam list have the same or similar data features in the case of receiving the downlink beam set or beam list with the same data feature identifier.

[0184] The beam set can include information of the first type of beam and information of the second type of beam. The information of the first type of beam can include a number of the first type of beam and an identifier of each of the plurality of first type of beam. The information of the second type of beam can include a number of the second type of beam and an identifier of each of the plurality of second type of beam.

[0185] The preset N types of beam sets can be preset N beam set identifiers. One beam set identifier can be used to indicate one type of beam set.

[0186] The pattern of the set in the embodiments of the present application can be a list or a sequence.

[0187] The cell set can include at least one cell. The cell set can include an identifier of each of the at least one cell. The preset N types of cell sets can be preset N cell set identifiers. One cell set identifier can be used to indicate one type of cell set. It should be understood that in the case where the cell set includes one cell, the cell set identifier can be the identifier of the cell.

[0188] The resource configuration can include a radio resource control (RRC) configuration, a medium access control-control element (MAC-CE) configuration, a downlink control information (DCI) configuration, and / or a beam configuration, etc.

[0189] The beam configuration can comprise a configuration of first type beams in a beam set A and a configuration of second type beams in a beam set B, both of which are transmitted by the network device (or a third device). Both the beam set A and the beam set B can be used for beam management.

[0190] Exemplarily, the configuration of the first type beams in the beam set A can comprise a number of the first type beams in the beam set A, a transmission order of the first type beams, and / or an orientation of the first type beams, etc.

[0191] The configuration of the second type beams in the beam set B can comprise a number of the second type beams in the beam set B, a transmission order of the second type beams, and / or an orientation of the second type beams, etc.

[0192] It should be understood that the beam configuration can comprise a spatial relationship between the beam set A and the beam set B. The spatial relationship between the beam set A and the beam set B can refer to Figure 4 or Figure 5 .

[0193] The beam set A can be referred to as a beam set A, and the beam set B can be referred to as a beam set B. The number of the first type beams in the beam set A is greater than the number of the second type beams in the beam set B. The first type beams in the beam set A can be referred to as dense beams. The second type beams in the beam set B can be referred to as sparse beams. The beam set A can comprise the second type beams in the beam set B, which can refer to Figure 4 . The beam set A can also not comprise the second type beams in the beam set B, which can refer to Figure 5 . For ease of understanding, Figure 4 and Figure 5 will be described later.

[0194] The preset N types of resource configurations can be preset N resource configuration identifiers. One resource configuration identifier can be used to indicate one type of resource configuration.

[0195] The speed in the N types of speeds can be understood as the motion speed of the terminal device. The preset N types of speeds can be preset N speed indication information.

[0196] The preset N types of time ranges can be preset N time range identifiers. One time range identifier can be used to indicate one type of time range.

[0197] The latitude and longitude range can be used to indicate the motion trajectory of the terminal device. The preset N types of latitude and longitude ranges can be preset N latitude and longitude range identifiers. One latitude and longitude range identifier can be used to indicate one type of latitude and longitude range.

[0198] Exemplarily, the data in the data set can be channel information, and / or beam information.

[0199] The data in the data set is channel information, and the training set of the data set can be used for training of an AI model implementing a channel state information feedback enhancement function.

[0200] The data in the data set is beam information, and the training set of the data set can be used for training of an AI model implementing a beam management function.

[0201] The channel information can include one or more of the following: beam signal quality, channel response information, and / or channel quality information.

[0202] The beam signal quality includes one or more of the following: reference signal received power (RSRP), signal to interference plus noise ratio (SINR), and reference signal received quality (RSRQ).

[0203] The RSRP can include layer 1 reference signal received power (L1-RSRP) and / or layer 3 reference signal received power (L3-RSRP).

[0204] The channel response information can include at least one of the following: precoding matrix indicator (PMI), precoding matrix, precoding vector, eigenvector, eigenmatrix, channel vector, channel matrix, layer information (LI), and rank information (RI).

[0205] The channel quality information can include at least one of the following: RSRP, SINR, and channel quality indicator (CQI).

[0206] Exemplarily, the channel information can be raw channel information. The raw channel information can be understood as channel state information (CSI) that has not been processed or initially measured, including channel quality indicator (CQI), channel gain, and / or noise level, etc.

[0207] The beam information can be a channel state information reference signal resource indicator (CRI) and / or a synchronization signal block resource indicator (SSBRI). The beam information can also be referred to as reference signal identification information.

[0208] Optionally, the categories in the data set can be classified based on any one of the following: preset N beam numbers, preset N cell coverage ranges, or preset N identifications of beams for data transmission.

[0209] Exemplarily, the data of the first category can be data with insufficient data quantity in the data set. For example, the data of the first category satisfies at least one of the following, or the screening condition of the first category includes at least one of the following:

[0210] The data quantity of the first category in the data set is less than a first threshold value;

[0211] The proportion of the data quantity of the first category in the data set is less than a second threshold value;

[0212] The ratio of the data quantity of the first category to the data quantity of a target category is less than a third threshold value, the target category is different from the first category, and the data of the target category is included in the data set;

[0213] The absolute value of the difference between the proportion of the first category and the proportion of the target category is greater than a fourth threshold value, the proportion of the first category is the proportion of the data quantity of the first category in the data set, and the proportion of the target category is the proportion of the data quantity of the target category in the data set.

[0214] The target category can be a category with the largest data quantity in the data set.

[0215] It can be understood that the first threshold value, the second threshold value, the third threshold value, and the fourth threshold value can be preset or predefined by the first device.

[0216] Exemplarily, the first information can include an identification of the first category, or the first information can be used to indicate the identification of the first category.

[0217] In a case where the categories in the data set are classified based on preset N data feature identifications, the identification of the first category can be one of the N data feature identifications.

[0218] In a case where the categories in the data set are classified based on preset N beam set identifications, the identification of the first category can be one of the N beam set identifications.

[0219] In a case where the categories in the data set are classified based on preset N cell set identifiers, the identifier of the first category can be one of the N cell set identifiers.

[0220] In a case where the categories in the data set are classified based on preset N resource configuration identifiers, the identifier of the first category can be one of the N resource configuration identifiers.

[0221] In a case where the categories in the data set are classified based on preset N speed indication information, the identifier of the first category can be one of the N speed indication information.

[0222] In a case where the categories in the data set are classified based on preset N time range identifiers, the identifier of the first category can be one of the N time range identifiers.

[0223] In a case where the categories in the data set are classified based on preset N beam numbers, the identifier of the first category can be one of the N beam numbers.

[0224] In a case where the categories in the data set are classified based on preset N cell coverage ranges, the identifier of the first category can represent one of the N cell coverage ranges. The identifier of the first category can be an identifier of a cell.

[0225] In a case where the categories in the data set are classified based on preset N latitude and longitude range identifiers, the identifier of the first category can be one of the N latitude and longitude range identifiers.

[0226] In a case where the categories in the data set are classified based on preset N beam identifiers for data transmission, the identifier of the first category can be one of the N beam identifiers for data transmission.

[0227] In a case where the categories in the data set are classified based on preset N beam configuration indication information, the identifier of the first category can represent one of the N beam configuration indication information.

[0228] The first device can select the first category from the data set collected by the first device according to the filtering condition of the first category. In a case where the first device selects the first category, the first device can perform S101 to instruct to collect data of the first category with insufficient data amount in the data set.

[0229] It can be understood that the training set for AI model training can include the data set collected by the first device before performing S101, and the data of the first category collected by using the data transmission method (such as S101-S102) provided in the embodiments of the present application.

[0230] The data transmission method provided in this application embodiment can achieve better balance in the data features of the training set. Therefore, using this training set for AI model training can result in better generalization performance of the AI ​​model.

[0231] Taking the example of the first device operating independently of the third device, in one possible implementation, to increase the amount of data in the first category that is insufficient in the dataset, the first device can send second information to the second device through the third device. This second information instructs the reporting of data for AI model training. Correspondingly, the second device can receive the second information from the first device through the third device.

[0232] The second device can report data from at least one category (from the first to the Nth category) to the first device via the third device. Correspondingly, the first device can receive data from the second device reported by the second device from at least one category (from the first to the Nth category).

[0233] Since the dataset contains sufficient data for all categories except the first category (e.g., any category from the second to the Nth category), the data reported by the second device for other categories upon receiving the second information is redundant. This results in wasted power consumption, storage, and reporting overhead for the second device in acquiring redundant data.

[0234] In the data transmission method provided in this application embodiment, in order to increase the amount of data in the first category that is insufficient in the dataset, the first device specifically collects data in the first category that is insufficient in the collected dataset. Upon receiving the first information, the second device reports the data in the first category and does not report data in other categories. This reduces the probability of wasted power consumption, storage, and reporting overhead in acquiring redundant data.

[0235] Figure 4 This diagram illustrates a spatial relationship between beam set A and beam set B provided in an embodiment of this application.

[0236] like Figure 4 As shown, Figure 4 Any number from 1 to 16 can be used as a beam number. Beam set A (or set A) can include 16 beams numbered 1-16. Beam set B (or set B) can include the beams corresponding to 1, 4, 7, and 10, i.e., beam set B can include 4 beams. All beams in beam set A are type I beams. All beams in beam set B are type II beams. Beam set A contains beams from beam set B.

[0237] Figure 5Another schematic diagram of the spatial relationship of the beam set A and the beam set B provided by the embodiments of the present application is shown.

[0238] As shown in Figure 5 Figure 5 A1-A8 and B1-B3 can be beam numbers. The beam set A (or set A) can include 8 beams with beam numbers A1-A8. The beam set B (or set B) can include 3 beams with beam numbers B1-B3. The beam set A does not contain the beams in the beam set B.

[0239] It can be understood that Figure 4 or Figure 5 The spatial relationship of the beam set A and the beam set B, the number of beams in the beam set A and the orientation of the beams, and the number of beams in the beam set B and the orientation of the beams shown are only an example and are not a limitation on the spatial relationship of the beam set A and the beam set B, the number of beams in the beam set A and the orientation of the beams, and the number of beams in the beam set B and the orientation of the beams. For example, the number of beams in the beam set A can be other values than 16 and 8. The number of beams in the beam set B can be other values than 4 and 3.

[0240] Exemplarily, the second device adopts the AI model for beam management, for example, the second device adopts the AI model for optimal beam prediction, the third device can transmit the beams in set A and the beams in set B to the second device. The second device can obtain the signal quality of the beams in set B, for example, the second device can obtain the signal quality of each beam in set B by measurement. The second device inputs the signal quality of each beam in set B into the AI model. The second device can obtain the optimal beam with the optimal signal quality output by the AI model. The optimal beam is one or more beams in set A. The second device can adopt the optimal beam for data transmission. In this way, the second device does not need to measure the signal quality of each beam in set A transmitted by the third device, and can obtain the optimal beam in set A, thereby reducing the power consumption of the second device.

[0241] The following still takes the AI node or the cloud server as the first device, the terminal device as the second device, and the network device as the third device, and the first device is independent of the third device as an example, and S101-S102 are described in combination with some embodiments.

[0242] Figure 6 A schematic diagram of a data collection scenario provided by an embodiment of the present application is shown.

[0243] In an embodiment of the present application, as shown in Figure 6 ​As shown, the first device 501 can train a plurality of AI models. For example, the first device 501 can train AI model 1 used by the second device 502. The first device 501 can train AI model 2 used by the second device 504. The second device 502 is a device within the service area of the third device 503. The second device 504 is a device within the service area of the third device 505.

[0244] The first device 501 adopts the data transmission method provided by the embodiments of the present application to collect data of the first category with insufficient data amount in the training set 1 in a targeted manner, and can also collect data of the first category with insufficient data amount in the training set 2 in a targeted manner.

[0245] The training set 1 can be used for training of AI model 1. The training set 2 can be used for training of AI model 2. The first device 501 collects data of the first category with insufficient data amount in the training set 1 in a targeted manner, and the first device 501 collects data of the first category with insufficient data amount in the training set 2 in a targeted manner. This can be performed simultaneously or not simultaneously. For example, it is performed in different time periods respectively.

[0246] For the first device 501 to collect data of the first category with insufficient data amount in the training set 1 in a targeted manner, as shown in Figure 6 The first device 501 can transmit the identifier of the first category to the third device 503. Correspondingly, the third device 503 can receive the identifier of the first category from the first device 501. In this way, the first device 501 can indicate to the third device 503 to collect data of the first category.

[0247] The third device 503 can broadcast the identifier of the first category to at least one second device 502 served by the third device 503. Correspondingly, the second device 502 can receive the identifier of the first category broadcast by the third device 503. In this way, the third device 503 can indicate to the second device 502 to report data of the first category. Upon receiving the identifier of the first category, the second device 502 can report data of the first category to the third device 503. The third device 503 can transmit the data of the first category reported by the second device 502 to the first device 501. In this way, the first device can collect data of the first category in the training set 1 in a targeted manner.

[0248] For the first device 501 to collect data of the first category with insufficient data amount in the training set 2 in a targeted manner, taking the identifier of the first category as the identifier of cell 1 and the data in the training set 2 as the signal quality of the beam as an example, as shown in Figure 6As shown, the first device 501 can transmit the identity of the cell 1 to the third device 505. Correspondingly, the third device 505 can receive the identity of the cell 1 from the first device 501. In this way, the first device 501 can indicate to the third device 505 to instruct the second device 504 accessing the cell 1 to report the signal quality of the obtained beam.

[0249] The third device 505 can broadcast the identity of the first category to at least one second device 504 served by the third device 505. For example, the third device 505 can broadcast the identity of the cell 1 in the cell associated with the third device 505, or the third device 505 can broadcast the identity of the cell 1 in the cell 1. The cell associated with the third device 505 is, for example, the cell 1, the cell 2 and the cell 3.

[0250] Correspondingly, the second device 504 can receive the identity of the cell 1 broadcasted by the third device 505. In this way, the third device 505 can indicate to the second device 504 to report the signal quality of the beam in the case of accessing the cell 1. In the case of accessing the cell 1, the second device 504 can report the signal quality of the obtained beam to the third device 505. The third device 505 can transmit the signal quality of the beam reported by the second device 504 to the first device 501. In this way, the first device can collect the data of the first category in the training set 2. It should be understood that the signal quality of the beam reported by the second device 504 belongs to the data of the first category in the training set 2 in the case of accessing the cell 1 by the second device 504.

[0251] It can be understood that the number of third devices is not limited in the embodiments of the present application. Taking the example of the first device collecting the data of the category with insufficient data amount in the training set for training of one AI model, the first device can instruct multiple third devices to report the data of the category with insufficient data amount in the training set for training of the one AI model at the same time, so as to shorten the collection time of the data of the category with insufficient data amount.

[0252] Taking the example of the first device collecting the data of the category with insufficient data amount in the training set for training of one AI model independently of the third device, Figure 7 Another interactive flowchart of the data transmission method provided by the embodiments of the present application is shown.

[0253] As shown, Figure 7 The data transmission method provided by the embodiments of the present application can include S701-S709:

[0254] S701, the first device can send second information to the third device. Correspondingly, the third device can receive the second information from the first device.

[0255] The second information can be used to indicate data for model training (e.g., AI model training).

[0256] For example, the second information can be used to indicate L1-RSRP. For example, the second information can include an identifier of L1-RSRP. In this way, the first device can indicate the third device to collect and report L1-RSRP by sending the second information to the third device.

[0257] For example, the second information can be used to indicate channel information. For example, the second information can include an identifier representing channel information. In this way, the first device can indicate the third device to collect and report channel information by sending the second information to the third device.

[0258] At S702, the third device can send the second information to the second device. Correspondingly, the second device can receive the second information sent by the third device. In this way, the second device can receive the second information from the first device.

[0259] For example, the third device can broadcast the second information. Correspondingly, at least one second device served by the third device can receive the second information broadcasted by the third device. The at least one second device served by the third device can include, for example, the second device A and the second device B as shown in FIG. 7B. In this way, the second device can report data indicated by the second information. Figure 7

[0260] For example, when the third device broadcasts the second information, the third device can also broadcast configuration related to the second information. Correspondingly, the second device can receive the second information and the configuration related to the second information.

[0261] For example, the configuration related to the second information can include configuration information of set A and set B. The configuration information of set A and set B can be, for example, configuration of the first type of beam in set A and configuration of the second type of beam in set B.

[0262] In this way, the third device can send each beam in set A and each beam in set B to the second device according to the configuration related to the second information, so that the second device can identify the beam number of each beam sent by the third device while obtaining the signal quality of each beam sent by the third device.

[0263] ​It should be understood that the third device can broadcast the second information and the configuration related to the second information simultaneously. The third device can also broadcast the second information and the configuration related to the second information sequentially. For example, the third device can broadcast the second information first and then broadcast the configuration related to the second information.

[0264] Exemplarily, the second information can further be used to indicate the first data amount. For example, the second information can further include a first data amount threshold. The first data amount threshold can represent the data amount that the first device needs to collect to perform S701, or can represent the data amount that the first device indicates to collect through the second information.

[0265] Optionally, the first data amount threshold can not be sent by the first device through the second information. The first data amount threshold can be pre-set on the third device, or the first data amount threshold can be pre-agreed by the first device and the third device.

[0266] Exemplarily, the third device can send each beam in set A and each beam in set B to the second device according to the configuration related to the second information multiple times to achieve that the data amount of the collected data for model training meets the first data amount threshold.

[0267] For example, the third device can pre-store a correspondence (or mapping relationship) between a data amount threshold and a beam transmission number. The beam transmission number is the number of times of transmitting each beam in set A and each beam in set B. One beam transmission can be understood as transmitting each beam in one set A and each beam in one set B according to the configuration related to the second information.

[0268] It should be understood that the correspondence between the data amount threshold and the beam transmission number can include multiple data amount thresholds and the beam transmission number corresponding to each data amount threshold in the multiple data amount thresholds.

[0269] The third device can select a first beam transmission number corresponding to the first data amount threshold from the correspondence between the data amount threshold and the beam transmission number.

[0270] The first beam transmission number can be referred to as the first transmission number.

[0271] The third device can send each beam in set A and each beam in set B to the second device according to the first transmission number and the configuration related to the second information. For example, taking the first transmission number m times as an example, the third device can send each beam in set A and each beam in set B to the second device m times according to the configuration related to the second information.

[0272] Wherein, m is an integer. The time interval between adjacent two times of m can be the same or different, and the embodiments of the present application do not make specific limitation.

[0273] Optionally, taking the channel information as an example for model training, the third device can pre-store a corresponding relationship between the data amount threshold and the broadcast number.

[0274] It should be understood that the corresponding relationship between the data amount threshold and the broadcast number can include multiple data amount thresholds and the broadcast number corresponding to each data amount threshold in the multiple data amount thresholds.

[0275] The third device can select the first broadcast number corresponding to the first data amount threshold from the corresponding relationship between the data amount threshold and the broadcast number.

[0276] The third device can periodically broadcast the second information according to the first broadcast number. So that the second device reports the channel information to the third device when receiving the second information, so as to realize that the data amount of the collected data for model training can reach the first data amount threshold.

[0277] S703, the second device can send the data for model training to the third device. Correspondingly, the third device can receive the data for model training reported by the second device.

[0278] Exemplarily, taking the L1-RSRP as an example for model training, the second device can send the L1-RSRP of the beam to the third device. Correspondingly, the third device can receive the L1-RSRP of the beam reported by the second device. It should be understood that the L1-RSRP of the beam sent by the second device is the L1-RSRP of the beam received by the second device.

[0279] Exemplarily, the second device can send the L1-RSRP of one beam to the third device every time the second device obtains the L1-RSRP of one beam.

[0280] Optionally, the second device can send the L1-RSRP of each beam in set A and the L1-RSRP of each beam in set B to the third device in the case of obtaining the L1-RSRP of each beam in set A and the L1-RSRP of each beam in set B. For example, the second device can send the L1-RSRP of each beam in set A and the L1-RSRP of each beam in set B in a package.

[0281] Exemplarily, in the case that the at least one second device served by the third device includes the second device A and the second device B, the second device A can send the L1-RSRP of the beam obtained by the second device A to the third device. The second device B can send the L1-RSRP of the beam obtained by the second device B to the third device.

[0282] The L1-RSRP of the second device A transmitting beam and the L1-RSRP of the second device B transmitting beam can be performed simultaneously or not simultaneously.

[0283] Optionally, taking channel information as an example (used for model training), the second device can send channel information to the third device. Correspondingly, the third device can receive the channel information reported by the second device.

[0284] For example, each time the second device receives the second information, the second device can send channel information to the third device once.

[0285] S704, the third device can send data for model training to the first device. Correspondingly, the first device can receive the data for model training sent by the third device. In this way, the first device can receive the data for model training reported by the second device.

[0286] For example, the third device may send data for model training to the first device at a preset frequency until the amount of data sent by the third device to the first device for model training is greater than or equal to a first data amount threshold.

[0287] Optionally, the third device may send the received data for model training to the first device when the amount of data received for model training is greater than or equal to a first data volume threshold.

[0288] In this way, the first device collects a dataset for model training. It should be understood that the dataset includes data for model training sent by the third device as shown in S704, or data for model training reported by the second device as shown in S703.

[0289] S705. The first device can determine the first category based on the dataset.

[0290] For example, the first device can classify the data in the dataset to obtain multiple categories and data for each category.

[0291] The first device can select the first category from multiple categories corresponding to the dataset according to the filtering criteria of the first category. The first device can execute S706.

[0292] In this way, categories with insufficient data can be selected from the dataset collected by the first device, so that targeted data collection can be carried out for these categories.

[0293] Exemplarily, the first device can classify the data in the data set based on one or more of preset N types of data features, preset N sets of beams, preset N sets of cells, preset N types of resource configurations, preset N types of speeds, preset N types of time ranges, preset N beam numbers, preset N cell coverage ranges, preset N latitude and longitude ranges, preset N identifiers of beams for data transmission, or preset N types of beam configurations, to obtain N categories and data in each of the N categories.

[0294] The first device can select the first category from the N categories corresponding to the data set according to the filtering condition of the first category.

[0295] Exemplarily, the first device can use a classification model (such as a classifier) to classify the data in the data set, to obtain the N categories and data in each of the N categories.

[0296] For example, the classification model can be based on one or more of preset N data feature identifiers, preset N set identifiers of beams, preset N set identifiers of cells, preset N resource configuration identifiers, preset N speed indication information, preset N time range identifiers, preset N latitude and longitude range identifiers, preset N beam numbers, preset N cell coverage ranges, preset N identifiers of beams for data transmission, or preset N beam configuration indication information, to classify the data in the data set.

[0297] It can be understood that the N categories can include a first category to an Nth category. The data set can include data of the first category to data of the Nth category.

[0298] Optionally, if the first device does not select the first category from the N categories corresponding to the data set, it can be indicated that the data amount of each of the N categories is sufficient, and there is no category with insufficient data amount in the data set. Then, the first device can use the data set to train a model (such as an AI model), without performing S706. In this way, the generalization of the AI model trained by the first device can be better while reducing the interaction process.

[0299] S706, the first device can send first information to the third device. Correspondingly, the third device can receive the first information from the first device.

[0300] The first information can include an identifier of the first category.

[0301] S707, the third device can send the first information to the second device. Correspondingly, the second device can receive the first information sent by the third device. In this way, the second device receives the first information from the first device.

[0302] Exemplarily, the third device can broadcast the first information. Correspondingly, each of the at least one second device served by the third device can receive the first information. For example, the second device A and the second device B can both receive the first information.

[0303] Exemplarily, the first information can also be used to indicate the second data amount. For example, the first information can also include a second data amount threshold. The second data amount threshold can represent the data amount required by the first device to perform S706 for targeted data collection, or can represent the data amount indicated by the first device to collect through the first information. In this way, the first category of data can be collected through targeted collection, so that the amount of data of the first category in the training set for model training is sufficient.

[0304] Exemplarily, the second data amount threshold can be the absolute value of the difference between the first threshold and the data amount of the first category.

[0305] The second data amount threshold can also be the absolute value of the difference between the data amount of the first category and the data amount of the target category.

[0306] The second data amount threshold can also be a value satisfying at least one of the following:

[0307] The sum of the data amount of the first category and the second data amount threshold in the data set is greater than or equal to the first threshold;

[0308] The proportion of the sum of the data amount of the first category and the second data amount threshold in the data set is greater than or equal to the second threshold;

[0309] The ratio of the sum of the data amount of the first category and the second data amount threshold to the data amount of the target category is greater than or equal to the third threshold;

[0310] The absolute value of the difference between the proportion of the sum of the data amount of the first category and the second data amount threshold and the proportion of the target category is less than or equal to the fourth threshold. The proportion of the sum of the data amount of the first category and the second data amount threshold is the proportion of the sum of the data amount of the first category and the second data amount threshold in the data set.

[0311] Exemplarily, taking the L1-RSRP of the first category of data as a beam as an example, the third device can select a second beam transmission number corresponding to the second data amount threshold from a pre-stored corresponding relationship between the data amount threshold and the beam transmission number.

[0312] The second beam transmission number can be referred to as the second transmission number.

[0313] After the third device broadcasts the first information, the third device can send each beam in set A and each beam in set B to the second device according to the second sending number and the configuration related to the second information. The specific implementation principles and technical effects can be referred to the specific implementation principles and technical effects of the third device sending each beam in set A and each beam in set B to the second device according to the first sending number and the configuration related to the second information.

[0314] Optionally, taking the channel information of the first type of data as an example, the third device can pre-store a corresponding relationship between the data amount threshold and the broadcast number.

[0315] The third device can select the second broadcast number corresponding to the second data amount threshold from the pre-stored corresponding relationship between the data amount threshold and the broadcast number.

[0316] The third device can periodically broadcast the first information according to the second broadcast number. In order to facilitate the second device to report the channel information associated with the first information to the third device when receiving the first information, so as to achieve that the amount of the first type of data collected in a targeted manner can reach the second data amount threshold.

[0317] S708, the second device can send the first type of data to the third device. Correspondingly, the third device can receive the first type of data reported by the second device.

[0318] For example, taking the L1-RSRP of the beam with beam number A5 as the first type of data, and taking the beam number A5 as the first type of identifier.

[0319] The second device can send the L1-RSRP of the beam with beam number A5 to the third device. Correspondingly, the third device can receive the L1-RSRP of the beam with beam number A5 reported by the second device.

[0320] It should be understood that the second device can send the L1-RSRP of the beam with beam number A5 to the third device every time the second device obtains the L1-RSRP of the beam with beam number A5.

[0321] In this way, the second device reports the L1-RSRP of the beam with beam number A5 and does not report the L1-RSRP of the beam with beam number other than A5, which can reduce the reporting overhead of redundant data when the second device obtains the L1-RSRP of each beam in set A and each beam in set B.

[0322] Optionally, in the case that the second device obtains the first information and obtains the beam sent by the third device, if the second device identifies that the beam number of the beam sent by the third device is not A5, the second device can not obtain the L1-RSRP of the beam with the beam number not being A5, and the power consumption and storage overhead of redundant data can be reduced.

[0323] Exemplarily, the first category of data is the L1-RSRP of the beam in the coverage range of the cell corresponding to the cell identifier C1, and the first category of identifier is the cell identifier C1.

[0324] The second device can send the L1-RSRP of the beam to the third device when accessing the cell with the cell identifier C1 and / or when being in the coverage range of the cell with the cell identifier C1. Correspondingly, the third device can receive the L1-RSRP of the beam reported by the second device.

[0325] It should be understood that the cell corresponding to the cell identifier C1 is one cell served by the third device, and the cell served by the third device can also include other cells. The identifier of the other cells is not C1.

[0326] In this way, the occurrence probability of the second device reporting the L1-RSRP of the beam in the other cells of the third device (such as the cells with the cell identifier not being C1) can be reduced, and the power consumption, the storage overhead of redundant data, and the reporting overhead of redundant data of the second device in the other cells of the third device can be reduced.

[0327] Exemplarily, the first category of data is the channel information in the coverage range of the cell corresponding to the cell identifier C1, and the first category of identifier is the cell identifier C1.

[0328] The second device can send the channel information to the third device when accessing the cell with the cell identifier C1 and / or when being in the coverage range of the cell with the cell identifier C1. Correspondingly, the third device can receive the channel information reported by the second device.

[0329] It should be understood that the second device can send the channel information to the third device once for each time of receiving the first information.

[0330] In this way, the occurrence probability of the second device reporting the channel information in the other cells of the third device (such as the cells with the cell identifier not being C1) can be reduced, and the power consumption, the storage overhead of redundant data, and the reporting overhead of redundant data of the second device in the other cells of the third device can be reduced.

[0331] Exemplarily, in a case that the at least one second device served by the third device includes the second device A and the second device B, the second device A can send the first category of data to the third device. The second device B can send the first category of data to the third device.

[0332] It should be understood that the first category of data reported by the second device can contain the identification of the first category.

[0333] S709, the third device can send the first category of data to the first device. Correspondingly, the first device can receive the first category of data sent by the third device. In this way, the first device can receive the first category of data reported by the second device.

[0334] Exemplarily, the third device can send the first category of data to the first device at a preset frequency until the amount of data of the first category of data sent by the third device to the first device is greater than or equal to the second data amount threshold.

[0335] Optionally, the third device can send the first category of data to the first device when the amount of data of the first category of data is greater than or equal to the second data amount threshold.

[0336] In this way, the first device can collect the first category of data in a targeted manner, so that the balance of the amount of data of each category in the training set used for model training is better, and thus the generalization of the model trained by using the training set can be better.

[0337] The data collection shown in S701-S704 can be referred to as indiscriminate data collection.

[0338] As Figure 7 shown, the data transmission method provided by the embodiments of the present application can obtain a data set with an amount of data reaching a first data amount threshold through the indiscriminate data collection as shown in S701-S704. By classifying the data set, a plurality of categories of data can be obtained. By selecting a first category with insufficient data amount from the plurality of categories, and by collecting the first category of data in a targeted manner through the manner as shown in S706-S709, the first category of data with an amount of data reaching a second data amount threshold can be collected. The balance of the amount of data of each category in the training set used for model training (such as AI model training) can be better. The training set contains the data set and the first category of data collected in a targeted manner as shown in S706-S709, and the amount of data of the training set reaches the sum of the first data amount threshold and the second data amount threshold. By implementing AI model training by using the training set, the generalization of the AI model can be better. The targeted data collection of the first category can be understood as the targeted collection of the data of the first category. In addition, the power consumption for obtaining redundant data, and the storage and reporting overheads can also be reduced.

[0339] Still taking the example that the first device collects data of a category with insufficient data amount in a training set for AI model training independently of the third device, Figure 8 Another interactive flowchart of the data transmission method provided by the embodiments of the present application is shown.

[0340] As Figure 8 shown, the data transmission method provided by the embodiments of the present application can include S807-S811:

[0341] S801, the first device can send third information to the third device. Correspondingly, the third device can receive the third information from the first device.

[0342] Since the communication environment can be dynamically changing. The change of the communication environment can cause the communication system to generate new data different from the data in the data set. In the embodiments of the present application, taking the new data as the data of the (N+1)th category as an example, the third information can be used to indicate the reporting of the data of the (N+1)th category and the data of at least one category of the first category to the Nth category of the data set. The (N+1)th category is different from the first category to the Nth category. The first category to the (N+1)th category can be a preset category. Among them, the first category to the Nth category can be a category of data that the known communication system can generate. The (N+1)th category can be a category of data that the research and development personnel analyze and speculate based on the possible changes of the communication environment. The (N+1)th category can also be a category of data that the known communication system can generate, but the data amount of the data of this category is less than a preset threshold (such as the eleventh threshold) compared to the data of the first category to the Nth category, or the probability of the data of this category appearing is less than a preset probability threshold compared to the data of the first category to the Nth category, but with the change of the communication environment, the data of this category can exceed the twelfth threshold, or the probability of the data of this category appearing can exceed the thirteenth threshold. Among them, the twelfth threshold can be greater than or equal to the eleventh threshold. The thirteenth threshold can be greater than or equal to the preset probability threshold.

[0343] By sending the third information from the first device to the third device to indicate the collection of the data of the (N+1)th category and the data of at least one category of the first category to the Nth category of the data set. On the one hand, the collection of the data of each category of the first category to the Nth category for model training can be realized, and on the other hand, it can be determined whether there is new data. If it is determined that there is new data, it means that the change of the communication environment causes the communication system to generate new data, so as to facilitate the targeted collection of the new data, so as to perform model training (such as AI model training) by using a training set containing sufficient data of known categories (such as the first category to the Nth category) and new data, so as to improve the generalization of the AI model, so that the AI model can perform better when applied in the changed communication environment.

[0344] For example, the data feature of the N+1th category can be determined based on the ith feature. The ith feature is included in the feature of the jth category data in the N categories of the data set. 1≤i and i is an integer, 1≤j≤N and j is an integer.

[0345] For example, in the case where the jth category includes the signal quality of each beam in multiple beams, the ith feature can be the maximum signal quality in the jth category.

[0346] The N+1th category can include the signal quality of the beam, and the absolute value of the difference between the signal quality of the N+1th category and the maximum signal quality is greater than the tenth threshold value.

[0347] For example, the first value determined based on the value associated with the kth feature of the data of the N+1th category and the value associated with the ith feature of the data of the jth category is less than the fifth threshold value, the first value is used to indicate the correlation between the value associated with the kth feature and the value associated with the ith feature, 1≤k and k is an integer.

[0348] For example, the first value can be the difference between the value associated with the kth feature and the value associated with the ith feature. The first value can also be the ratio of the value associated with the kth feature and the value associated with the ith feature. The first value can also be a value obtained by processing the value associated with the kth feature and the value associated with the ith feature using an existing correlation statistical method.

[0349] For example, in the case where the jth category includes the signal quality of each beam in 3 beams, the 3 beams in the jth category are beam a, beam b and beam c respectively, and the signal quality of beam a>the signal quality of beam b>the signal quality of beam c, the value associated with the ith feature can be the signal quality of beam a. For example, the value associated with the kth feature of the data of the N+1th category is the same as the signal quality of beam b, and the first value determined based on the signal quality of beam a and the signal quality of beam b is less than the fifth threshold value. That is, in the case where the data of the jth category is obtained, if the first value determined based on the signal quality of beam a and the signal quality of beam b is less than the fifth threshold value, it can be determined that the signal quality of beam b can be the data of the N+1th category.

[0350] For example, the N+1th category can include the signal quality of each beam in multiple beams. The absolute value of the difference between the two largest signal qualities in the N+1th category is greater than the sixth threshold value. The difference between the two largest signal qualities in the N+1th category can be understood as follows: the difference between the signal qualities of the top 2 beams in the sequence of signal qualities obtained by arranging the multiple beams included in the N+1th category in descending order of signal quality.

[0351] Alternatively, the N+1th category can include a signal to interference plus noise ratio (SINR), and an absolute value of a difference between the SINR of the N+1th category and a second value of the first preset range can be greater than a seventh threshold value. The second value can be a minimum value or a maximum value of the first preset range.

[0352] The data of the first category to the data of the Nth category can be within the first preset range.

[0353] The data feature of the N+1th category can be previously agreed upon by the first device, the second device, and the third device.

[0354] Exemplarily, the information related to the data feature of the N+1th category can be previously agreed upon by the first device, the second device, and the third device, can be preset on the first device, the second device, and the third device, or can be carried in the third information and sent by the first device to the third device and / or the second device. The information related to the data feature of the N+1th category, for example, the fifth threshold value, the sixth threshold value, the seventh threshold value, the tenth threshold value, the eleventh threshold value, the twelfth threshold value, the thirteenth threshold value, and / or the preset probability threshold value.

[0355] Optionally, the third information can further include a data reporting condition. The data reporting condition can include at least one of the following:

[0356] When the N+1th category of data is obtained, the N+1th category of data is reported;

[0357] When a number of times of obtaining the N+1th category of data is greater than or equal to an eighth threshold value, the N+1th category of data is reported. It should be understood that the N+1th category of data obtained multiple times can be the same or different;

[0358] When a data amount of the N+1th category of data obtained is greater than or equal to a ninth threshold value, the N+1th category of data is reported.

[0359] The eighth threshold value and the ninth threshold value can be previously agreed upon by the first device, the second device, and the third device, can be preset on the first device, the second device, and the third device, or can be carried in the third information and sent by the first device to the third device and / or the second device.

[0360] Exemplarily, the third information can include an identifier of each category of the first category to the Nth category.

[0361] Optionally, the third information can further include an identifier of the N+1th category.

[0362] S802, the third device can send the third information to the second device. Correspondingly, the second device can receive the third information sent by the third device. In this way, the second device can receive the third information from the first device.

[0363] Exemplarily, the third device can broadcast the third information. Correspondingly, at least one second device served by the third device can receive the third information broadcasted by the third device. The at least one second device served by the third device can include the second device A and the second device B as shown. Figure 8 In order to facilitate the second device to report the data indicated by the third information.

[0364] Exemplarily, taking the data used for model training as L1-RSRP for example, when the third device broadcasts the third information, the third device can also broadcast the configuration related to the third information. Correspondingly, the second device can receive the third information and the configuration related to the third information.

[0365] The configuration related to the third information can include the configuration information of set A and set B.

[0366] In this way, the third device can send each beam in set A and each beam in set B to the second device according to the configuration related to the third information, so that the second device can obtain the signal quality of each beam sent by the third device and identify the beam number of each beam sent by the third device at the same time.

[0367] It should be understood that the third device can broadcast the third information and the configuration related to the third information at the same time. The third device can also broadcast the third information and the configuration related to the third information sequentially.

[0368] Exemplarily, the third information can also be used to indicate the third data amount. For example, the third information can also include the third data amount threshold. The third data amount threshold can represent the data amount that the first device needs to collect when performing S801, or can represent the data amount indicated by the first device to collect through the third information.

[0369] The third data amount threshold can be greater than or equal to the first data amount threshold.

[0370] Optionally, the third data amount threshold can not be sent by the first device through the third information. The third data amount threshold can be pre-set on the third device, or the third data amount threshold can be agreed by the first device and the third device in advance.

[0371] The specific implementation principle of the third data amount threshold in this step can be referred to the specific implementation principle of the first data amount threshold in S702, which will not be repeated here. The specific implementation of the third data amount threshold in this step can make the data amount of the data indicated by the third information collected by the third device meet the third data amount threshold.

[0372] S803, the second device can send data of at least one of the first category to the Nth category to the third device. Correspondingly, the third device can receive the data of at least one of the first category to the Nth category reported by the second device.

[0373] Among the data of at least one of the first category to the Nth category sent by the second device, the identification of the category to which each data belongs can be contained. The data of at least one category, for example, L1-RSRP of at least one category.

[0374] S804, the third device can send the data reported by the second device to the first device. Correspondingly, the first device can receive the data reported by the second device sent by the third device. In this way, the first device can receive the data reported by the second device in S803.

[0375] The specific implementation principle and technical effect of S803-S804 can be referred to the specific implementation principle and technical effect of S703-S704, which will not be repeated here.

[0376] S805, the second device obtains the data of the N+1th category.

[0377] Exemplarily, if the second device judges that the obtained data has the data characteristics of the N+1th category, the second device determines to obtain the data of the N+1th category.

[0378] Optionally, the second device can judge whether the obtained data of the N+1th category meets the data reporting condition. If it meets, the second device can send the data of the N+1th category to the third device or the first device. For example, if it meets, the second device can perform S806.

[0379] It should be understood that the second device can be any one of at least one second device served by the third device. For example, the second device can be Figure 8 the second device B shown.

[0380] Optionally, S805 and S803 can be performed concurrently.

[0381] S806, the second device can send the data of the N+1th category to the third device. Correspondingly, the third device can receive the data of the N+1th category sent by the second device.

[0382] In this way, the second device can receive the N+1th category of data reported by the second device. The N+1th category of data sent by the second device can include an identifier of the N+1th category, for example, the second device reports the N+1th category of data in the manner of reporting any one of the first category to the Nth category of data, and fills the field for indicating the identifier of the category with information indicating the identifier of the N+1th category, so as to realize the definition of the identifier of the N+1th category by the second device.

[0383] Optionally, in the case where the identifier of the N+1th category is defined by the first device, the third information can also carry the identifier of the N+1th category.

[0384] Optionally, S806 and S803 can be performed concurrently.

[0385] S807, the third device can send the N+1th category of data to the first device. Correspondingly, the first device can receive the N+1th category of data sent by the third device. In this way, the first device can receive the N+1th category of data reported by the second device.

[0386] After the first device receives the N+1th category of data, the first device can determine that there is new data, which can indicate that the change of the communication environment causes the communication system to generate new data. Then the first device collects the new data, for example, the first device can perform S808, so as to perform model training (such as AI model training) by using a training set containing sufficient data of known categories (such as the first category to the Nth category) and new data, so as to improve the generalization of the AI model, so that the AI model can perform better when applied in the changed communication environment.

[0387] S808, the first device can send the first information to the third device. Correspondingly, the third device can receive the first information from the first device.

[0388] The first information is also used to indicate the reporting of the N+1th category of data.

[0389] For example, taking the first category with insufficient data amount in the first category of data to the Nth category of data collected in S801-S804 as an example, the first device can determine the first category with insufficient data amount in the N categories of data collected in S801-S804, and in the case where the first device receives the N+1th category of data, the first device can send the first information to the third device. Correspondingly, the third device can receive the first information from the first device.

[0390] The first device determines the specific implementation principle of the first category with insufficient data amount from the data set containing N categories collected from S801-S804. For example, the first device selects the specific implementation principle of the first category from the plurality of categories corresponding to the data set according to the filtering condition of the first category.

[0391] In this way, the first category with insufficient data amount in the collected data set and the new category (e.g., the N+1 category) caused by the change of the communication environment can be collected in a targeted manner, so that the data amount of each category from the first category to the N+1 category included in the training set for AI model training is sufficient.

[0392] S809, the third device can send the first information to the second device. Correspondingly, the second device can receive the first information sent by the third device.

[0393] For example, the first information can also be used to indicate the fourth data amount. For example, the first information can also include the fourth data amount threshold. The fourth data amount threshold can represent the data amount of the N+1 category collected by the first device through the first information.

[0394] It should be understood that the first information can include the second data amount threshold and the fourth data amount threshold. In this way, the data of the first category and the N+1 category can be collected in a targeted manner, so that the data amount of the first category and the N+1 category in the training set for model training is sufficient.

[0395] The fourth data amount threshold can be the same as the first threshold, or can be the same as the data amount of the target category.

[0396] The fourth data amount threshold can also be a value satisfying at least one of the following conditions:

[0397] The sum of the collected data amount of the N+1 category and the fourth data amount threshold is greater than or equal to the first threshold, and the collected data amount of the N+1 category, for example, the data amount of the N+1 category collected in S807;

[0398] The proportion of the sum of the collected data amount of the N+1 category and the fourth data amount threshold in the collected data is greater than or equal to the second threshold, and the collected data can include the data set and the collected data of the N+1 category;

[0399] The ratio of the sum of the collected data amount of the N+1 category and the fourth data amount threshold to the data amount of the target category is greater than or equal to the third threshold;

[0400] The absolute value of the difference between the proportion of the sum of the amount of the N+1th category of data collected and the fourth data amount threshold and the proportion of the target category is less than or equal to a fourth threshold. The proportion of the sum of the amount of the N+1th category of data collected and the fourth data amount threshold is the proportion of the sum of the amount of the N+1th category of data collected and the fourth data amount threshold in the data set.

[0401] Exemplarily, taking the L1-RSRP of the beam of the data of the first category as an example, the third device can select the third beam transmission number from the pre-stored correspondence between the data amount threshold and the beam transmission number. The third beam transmission number can be referred to as the third transmission number. In the correspondence between the data amount threshold and the beam transmission number, the sum of the second data amount threshold and the fourth data amount threshold corresponds to the third transmission number.

[0402] After the third device broadcasts the first information, the third device can transmit each beam in set A and each beam in set B to the second device according to the configuration related to the third transmission number and the third information. For specific implementation principles and technical effects, please refer to the specific implementation principles and technical effects of the third device transmitting each beam in set A and each beam in set B to the second device according to the configuration related to the first transmission number and the second information.

[0403] Optionally, taking the channel information of the data of the first category as an example, the third device can select the third broadcast number from the pre-stored correspondence between the data amount threshold and the broadcast number. In the correspondence between the data amount threshold and the beam transmission number, the sum of the second data amount threshold and the fourth data amount threshold corresponds to the third broadcast number.

[0404] The third device can periodically broadcast the first information according to the third broadcast number. So that when the second device receives the first information, it reports the channel information associated with the first information to the third device, so that the amount of the first category of data collected for the purpose can reach the second data amount threshold. The amount of the N+1th category of data collected for the purpose can reach the fourth data amount threshold.

[0405] Exemplarily, the first information can also be used to indicate the identification of the N+1th category. For example, the first information can also include the identification of the N+1th category. To indicate the third device to collect the data of the N+1th category and / or indicate the second device to report the data of the N+1th category.

[0406] S810, the second device can send the first category of data and / or the N+1th category of data to the third device. Correspondingly, the third device can receive the first category of data and / or the N+1th category of data reported by the second device.

[0407] Exemplarily, the second device can send the first category of data to the third device. Correspondingly, the third device can receive the first category of data reported by the second device. The specific implementation principle and technical effects can be referred to the specific implementation principle and technical effects of S708, and will not be repeated here.

[0408] The second device can send the N+1th category of data to the third device in a case that the obtained data is the N+1th category of data. Correspondingly, the third device can receive the N+1th category of data reported by the second device.

[0409] Optionally, the first information can further include a data reporting condition. The second device can send the N+1th category of data to the third device in a case that the obtained data is the N+1th category of data and the data reporting condition is met. Correspondingly, the third device can receive the N+1th category of data reported by the second device.

[0410] It can be understood that the second device sending the first category of data to the third device and the second device sending the N+1th category of data to the third device can be performed simultaneously or sequentially.

[0411] For example, the second device can send the first category of data to the third device first, and then send the N+1th category of data to the third device. The second device can send the N+1th category of data to the third device first, and then send the first category of data to the third device.

[0412] The specific implementation principle of the second device sending the N+1th category of data to the third device can be referred to the specific implementation principle of S806.

[0413] S811, the third device can send the first category of data and / or the N+1th category of data to the first device. Correspondingly, the first device can receive the data and / or the N+1th category of data sent by the third device. In this way, the first device can receive the reported first category of data and the N+1th category of data.

[0414] Exemplarily, the third device can send the first category of data to the first device in a case that the data amount of the first category of data received is greater than or equal to the second data amount threshold.

[0415] The third device can send the N+1th category of data to the first device in a case that the data amount of the N+1th category of data received is greater than or equal to the fourth data amount threshold.

[0416] Exemplarily, in a case that the third device receives data of the first category in an amount greater than or equal to the second data amount threshold, and the third device receives data of the N+1th category in an amount greater than or equal to the fourth data amount threshold, the third device can send, to the first device, the data of the first category and the data of the N+1th category received by the third device.

[0417] Exemplarily, in a case that the third device receives data of the first category in an amount greater than or equal to the second data amount threshold, and the third device receives data of the N+1th category in an amount greater than or equal to the fourth data amount threshold, the third device can send, to the first device, the data of the first category and the data of the N+1th category received by the third device.

[0418] In this way, the first device can collect the data of the first category and the data of the N+1th category in a targeted manner, so that the data amount of each category in a training set used for model training is better balanced, and thus the generalization of a model trained by using the training set can be better.

[0419] As shown in FIG. 1, Figure 8 The data transmission method provided by the embodiment of the present application can collect data of N categories, select a first category with insufficient data amount from the N categories, and collect data of the first category in a targeted manner. When data of an N+1th category is collected, it is determined that the change of the communication environment causes the data of the N+1th category, and the data of the N+1th category is collected in a targeted manner. The data of the N+1th category can be collected in a targeted manner, or new data can be collected in a targeted manner. The data amount of each category in a training set used for model training (such as AI model training) can be better balanced. The training set can include data of each category from the first category to the N+1th category collected by the embodiment of the present application, and the data amount of each category is sufficient. The AI model trained by using the training set can have better generalization, for example, the AI model can have better performance in the changed communication environment. In addition, the power consumption for obtaining redundant data, and the storage and reporting overhead can be reduced.

[0420] Still taking the example that the first device independently collects data of a category with insufficient data amount in a training set used for AI model training, Figure 9 FIG. 1 shows another interactive flowchart of the data transmission method provided by the embodiment of the present application.

[0421] As shown in FIG. 1, Figure 9As shown, the data transmission method provided by the embodiments of the present application can include S807-S807, S901-S904:

[0422] Figure 9 The embodiments shown Figure 8 The difference between the embodiments shown Figure 9 In the first device, if no first category with insufficient data amount is selected from the N-category data set collected from S801-S804, it can be indicated that the data amount of each category in the collected N-category data set is sufficient, and there is no category with insufficient data amount in the data set. Then the first device can perform targeted data collection on a new category (such as the N+1 category), which can be seen from S901-S904.

[0423] S901, the first device can send fourth information to the third device. Correspondingly, the third device can receive the fourth information from the first device.

[0424] The fourth information is used to indicate the reporting of the N+1 category data.

[0425] Exemplarily, when the first device receives the N+1 category data, the first device can send the fourth information to the third device. Correspondingly, the third device can receive the fourth information from the first device.

[0426] In this way, targeted data collection can be performed on the new category (such as the N+1 category) caused by the change of the communication environment, so that the data amount of each category from the first category to the N+1 category included in the training set for AI model training is sufficient.

[0427] Exemplarily, the fourth information can include an identifier of the N+1 category.

[0428] Optionally, the fourth information can also include a data reporting condition.

[0429] S902, the third device can send the fourth information to the second device. Correspondingly, the second device can receive the fourth information sent by the third device.

[0430] Exemplarily, the fourth information can also be used to indicate a fourth data amount. For example, the fourth information can also include a fourth data amount threshold. The fourth data amount threshold can represent the data amount of the N+1 category collected by the first device through the fourth information. In this way, by collecting the N+1 category data, the data amount of the N+1 category in the training set for model training can be sufficient.

[0431] Exemplarily, taking the L1-RSRP of the beam of the first category of data as an example, the third device can select the fourth beam transmission number corresponding to the fourth data amount threshold from the pre-stored correspondence between the data amount threshold and the beam transmission number.

[0432] After the third device broadcasts the fourth information, the third device can transmit each beam in set A and each beam in set B to the second device according to the configuration related to the fourth transmission number and the third information. For specific implementation principles and technical effects, please refer to the specific implementation principles and technical effects of the third device transmitting each beam in set A and each beam in set B to the second device according to the configuration related to the first transmission number and the second information.

[0433] Optionally, taking the channel information of the first category of data as an example, the third device can select the fourth broadcast number corresponding to the fourth data amount threshold from the pre-stored correspondence between the data amount threshold and the broadcast number.

[0434] The third device can periodically broadcast the fourth information according to the fourth broadcast number. In order to facilitate the second device to report the channel information associated with the fourth information to the third device when receiving the fourth information, so as to achieve that the amount of the N+1 category of data collected in a targeted manner reaches the fourth data amount threshold.

[0435] S903, the second device can send the N+1 category of data to the third device. Correspondingly, the third device can receive the N+1 category of data reported by the second device.

[0436] Exemplarily, the second device can send the N+1 category of data to the third device in a case where it is determined that the obtained data is the N+1 category of data. Correspondingly, the third device can receive the N+1 category of data reported by the second device. For specific implementation principles, please refer to the specific implementation principles of S806.

[0437] S904, the third device can send the N+1 category of data to the first device. Correspondingly, the first device can receive the N+1 category of data sent by the third device. In this way, the first device can receive the reported first category of data and the N+1 category of data.

[0438] Exemplarily, the third device can send the N+1 category of data to the first device every time the N+1 category of data is received.

[0439] Exemplarily, the third device can send the first category of data to the first device when the amount of the N+1 category of data received is greater than or equal to the fourth data amount threshold.

[0440] Exemplarily, in a case where the third device receives a data amount of the (N+1)th category of data greater than or equal to the fourth data amount threshold, the third device can send the (N+1)th category of data to the first device in batches at a preset frequency until the third device receives all the (N+1)th category of data.

[0441] In this way, the first device can collect the (N+1)th category of data in a targeted manner, so that the balance of the data amount of each category in the training set used for model training is better, and thus the generalization of the model trained by using the training set can be better.

[0442] As Figure 9 indicated, the data transmission method provided by the embodiments of the present application collects the data of the N categories and the (N+1)th category of data, and in a case where no first category with insufficient data amount is selected from the data set and it is determined that the change of the communication environment causes the generation of data of a new category (such as the (N+1)th category), the (N+1)th category is collected in a targeted manner. The balance of the data amount of each category in the training set used for model training (such as AI model training) can be better. The training set can include the data of each category from the first category to the (N+1)th category collected by the embodiments of the present application, and the data amount of each category is sufficient. In addition, the power consumption for obtaining redundant data, and the storage and reporting overhead can be reduced.

[0443] Still taking the example that the first device independently collects data of a category with insufficient data amount in the training set used for AI model training from the third device, the data transmission method provided by the embodiments of the present application can include S701-S704, S1001, S808-S811.

[0444] The data transmission method provided by the embodiments of the present application is different from the data transmission method provided by the embodiments shown in Figure 8 The data transmission method provided by the embodiments of the present application is different from the data transmission method provided by the embodiments shown in

[0445] S1001, the first device can classify the collected data by using a classification model to obtain the data of the first category to the Nth category, and also obtain the data of the (N+1)th category. The first device can select the first category with insufficient data amount from the data of the first category to the Nth category.

[0446] When the first category is selected and the data of the N+1 category is obtained, the first device can execute S808-S811 to achieve targeted data collection for the first category and the N+1 category, so that the data volume of each category in the training set used for model training (such as AI model training) is well balanced.

[0447] Taking the example of the first device being independent of the third device, and the first device collecting data specifically for categories with insufficient data in the training set used for training an AI model, the data transmission method provided in this application embodiment may include S701-S704, S1101, and S901-S904.

[0448] The data transmission method and embodiments of this application provide Figure 9 The difference in the data transmission method provided in the illustrated embodiment is that, in the data transmission method provided in this application embodiment, when data in the dataset is collected through S701-S704, the first device can use a classification model to classify the data in the dataset, obtaining not only data from the first category to the Nth category, but also data from the N+1th category. See S1101, which is not shown in the accompanying drawings.

[0449] S1101. The first device can use a classification model to classify the collected data to obtain data from the first category to the Nth category, and also obtain data from the N+1th category.

[0450] If the first device fails to select the first category (which has insufficient data) from the data of the first category to the data of the Nth category, the first device can execute S901-S904 to collect targeted data for the N+1th category, so that the data volume of each category in the training set used for model training (such as AI model training) is well balanced.

[0451] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in the embodiments of this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0452] The data transmission method of the embodiments of this application has been described above. The apparatus for executing the above method provided in the embodiments of this application is described below. Those skilled in the art will understand that the methods and apparatus can be combined with and referenced by each other, and the related apparatus provided in the embodiments of this application can execute the steps in the above method.

[0453] The data transmission method provided by the embodiments of the present application can be applied to an AI node or a terminal device with a communication function. The AI node can include a first device. The terminal device can include a second device. The specific form of the AI node and the specific device form of the terminal device can refer to the related description above, and will not be described here.

[0454] The embodiments of the present application provide a first device, which includes one or more processors and a memory. The memory is coupled to the one or more processors, and the memory is configured to store computer program code including computer instructions. The one or more processors invoke the computer instructions to cause the first device to perform the above method.

[0455] The embodiments of the present application provide a second device, which includes one or more processors and a memory. The memory is coupled to the one or more processors, and the memory is configured to store computer program code including computer instructions. The one or more processors invoke the computer instructions to cause the second device to perform the above method.

[0456] The embodiments of the present application provide a third device, which includes one or more processors and a memory. The memory is coupled to the one or more processors, and the memory is configured to store computer program code including computer instructions. The one or more processors invoke the computer instructions to cause the third device to perform the above method.

[0457] The embodiments of the present application provide a chip system applied to the first device or the second device, which includes one or more processors. The one or more processors are configured to invoke computer instructions to cause the first device to perform the above method or cause the second device to perform the above method.

[0458] The embodiments of the present application provide a computer readable storage medium including computer instructions. When the computer instructions run on the first device, the first device performs the above method. Or when the computer instructions run on the second device, the second device performs the above method. Or when the computer instructions run on the third device, the third device performs the above method.

[0459] The embodiments of the present application provide a computer program product including computer program code. When the computer program code runs on the first device, the first device performs the above method. Or when the computer program code runs on the second device, the second device performs the above method. Or when the computer program code runs on the third device, the third device performs the above method.

[0460] The embodiments of the present application also provide a communication system, which can include a first device and a second device.

[0461] The first device can send first information to the second device. Correspondingly, the second device can receive the first information from the first device. The first information is used to indicate reporting data of a first category, and the first category is one of multiple categories, and data of each category in the multiple categories is used for model (such as AI model) training.

[0462] The second device can send data of the first category to the first device. Correspondingly, the first device can receive the data of the first category from the second device.

[0463] The embodiments of the present application also provide a communication system, which can include a first device, a second device and a third device.

[0464] The first device can send first information to the third device. Correspondingly, the third device can receive the first information from the first device. The first information is used to indicate reporting data of a first category, and the first category is one of multiple categories, and data of each category in the multiple categories is used for model (such as AI model) training.

[0465] The third device can send the first information to the second device. Correspondingly, the second device can receive the first information sent by the third device.

[0466] The second device can send data of the first category to the third device. Correspondingly, the third device can receive the data of the first category sent by the second device.

[0467] The third device can send the data of the first category to the first device. Correspondingly, the first device can receive the data of the first category sent by the third device.

[0468] The embodiments of the present application also provide a computer readable storage medium. The computer readable storage medium stores a computer program. The computer program is executed by a processor to implement the above method. The method described in the above embodiments can be implemented by software, hardware, firmware or any combination thereof, in whole or in part. If implemented in software, the functions can be stored as one or more instructions or codes on a computer readable medium or transmitted on a computer readable medium. The computer readable medium can include computer storage medium and communication medium, and can also include any medium that can transfer computer programs from one place to another. The storage medium can be any target medium accessible by a computer.

[0469] In a possible implementation, the computer readable medium can include a RAM, a ROM, a compact disc read-only memory (CD-ROM) or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that is suitable for storing desired program code in the form of instructions or data structures and that can be accessed by a computer. Also, any connection is properly termed a computer readable medium. For example, if the software is transmitted from a website, server, or other remote source using a coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave, then the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of medium. Disk and disc, as used herein, include compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk, and Blu-ray® disc where disks usually reproduce data magnetically, while discs reproduce data optically with lasers. Combinations of the above should also be included within the scope of computer readable media.

[0470] The embodiment of the present application provides a computer program product, which comprises a computer program, and when the computer program is executed, the computer program causes the computer to execute the method.

[0471] It can be understood that the embodiment of the present application provides a data transmission apparatus, which can comprise a first device, a second device, a third device, a chip system, a computer readable storage medium and / or a communication system.

[0472] The embodiment of the present application is described with reference to the flowcharts and / or block diagrams of the method, device (system), and computer program product according to the embodiment of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to produce a machine, so that the instructions executed by the computer or other programmable data processing devices produce a device for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 The function of one flow or multiple flows and / or blocks Figure 1 The function of one flow or multiple flows and / or blocks

[0473] The above detailed description of the specific embodiments of the present application is provided for the purpose of further explaining the objects, technical solutions and advantages of the present application, and it should be understood that the above is only a specific embodiment of the present application and is not used to limit the protection scope of the present application, and any modification, equivalent replacement, improvement, etc. made on the basis of the technical solutions of the present application should be included in the protection scope of the present application.

Claims

1. A data transmission method, characterized in that, Applied to a first device, the method includes: Send a first message, which is used to instruct the reporting of data of a first category, where the first category is one of multiple categories, and the data of each of the multiple categories are used for model training; Receive data from the first category.

2. The method according to claim 1, characterized in that, The first device collects a dataset, and the first category of data is: data in the dataset that is insufficient in quantity, and the dataset includes data from the multiple categories.

3. The method according to claim 2, characterized in that, The data in the first category satisfy at least one of the following: In the dataset, the amount of data in the first category is less than a first threshold; The proportion of data in the first category in the dataset is less than the second threshold; The ratio of the amount of data in the first category to the amount of data in the target category is less than a third threshold, the target category is different from the first category, and the data of the target category is included in the dataset; The absolute value of the difference between the proportion of the first category and the proportion of the target category is greater than the fourth threshold. The proportion of the first category is the proportion of the data volume of the first category in the dataset, and the proportion of the target category is the proportion of the data volume of the target category in the dataset.

4. The method according to claim 2 or 3, characterized in that, The dataset also includes data from the second category to the Nth category, where N is an integer greater than 2.

5. The method according to claim 4, characterized in that, The categories in the dataset are obtained based on one or more of the following classifications: preset N types of data features, preset N types of beam sets, preset N types of cell sets, preset N types of resource configurations, preset N types of speeds, preset N types of time ranges, or preset N types of latitude and longitude ranges.

6. The method according to any one of claims 2-5, characterized in that, The data in the dataset includes: channel information, and / or, beam information.

7. The method according to claim 6, characterized in that, The channel information includes: the signal quality of the beam, which includes one or more of the following: Reference Signal Received Power (RSRP), Signal-to-Interference-plus-Noise Ratio (SINR), and Reference Signal Received Quality (RSRQ).

8. The method according to any one of claims 1-7, characterized in that, Before sending the first information, the method further includes: Send a second message, the second message being used to instruct the reporting of data for training the model; Receive data for training the model.

9. The method according to any one of claims 2-7, characterized in that, The first information is also used to instruct the reporting of data in the N+1th category, the dataset including data from the first category to the Nth category, where N is an integer greater than 2.

10. The method according to claim 9, characterized in that, Also includes: Receive data from the N+1th category.

11. The method according to claim 10, characterized in that, The data of the (N+1)th category is determined based on the i-th feature, which is the feature of the j-th category of the N categories in the dataset, where 1 ≤ i and i is an integer, and 1 ≤ j ≤ N and j is an integer.

12. The method according to claim 11, characterized in that, The first value determined based on the numerical value associated with the k-th feature of the data in the (N+1)-th category and the numerical value associated with the i-th feature of the data in the j-th category is less than the fifth threshold. The first value is used to indicate the correlation between the numerical value associated with the k-th feature and the numerical value associated with the i-th feature, where 1 ≤ k and k is an integer.

13. The method according to any one of claims 9-12, characterized in that, Before sending the first information, the method further includes: Send a third message, which is used to indicate the reporting of data of category N+1 and data of at least one category from category 1 to category N of the dataset, wherein category N+1 is distinct from category 1 to category N, and N is an integer greater than 2; Receive data from the (N+1)th category and data from at least one of the first category to the Nth category.

14. The method according to any one of claims 1-13, characterized in that, The first information includes the identifier of the first category.

15. A data transmission method, characterized in that, Applied to a second device, the method includes: Receive first information, the first information being used to instruct the reporting of data of a first category, the first category being one of multiple categories, the data of each of the multiple categories being used for model training; Send the data for the first category.

16. The method according to claim 15, characterized in that, Before receiving the first information, the method further includes: Receive second information, which instructs the reporting of data used for training the model; Send the data used for training the model.

17. The method according to claim 15, characterized in that, The first information is also used to instruct the reporting of data of the N+1th category, the dataset including data of the first category to the Nth category, where N is an integer greater than 2, the dataset being used for model training, and the data in the dataset being reported before the first information is received; The method further includes sending data of the N+1th category.

18. The method according to claim 17, characterized in that, Before receiving the first information, the method further includes: Receive third information, the third information being used to indicate the reporting of data of category N+1 and data of at least one category from category 1 to category N of the dataset, wherein category N+1 is distinct from category 1 to category N; Send data from category N+1 and data from the first category to at least one of the categories N.

19. A first device, characterized in that, The first device includes: one or more processors and memory; The memory is coupled to the one or more processors, the memory being used to store computer program code, the computer program code including computer instructions, the one or more processors invoking the computer instructions to cause the first device to perform the method as described in any one of claims 1 to 14.

20. A second device, characterized in that, The second device includes: one or more processors and memory; The memory is coupled to the one or more processors, the memory being used to store computer program code, the computer program code including computer instructions, the one or more processors invoking the computer instructions to cause the second device to perform the method as described in any one of claims 15 to 18.

21. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes computer instructions that, when executed on a first device, cause the first device to perform the method as described in any one of claims 1 to 14; or, when executed on a second device, cause the second device to perform the method as described in any one of claims 15 to 18.

22. A computer program product, characterized in that, The computer program product includes computer program code that, when executed on a first device, causes the first device to perform the method as described in any one of claims 1 to 14; or, when executed on a second device, causes the second device to perform the method as described in any one of claims 15 to 18.

Images