Communication method and related apparatus

Abstract

A communication method and a related apparatus. In the method, after a first communication apparatus receives a first signal, the first communication apparatus may estimate, on the basis of first information indicating channel sparsity, a channel corresponding to the first signal, so as to obtain first estimation information; then, the first communication apparatus may send second information indicating the first estimation information and the first signal, such that a receiver of the second information may use the second information for model management of a first model, the first model being used for channel estimation, that is, the first model may be a channel estimation model; in other words, the second information sent by the first communication apparatus may be used for model management of a channel estimation model. In this way, a communication device on which a channel estimation model is deployed can perform channel estimation on a received signal on the basis of the channel estimation model, thereby improving the processing efficiency in sparse signal estimation.

Description

A communication method and related apparatus

[0001] This application claims priority to Chinese Patent Application No. 202410940653.1, filed on July 12, 2024, entitled "A Communication Method and Related Device", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of communications, and more particularly to a communication method and related apparatus. Background Technology

[0003] Wireless communication can be a transmission communication between two or more communication devices that does not propagate through conductors or cables. Generally, the two or more communication devices include network devices and terminal devices, or the two or more communication devices include different terminal devices.

[0004] Currently, in the transmission of communication signals, sparse signal estimation can instruct the signal receiver to recover or estimate a sparse signal from a limited set of observations. A sparse signal can be represented as a linear combination of a few non-zero or significant elements, while the remainder is zero or close to zero. Through sparse signal estimation, the number of observations or measurements required by the signal receiver can be reduced, thereby lowering energy consumption and improving efficiency.

[0005] However, how to improve the processing efficiency of sparse signal estimation is a technical problem that urgently needs to be solved. Summary of the Invention

[0006] This application provides a communication method and related apparatus for improving the processing efficiency of sparse signal estimation.

[0007] This application provides a communication method executed by a first communication device. The first communication device can be a communication equipment (such as a terminal device or network device), or it can be a component of the communication equipment (e.g., a circuit or chip responsible for communication functions, such as a modem chip, also known as a baseband chip, or a system-on-a-chip (SoC) chip containing a modem core, or a system-in-package (SIP) chip, etc.), or it can be a logic module or software capable of implementing all or part of the functions of the communication equipment. In this method, the first communication device receives first information indicating channel sparsity; the first communication device receives a first signal; the first communication device sends second information indicating first estimation information and the first signal; the second information is used for model management of a first model, which is used for channel estimation; the first estimation information is obtained by estimating the channel corresponding to the first signal based on the first information.

[0008] Based on the above scheme, after receiving the first signal, the first communication device can estimate the channel corresponding to the first signal based on first information indicating channel sparsity, thus obtaining first estimation information. Subsequently, the first communication device can send second information indicating the first estimation information and the first signal, enabling the receiver of the second information to use it for model management of the first model, which is used for channel estimation; that is, the first model can be a channel estimation model. In other words, the second information sent by the first communication device can be used for model management of the channel estimation model. In this way, communication equipment deploying the channel estimation model can perform channel estimation on the received signal based on the channel estimation model, thereby improving the processing efficiency of sparse signal estimation.

[0009] Furthermore, the first estimation information indicated by the second information sent by the first communication device is determined based on the first information indicating channel sparsity. This second information can also be used for model management of the channel estimation model. In this way, the first communication device can perform channel estimation based on the channel sparsity specified by other communication devices to obtain the first estimation information, and send information containing this first estimation information. This allows other communication devices (e.g., the receiver of the second information) to perform model management of the channel estimation model based on the first estimation information corresponding to the specified channel sparsity, thereby improving model management efficiency.

[0010] It should be understood that when a communication device performs sparse signal estimation on a signal, it can be interpreted as the communication device estimating the channel (or transmission channel) corresponding to that signal. Correspondingly, the result of sparse signal estimation for a signal can be understood as the estimation result (or channel estimation result) of the channel corresponding to that signal.

[0011] In this application, the model (e.g., the first model, the second model described below, etc.) may include a mathematical model, an AI model, a neural network model, an AI neural network model, a machine learning model, or an AI processing model, etc.

[0012] For example, a model (e.g., a first model, a second model described later, etc.) used for channel estimation can be understood as follows: the input of the model may include the signal received by the signal receiver and transmitted through the wireless channel, and the output of the model may include an estimation result of the wireless channel (e.g., the second estimation information described later, etc.). For example, the estimation result may include one or more of the following: the estimated index of the non-zero element, the estimated value of the non-zero element, OR, and residual error.

[0013] In this application, model management may include one or more of the following: model monitoring, model training, model inference, model scheduling, model deployment, model updating, model generation, model switching, or function rollback.

[0014] As an example, when the first communication device deploys the first model, after the first communication device sends the second information, the recipient of the second information (e.g., the second communication device) can perform model management on the first model deployed on the first communication device based on the second information. The model management may include one or more of the following: model monitoring, model scheduling, model switching, function rollback, or model generation.

[0015] For example, the receiver can determine the performance of the first model based on the second information, thereby enabling model monitoring (e.g., monitoring of model performance) of the first model deployed on the first communication device.

[0016] For example, the receiver can schedule whether the first communication device performs channel estimation processing for one or more other signal corresponding channels based on the first model based on the second information, thereby realizing model scheduling of the first model deployed on the first communication device.

[0017] For example, if the first communication device deploys two or more models for channel estimation, the receiver can determine the performance of the first model based on the second information, and determine whether to instruct the first communication device to perform channel estimation processing on the channel corresponding to the signal based on other models based on the performance of the first model, thereby realizing model switching of the first model deployed on the first communication device.

[0018] As another example, when the first communication device deploys the second model, after the first communication device sends the second information, the recipient of the second information (e.g., the second communication device) can perform model management on the first model deployed on the second communication device based on the second information. The model management may include one or more of the following: model monitoring, model scheduling, model switching, function rollback, model generation, model updating, or model training.

[0019] For example, the receiver can determine the performance of the first model based on the second information, thereby enabling model monitoring (e.g., monitoring of model performance) of the first model deployed on the second communication device.

[0020] For example, the receiver can determine the performance of the first model based on the second information, and determine whether to perform channel estimation processing on the received signal based on the model, thereby realizing model management of whether to perform function rollback on the first model deployed on the second communication device.

[0021] For example, the receiver can perform model processing (such as fine-tuning, adjustment, and updating) on ​​the first model based on the second information to update the first model deployed on the second communication device.

[0022] It should be understood that when a model is deployed on a communication device (e.g., the first model is deployed on the first communication device, the first model is deployed on the second communication device, etc.), it can be understood that after the communication device obtains the model parameters of the model, it obtains / generates / constructs the model based on the model parameters of the model, and then the communication device can perform data processing based on the model.

[0023] It should be understood that after a signal sender transmits a signal, and the signal is transmitted through a wireless channel, the signal receiver can receive the signal. During this process, the wireless channel will inevitably affect the signal transmitted by the signal sender (e.g., multipath propagation, interference, attenuation, etc.), which may result in the signal received by the signal receiver being different from the signal transmitted by the signal sender. Accordingly, in the above scheme, the first signal received by the first communication device may be different from the first signal transmitted by the signal sender (e.g., the second communication device). In the above scheme, the first signal included in the second information transmitted by the first communication device is specifically the first signal received (or parsed, or detected) by the first communication device.

[0024] In this application, channel sparsity can indicate the number of elements to be estimated (or the number of non-zero elements) in the channel information, or the ratio of the number of elements to be estimated (or the number of non-zero elements) in the channel information to the dimension of the channel information.

[0025] Optionally, in addition to indicating channel sparsity, the first information may also indicate at least one of the following: residual error, capability configuration of the first communication device (e.g., the capability configuration indicates a certain capability of the first communication device, including one or more of supported bandwidth, number of antennas, channel estimation algorithm, channel estimation quality or performance, enabling the first communication device to subsequently process the first signal and obtain the first estimation information based on the configuration indicated by the capability configuration), and configuration information of the reference signal. In this way, the first communication device can perform channel estimation based on parameters specified by other communication devices to obtain the first estimation information, and send second information indicating the first estimation information, enabling the other communication devices to perform model management on the channel estimation model based on the first estimation information corresponding to the specified parameters, thereby improving model management efficiency.

[0026] Optionally, before receiving the first information, the first communication device may send one or more capability information supported by the first communication device (e.g., at least one of one or more residual errors, one or more capability information, and configuration information of one or more reference signals) so that the receiver (e.g., the second communication device) can determine the capability configuration indicated by the first information based on the received one or more capability information.

[0027] It should be understood that residual error can indicate the error, difference, or deviation between the signal actually received by the signal receiver and the signal constructed by the signal receiver based on the estimated information obtained from channel estimation.

[0028] In one possible implementation of the first aspect, the method further includes: the first communication device transmitting third information for determining the first information; wherein the third information indicates one or more estimations of a reference signal, each estimation indicating a residual error corresponding to one of the one or more channel sparsities.

[0029] It should be understood that one or more estimation information of the reference signal can be understood as one or more estimation information obtained by the first communication device after receiving the reference signal and performing one or more channel estimations based on the reference signal.

[0030] Based on the above scheme, the first communication device can also send third information for determining the first information, wherein the third information indicates one or more estimation information of the reference signal, and different estimation information can indicate the residual error corresponding to different channel sparsity. In this way, the receiver of the third information (e.g., the second communication device) can determine the residual error corresponding to one or more channel sparsities through the third information, so that the receiver can instruct the first communication device to obtain estimation information based on one of the channel sparsities through the first information, and then the receiver can obtain the first estimation information corresponding to the specified channel sparsity through the second information, and perform model management on the channel estimation model based on the first estimation information corresponding to the specified channel sparsity, so as to improve the model management efficiency.

[0031] In one possible implementation of the first aspect, the model parameters of the first model are determined based on the third information.

[0032] Based on the above scheme, the third information sent by the first communication device can be used to determine the model parameters of the first model (such as the number of neural network levels and layers contained in the first model). That is, the receiver of the third information (such as the second communication device) can use the third information for the management of the first model (such as model update, model generation, etc.), so that the receiver obtains a first model that is compatible with the estimation information obtained by the first communication device from channel estimation of the reference signal.

[0033] Optionally, model parameters may include one or more of the following: model hyperparameters, model dataset (including model input data and corresponding label data), model structure, model weights, and model parameters.

[0034] In one possible implementation of the first aspect, the model parameters of the first model are determined based on the first information.

[0035] Based on the above scheme, the first information received by the first communication device can be used to determine the model parameters of the first model (such as the number of neural network levels and layers contained in the first model). That is, the first communication device can use the first information for the management of the first model (such as model update, model generation, etc.), so that the first communication device obtains a first model that is compatible with the channel coefficient indicated by the first information.

[0036] In one possible implementation of the first aspect, the method further includes: the first communication device receiving fourth information, the fourth information being used to instruct a second model, the second model being obtained by training a model based on the first model.

[0037] Based on the above scheme, the first communication device can also receive fourth information indicating the second model, which can be obtained by training the first model. In other words, the recipient of the first information (e.g., the second communication device) can train the first model based on the first information to obtain the second model, and deploy the second model to the first communication device through the fourth information to achieve model training and model deployment.

[0038] In one possible implementation of the first aspect, the method further includes: the first communication device sending fifth information indicating the difference between the first estimation information and the second estimation information, the second estimation information being estimated based on the first model.

[0039] Based on the above scheme, the first communication device can estimate the received first signal using a first model to obtain second estimation information. Furthermore, the first communication device can estimate the received first signal using other methods to obtain first estimation information. Subsequently, the first communication device sends fifth information to indicate the difference between the first and second estimation information, enabling the recipient of the fifth information to obtain the difference based on the fifth information and to perform model management on the first model based on the difference.

[0040] Optionally, when the first model is deployed on the first communication device, the first communication device can obtain the second estimation information based on the first model deployed locally and send the fifth information, so that the second communication device can perform model management on the first model based on the differences indicated by the fifth information.

[0041] Optionally, if the first model is deployed on another communication device (e.g., the second communication device), the second communication device can obtain the second estimation information by receiving information from the other communication device (or by obtaining the first model deployed by itself), so that the second communication device can perform model management on the first model based on the first estimation information and the second estimation information (e.g., the difference between the two), that is, the first communication device may not send the fifth information.

[0042] It should be understood that the first communication device can estimate the received first signal in a way different from the first model to obtain the first estimation information, which will be described below with reference to some implementation examples.

[0043] As an example of implementation, this other approach can be a traditional channel estimation method, including but not limited to channel estimation methods based on orthogonal matching pursuit (OMP) algorithms, Bayesian learning-based algorithms, etc. In this way, the first estimation information can serve as the target value (or expected value, or ground truth, etc.) corresponding to the second estimation information, enabling the receiver of the fifth information to manage the first model based on the difference between the first estimation information indicated by the fifth information and the target value.

[0044] As another implementation example, this other approach can be a processing method for models other than the first model.

[0045] For example, if the model capability of the other model is superior to that of the first model, the model inference accuracy and / or the communication performance (e.g., system performance, link performance) corresponding to the model output of the former may be superior to that of the latter. Accordingly, the first estimated information obtained from the output of the other model can be used as the target value (or expected value, or benchmark truth, etc.) corresponding to the second estimated information, so that the receiver of the fifth information can manage the first model based on the difference between the first estimated information indicated by the fifth information and the target value.

[0046] For example, if the model capability of the other model is inferior to that of the first model, the model inference accuracy and / or the communication performance (e.g., system performance, link performance) corresponding to the model output of the former may be inferior to that of the latter. Accordingly, the second estimated information obtained by the output of the first model can be used as the target value (or expected value, or benchmark truth, etc.) corresponding to the first estimated information obtained by the output of the other model, so that the receiver of the fifth information can realize model management of the other model based on the difference between the first estimated information indicated by the fifth information and the target value.

[0047] A second aspect of this application provides a communication method executed by a second communication device. The second communication device can be a communication equipment (e.g., a terminal device or a network device), or it can be a component of the communication equipment (e.g., a circuit or chip responsible for communication functions, such as a modem chip, also known as a baseband chip, or a SoC chip or SIP chip containing a modem core, etc.), or it can be a logic module or software capable of implementing all or part of the functions of the communication equipment. In this method, the second communication device sends first information indicating channel sparsity; the second communication device sends a first signal; the second communication device receives second information indicating first estimation information and the first signal; the second information is used for model management of a first model, which is used for channel estimation; the first estimation information is obtained by estimating the channel corresponding to the first signal based on the first information.

[0048] Based on the above scheme, the second communication device can send first information and a first signal to the first communication device, enabling the first communication device to estimate the channel corresponding to the first signal based on the first information indicating channel sparsity, thereby obtaining first estimation information. Subsequently, the first communication device can send second information indicating the first estimation information and the first signal to the second communication device, allowing the second communication device to use the second information for model management of the first model, which is used for channel estimation; that is, the first model can be a channel estimation model. In other words, the second information received by the second communication device can be used for model management of the channel estimation model. In this way, communication equipment deploying a channel estimation model can perform channel estimation on received signals based on the channel estimation model, thereby improving the processing efficiency of sparse signal estimation.

[0049] Furthermore, the first estimation information indicated by the second information received by the second communication device is determined based on the first information indicating channel sparsity. This second information can also be used for model management of the channel estimation model. In this way, the first communication device can perform channel estimation based on the channel sparsity specified by the second communication device to obtain the first estimation information, and then send information containing this first estimation information. This allows the second communication device to perform model management of the channel estimation model based on the first estimation information corresponding to the specified channel sparsity, thereby improving model management efficiency.

[0050] Optionally, in addition to indicating channel sparsity, the first information may also indicate at least one of the following: residual error, capability configuration of the first communication device (e.g., the capability configuration indicates a certain capability of the first communication device, including one or more of supported bandwidth, number of antennas, channel estimation algorithm, channel estimation quality or performance, enabling the first communication device to subsequently process the first signal and obtain the first estimation information based on the configuration indicated by the capability configuration), and configuration information of the reference signal. In this way, the first communication device can perform channel estimation based on parameters specified by the second communication device to obtain the first estimation information, and send second information containing the first estimation information, enabling the second communication device to perform model management on the channel estimation model based on the first estimation information corresponding to the specified parameters, thereby improving model management efficiency.

[0051] Optionally, before receiving the first information, the first communication device may send one or more capability information supported by the first communication device (e.g., at least one of one or more residual errors, one or more capability information, and configuration information of one or more reference signals) so that the second communication device can determine the capability configuration indicated by the first information based on the received one or more capability information.

[0052] In one possible implementation of the second aspect, the method further includes: the second communication device receiving third information for determining the first information; wherein the third information indicates one or more estimation information of a reference signal, each estimation information indicating a residual error corresponding to one of the one or more channel sparsities.

[0053] Based on the above scheme, the second communication device can also receive third information for determining the first information, wherein the third information indicates one or more estimation information of the reference signal, and different estimation information can indicate residual errors corresponding to different channel sparsities. In this way, the second communication device can determine the residual errors corresponding to one or more channel sparsities through the third information, so that the second communication device can instruct the first communication device to obtain estimation information based on one of the channel sparsities through the first information, thereby enabling the second communication device to obtain the first estimation information corresponding to the specified channel sparsity through the second information, and to perform model management on the channel estimation model based on the first estimation information corresponding to the specified channel sparsity, so as to improve the model management efficiency.

[0054] In one possible implementation of the second aspect, the model parameters of the first model are determined based on the third information.

[0055] Based on the above scheme, the third information received by the second communication device can be used to determine the model parameters of the first model (such as the number of neural network levels and layers contained in the first model). That is, the second communication device can use the third information for the management of the first model (such as model update, model generation, etc.), so that the second communication device obtains a first model that is compatible with the estimation information obtained by the first communication device from channel estimation of the reference signal.

[0056] In one possible implementation of the second aspect, the model parameters of the first model are determined based on the first information.

[0057] Based on the above scheme, the first information sent by the second communication device can be used to determine the model parameters of the first model (e.g., the number of neural network levels and layers contained in the first model). That is, the receiver of the first information (e.g., the first communication device) can use the first information for the management of the first model (e.g., model update, model generation, etc.), so that the receiver of the first information obtains a first model that is compatible with the channel coefficient indicated by the first information.

[0058] In one possible implementation of the second aspect, the method further includes: the second communication device sending fourth information, the fourth information being used to instruct a second model, the second model being obtained by training the model based on the first model.

[0059] Based on the above scheme, the second communication device can also send fourth information indicating the second model, which can be obtained by training the first model. In other words, the second communication device can train the first model based on the first information to obtain the second model, and deploy the second model to the first communication device through the fourth information, so as to realize model training and model deployment.

[0060] In one possible implementation of the second aspect, the method further includes: the second communication device receiving fifth information indicating the difference between the first estimation information and the second estimation information, the second estimation information being estimated based on the first model.

[0061] Based on the above scheme, the first communication device can estimate the received first signal using a first model to obtain second estimation information. Furthermore, the first communication device can estimate the received first signal using other methods to obtain first estimation information. Subsequently, the first communication device sends fifth information to indicate the difference between the first and second estimation information, enabling the second communication device to obtain the difference based on the fifth information and to perform model management on the first model based on the difference.

[0062] A third aspect of this application provides a communication method executed by a first communication device. The first communication device can be a communication equipment (e.g., a terminal device or a network device), or it can be a component of the communication equipment (e.g., a circuit or chip responsible for communication functions, such as a modem chip, also known as a baseband chip, or a SoC chip or SIP chip containing a modem core, etc.), or it can be a logic module or software capable of implementing all or part of the functions of the communication equipment. In this method, the first communication device receives a second signal; the first communication device sends sixth information, which indicates one or more sparsity-corresponding estimation information and the second signal; the sixth information is used for model management of a first model, which is used for channel estimation; the estimation information is obtained by estimating the channel corresponding to the second signal.

[0063] Based on the above scheme, after receiving the second signal, the first communication device can estimate the channel corresponding to the second signal to obtain estimation information corresponding to one or more sparsities. Subsequently, the first communication device can send sixth information indicating the one or more sparsity-corresponding estimation information and the second signal, so that the receiver of the sixth information can use it for model management of the first model, which is used for channel estimation; that is, the first model can be a channel estimation model. In other words, the sixth information sent by the first communication device can be used for model management of the channel estimation model. In this way, communication equipment deploying the channel estimation model can perform channel estimation on the received signal based on the channel estimation model, thereby improving the processing efficiency of sparse signal estimation.

[0064] Furthermore, the estimation information indicated by the sixth information sent by the first communication device corresponds to one or more sparsities; that is, the estimation information indicated by the sixth information can indicate the estimation information corresponding to different channel sparsities. In this way, the receiver of the sixth information (e.g., the second communication device) can determine the estimation information corresponding to one or more channel sparsities through the sixth information, thereby enabling the receiver to manage the channel estimation model using the estimation information corresponding to one or more channel sparsities, thus improving model management efficiency.

[0065] It should be understood that the first communication device can estimate the received second signal in a manner different from the first model to obtain estimation information corresponding to one or more sparsities. The following will describe this in conjunction with some implementation examples.

[0066] As an example of implementation, this other approach can be a traditional channel estimation method, including but not limited to channel estimation methods based on orthogonal matching pursuit (OMP) algorithms, Bayesian learning-based algorithms, etc.

[0067] As another implementation example, this other approach can be a processing method for models different from the first model. For example, if the model capability of the other model is superior to that of the first model, the former's model inference accuracy and / or the communication performance (e.g., system performance, link performance) corresponding to the model output can be superior to the latter. Accordingly, the estimated information corresponding to one or more sparsities obtained from the output of the other model can be used as the target value (or expected value, or baseline truth, etc.) corresponding to the estimated information output by the first model, so that the receiver of the sixth information can manage the first model based on the difference between the estimated information output by the first model indicated by the sixth information and the target value.

[0068] For example, if the model capability of the other model is inferior to that of the first model, the model inference accuracy and / or the communication performance (e.g., system performance, link performance) corresponding to the model output of the former may be inferior to that of the latter. Accordingly, the estimated information obtained by the output of the first model can be used as the target value (or expected value, or benchmark truth, etc.) corresponding to one or more estimated information output by the other model, so that the recipient of the sixth information can realize model management of the other model based on the difference between one or more estimated information indicated by the sixth information and the target value.

[0069] Optionally, the sixth message transmitted by the first communication device includes at least one of the following:

[0070] The first indication information indicates the one or more sparsities;

[0071] The second indication information indicates one or more residual errors corresponding to the one or more sparsities;

[0072] The third instruction information indicates the capability of the first communication device; or...

[0073] The fourth indication information is the configuration information of the reference signal.

[0074] In one possible implementation of the third aspect, the model parameters of the first model are determined based on the sixth information.

[0075] Based on the above scheme, the sixth information sent by the first communication device can be used to determine the model parameters of the first model (such as the number of neural network levels and layers contained in the first model). That is, the receiver of the sixth information (such as the second communication device) can use the sixth information for the management of the first model (such as model update, model generation, etc.), so that the receiver obtains a first model that is compatible with the estimation information obtained by the first communication device from channel estimation of the second signal.

[0076] In one possible implementation of the third aspect, the method further includes: the first communication device receiving seventh information, the seventh information being used to instruct a second model, the second model being obtained by training the model based on the first model.

[0077] Based on the above scheme, the first communication device can also receive seventh information indicating the second model, which can be obtained by training the first model. In other words, the recipient of the sixth information (e.g., the second communication device) can train the first model based on the sixth information to obtain the second model, and deploy the second model to the first communication device through the seventh information, so as to realize model training and model deployment.

[0078] A fourth aspect of this application provides a communication method performed by a second communication device. The second communication device can be a communication device (e.g., a terminal device or a network device), or it can be a component of the communication device (e.g., a circuit or chip responsible for communication functions, such as a modem chip, also known as a baseband chip, or a SoC chip or SIP chip containing a modem core, etc.), or it can be a logic module or software capable of implementing all or part of the functions of the communication device. In this method, the second communication device transmits a second signal; the second communication device receives sixth information, which indicates one or more sparsity-corresponding estimation information and the second signal; the sixth information is used for model management of a first model, which is used for channel estimation; the estimation information is obtained by estimating the channel corresponding to the second signal.

[0079] Based on the above scheme, after the second communication device sends a second signal to the first communication device, the first communication device can estimate the channel corresponding to the second signal to obtain estimation information corresponding to one or more sparsities. Subsequently, the first communication device can send the estimation information corresponding to the one or more sparsities and a sixth piece of information about the second signal to the second communication device, enabling the second communication device to use the sixth piece of information for model management of a first model. This first model is used for channel estimation; that is, the first model can be a channel estimation model. In other words, the sixth piece of information received by the second communication device can be used for model management of the channel estimation model. In this way, communication devices deploying a channel estimation model can perform channel estimation on received signals based on this model, thereby improving the processing efficiency of sparse signal estimation.

[0080] Furthermore, the estimation information indicated by the sixth information received by the second communication device corresponds to one or more sparsities; that is, the estimation information indicated by the sixth information can indicate the estimation information corresponding to different channel sparsities. In this way, the second communication device can determine the estimation information corresponding to one or more channel sparsities through the sixth information, thereby enabling the second communication device to perform model management on the channel estimation model using the estimation information corresponding to one or more channel sparsities, thus improving model management efficiency.

[0081] Optionally, the sixth message transmitted by the first communication device includes at least one of the following:

[0082] The first indication information indicates the one or more sparsities;

[0083] The second indication information indicates one or more residual errors corresponding to the one or more sparsities;

[0084] The third instruction information indicates the capability of the first communication device; or...

[0085] The fourth indication information is the configuration information of the reference signal.

[0086] In one possible implementation of the fourth aspect, the model parameters of the first model are determined based on the sixth information.

[0087] Based on the above scheme, the sixth information received by the second communication device can be used to determine the model parameters of the first model (such as the number of neural network levels and layers contained in the first model). That is, the second communication device can use the sixth information for the management of the first model (such as model update, model generation, etc.), so that the second communication device obtains a first model that is compatible with the estimation information obtained by the first communication device from channel estimation of the second signal.

[0088] In one possible implementation of the fourth aspect, the method further includes: the second communication device sending seventh information, the seventh information being used to instruct a second model, the second model being obtained by training the model based on the first model.

[0089] Based on the above scheme, the second communication device can also send a seventh message indicating the second model to the first communication device. This second model can be obtained by training the first model. In other words, the second communication device can train the first model based on the sixth message to obtain the second model, and then deploy the second model to the first communication device through the seventh message, thereby realizing model training and model deployment.

[0090] A fifth aspect of this application provides a communication method performed by a third communication device. The third communication device can be a communication device (e.g., a terminal device or a network device), or it can be a component of the communication device (e.g., a circuit or chip responsible for communication functions, such as a modem chip, also known as a baseband chip, or a SoC chip or SIP chip containing a modem core, etc.), or it can be a logic module or software capable of implementing all or part of the functions of the communication device. In this method, the third communication device receives a third signal; the third communication device processes the third signal based on a first model to obtain estimation information corresponding to the third signal; wherein the first model is used for channel estimation.

[0091] Optionally, the third communication device receives eighth information, which is used to indicate the first model. Alternatively, the third communication device obtains / determines the first model through a pre-configured method.

[0092] Based on the above scheme, after receiving the third signal, the third communication device can estimate the channel corresponding to the third signal to obtain the estimated information corresponding to the third signal. The first model is used for channel estimation; that is, the first model can be a channel estimation model. In this way, communication equipment deploying the channel estimation model can perform channel estimation on the received signal based on the channel estimation model, thereby improving the processing efficiency of sparse signal estimation.

[0093] A sixth aspect of this application provides a communication method performed by a fourth communication device. The fourth communication device can be a communication device (e.g., a terminal device or a network device), or it can be a component of the communication device (e.g., a circuit or chip responsible for communication functions, such as a modem chip, also known as a baseband chip, or a SoC chip or SIP chip containing a modem core, etc.), or it can be a logic module or software capable of implementing all or part of the functions of the communication device. In this method, the fourth communication device sends eighth information, which is used to indicate a first model; wherein the first model is used to process a signal (e.g., a third signal) to obtain estimation information corresponding to the signal; wherein the first model is used for channel estimation.

[0094] Based on the above scheme, after the fourth communication device sends the eighth information, the receiver of the eighth information (e.g., the third communication device) can estimate the channel corresponding to the received signal to obtain the estimated information corresponding to the signal. Here, the first model is used for channel estimation; that is, the first model can be a channel estimation model. In this way, communication devices deploying the channel estimation model can perform channel estimation on the received signal based on the channel estimation model, thereby improving the processing efficiency of sparse signal estimation.

[0095] In one possible implementation of any of the first to sixth aspects, the first model includes an N-layer neural network, wherein the input of the (i+1)th layer of the N-layer neural network is determined based on the output of the i-th layer of the neural network, where N is a positive integer and i takes values ​​from 1 to N-1.

[0096] Based on the above scheme, the first model can contain N layers of neural networks. The two adjacent neural networks can satisfy the above correlation, so that the next layer of neural network can obtain the feature information accumulated by the previous layer of neural network, thereby improving the processing performance of the next layer of neural network.

[0097] In one possible implementation of any of the first to sixth aspects, the first model includes a P-level neural network, wherein one level of the P-level neural network includes the N-layer neural network, and the different levels of the neural network contain different neural networks, where P is a positive integer.

[0098] Optionally, in a P-level neural network, the number of neural network layers contained in different levels of neural networks is N.

[0099] Based on the above scheme, the first model may include a P-level neural network to achieve sparse signal estimation through joint processing of one or more levels of neural networks.

[0100] In any possible implementation of any of the first to sixth aspects, the P-level neural network satisfies any of the following:

[0101] In this P-level neural network, the channel estimation parameters corresponding to one or more layers of neural networks contained in any level of the neural network are the same;

[0102] In this P-level neural network, the channel estimation parameters corresponding to one or more layers of neural networks contained in the (p+1)th level neural network are determined based on the channel estimation parameters corresponding to one or more layers of neural networks contained in the p-level neural network; or,

[0103] In this P-level neural network, the channel estimation parameters corresponding to the j-th layer of the p+1-th level neural network are determined based on the channel estimation parameters corresponding to the j-th layer of the p-th level neural network, where j is a positive integer.

[0104] Based on the above scheme, in the P-level neural network included in the first model, the channel estimation parameters corresponding to different layers of the same level neural network can be the same, so that different layers of the same level neural network can obtain channel estimation results based on the same channel estimation parameters, thereby simplifying the implementation and improving the model processing efficiency.

[0105] And / or, in the P-level neural network included in the first model, the channel estimation parameters corresponding to two adjacent neural networks can satisfy the above correlation, so that the next level neural network can obtain the feature information accumulated by the previous level neural network, thereby improving the processing performance of the next level neural network.

[0106] Optionally, in any one or more layers of the P-level neural network, the j-th layer satisfies at least one of the following conditions, where j is a positive integer:

[0107] The j-th layer neural network includes an attention module, and the input of the attention module includes the j-th key K. j and the j-th value V j K j and the V j It is determined based on the channel estimation parameters corresponding to the j-th layer neural network;

[0108] The j-th layer neural network includes an attention module, and the input to the attention module includes the j-th query Q. j Q j Satisfy: When j takes the value 1, the Q j It is determined based on the received signal; and / or, when j is greater than 1, this Q... jIt is determined based on the first output of the feedforward module contained in the (j-1)th layer of the neural network;

[0109] The j-th layer neural network includes a feedforward module, and the first output of the feedforward module in the j-th layer neural network is used to determine the (j+1)-th query Q input to the attention module in the (j+1)-th layer neural network. j+1 ;

[0110] The j-th layer neural network includes an attention module, and the second output of the attention module in the j-th layer neural network is used to determine the channel estimation result;

[0111] The j-th layer neural network includes an attention module and a feedforward module. The input of the feedforward module in the j-th layer neural network includes the third output of the attention module in the j-th layer neural network, and / or, the fourth output of the feedforward module in the (j-1)-th layer neural network; or,

[0112] The j-th layer neural network includes a feedforward module, and the fourth output of the feedforward module contained in the j-th layer neural network is used to determine the fourth output of the feedforward module contained in the (j+1)-th layer neural network.

[0113] In any possible implementation of any of the first to sixth aspects, any estimation information output by the first model (such as the second estimation information described above) includes: a set of some or all of the channel estimation results from the N channel estimation results obtained by the N-layer neural network included in the first model; and some or all of the channel estimation results from the N channel estimation results obtained by the N-layer neural network included in the first model.

[0114] Based on the above scheme, after the first model performs channel estimation processing on the input received signal, any estimated information output can be realized in the above multiple ways to improve the flexibility of the scheme implementation.

[0115] A seventh aspect of this application provides a communication device, which is a first communication device, comprising a transceiver unit and a processing unit; the transceiver unit is configured to receive first information, the first information being used to indicate channel sparsity; the transceiver unit is also configured to receive a first signal; the processing unit is configured to determine second information; the transceiver unit is also configured to transmit the second information, the second information indicating first estimation information and the first signal; the second information is used for model management of a first model, the first model being used for channel estimation; the first estimation information is obtained by estimating the channel corresponding to the first signal based on the first information.

[0116] In the seventh aspect of this application, the constituent modules of the communication device can also be used to perform the steps executed in various possible implementations of the first aspect and achieve the corresponding technical effects. For details, please refer to the first aspect, which will not be repeated here.

[0117] An eighth aspect of this application provides a communication device, which is a second communication device, comprising a transceiver unit and a processing unit; the processing unit is configured to determine first information and a first signal; the transceiver unit is configured to transmit the first information, which indicates channel sparsity; the transceiver unit is further configured to transmit the first signal; the transceiver unit is further configured to receive second information, which indicates first estimation information and the first signal; the second information is used for model management of a first model, which is used for channel estimation; the first estimation information is obtained by estimating the channel corresponding to the first signal based on the first information.

[0118] In the eighth aspect of this application, the constituent modules of the communication device can also be used to perform the steps executed in various possible implementations of the second aspect and achieve the corresponding technical effects. For details, please refer to the second aspect, which will not be repeated here.

[0119] The ninth aspect of this application provides a communication device, which is a first communication device, comprising a transceiver unit and a processing unit; the transceiver unit is used to receive a second signal; the processing unit is used to determine sixth information; the transceiver unit is also used to transmit the sixth information, which indicates one or more sparsity-corresponding estimation information and the second signal; the sixth information is used for model management of a first model, which is used for channel estimation; the estimation information is obtained by estimating the channel corresponding to the second signal.

[0120] In the ninth aspect of this application, the constituent modules of the communication device can also be used to perform the steps executed in various possible implementations of the third aspect and achieve the corresponding technical effects. For details, please refer to the third aspect, which will not be repeated here.

[0121] The tenth aspect of this application provides a communication device, which is a second communication device, comprising a transceiver unit and a processing unit; the processing unit is used to determine a second signal; the transceiver unit is used to transmit the second signal; the transceiver unit is also used to receive sixth information, the sixth information indicating one or more sparsity-corresponding estimation information and the second signal; the sixth information is used for model management of a first model, the first model being used for channel estimation; the estimation information is obtained by estimating the channel corresponding to the second signal.

[0122] In the tenth aspect of this application, the constituent modules of the communication device can also be used to perform the steps executed in various possible implementations of the fourth aspect and achieve the corresponding technical effects. For details, please refer to the fourth aspect, which will not be repeated here.

[0123] The eleventh aspect of this application provides a communication device, which is a third communication device, comprising a transceiver unit and a processing unit; the transceiver unit is used to receive a third signal; the processing unit is used to process the third signal based on a first model to obtain estimation information corresponding to the third signal; wherein the first model is used for channel estimation.

[0124] In the eleventh aspect of this application, the constituent modules of the communication device can also be used to perform the steps executed in various possible implementations of the fifth aspect and achieve the corresponding technical effects. For details, please refer to the fifth aspect, which will not be repeated here.

[0125] The twelfth aspect of this application provides a communication device, which is a fourth communication device, comprising a transceiver unit and a processing unit; the processing unit is used to determine eighth information; the transceiver unit is used to transmit the eighth information, which is used to indicate a first model; wherein the first model is used to process a signal (e.g., a third signal) to obtain estimation information corresponding to the signal; wherein the first model is used for channel estimation.

[0126] In the twelfth aspect of this application, the constituent modules of the communication device can also be used to perform the steps executed in various possible implementations of the sixth aspect and achieve the corresponding technical effects, all of which can be referred to the sixth aspect, and will not be repeated here.

[0127] The thirteenth aspect of this application provides a communication device including at least one processor for executing computer programs or instructions to enable the communication device to implement the method described in any possible implementation of any of the first to sixth aspects. Optionally, the communication device may include the memory, and / or the at least one processor is coupled to the memory; wherein the memory is used to store programs or instructions.

[0128] The fourteenth aspect of this application provides a communication device including at least one logic circuit and an input / output interface; the logic circuit is used to perform the method as described in any one of the possible implementations of the first to sixth aspects described above.

[0129] The fifteenth aspect of this application provides a communication system, which includes the first communication device and the second communication device described above.

[0130] Alternatively, the communication system may include the aforementioned third and / or fourth communication devices.

[0131] The sixteenth aspect of this application provides a computer-readable storage medium for storing one or more computer-executable instructions, which, when executed by a processor, perform the method as described in any possible implementation of any of the first to sixth aspects described above.

[0132] The seventeenth aspect of this application provides a computer program product (or computer program) that, when executed by a processor, performs the method described in any possible implementation of any of the first to sixth aspects described above.

[0133] The eighteenth aspect of this application provides a chip or chip system including at least one processor for supporting a communication device in implementing the methods described in any possible implementation of any of the first to fourth aspects. For example, the chip may be a baseband chip, a modem chip, a system-on-chip (SoC) chip containing a modem core, a system-in-package (SIP) chip, or a communication module, etc.

[0134] In one possible design, the chip or chip system may further include a memory for storing program instructions and data necessary for the communication device. The chip system may be composed of chips or may include chips and other discrete devices. Optionally, the chip system may also include interface circuitry that provides program instructions and / or data to the at least one processor.

[0135] The technical effects of any of the design methods in aspects seven through eighteen can be found in the technical effects of the different design methods in aspects one through six above, and will not be repeated here. Attached Figure Description

[0136] Figures 1a to 1c are schematic diagrams of the communication system provided in this application;

[0137] Figures 2a to 2g are schematic diagrams of the AI ​​processing involved in this application;

[0138] Figure 3 is an interactive schematic diagram of the communication method provided in this application;

[0139] Figures 4a to 4k are some schematic diagrams of the model provided in this application;

[0140] Figures 5 and 6 are some interactive schematic diagrams of the communication method provided in this application;

[0141] Figures 7 to 11 are schematic diagrams of the communication device provided in this application. Detailed Implementation

[0142] First, some terms used in the embodiments of this application will be explained to facilitate understanding by those skilled in the art.

[0143] (1) Terminal device: can be a wireless terminal device that can receive network device scheduling and instruction information. The wireless terminal device can be a device that provides voice and / or data connectivity to the user, or a handheld device with wireless connection function, or other processing device connected to a wireless modem.

[0144] Terminal devices can communicate with one or more core networks or the Internet via a radio access network (RAN). Terminal devices can be mobile terminal devices, such as mobile phones (or "cellular" phones), computers, and data cards. For example, they can be portable, pocket-sized, handheld, computer-embedded, or vehicle-mounted mobile devices that exchange voice and / or data with the RAN. Examples include personal communication service (PCS) phones, cordless phones, session initiation protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistants (PDAs), tablets, and computers with wireless transceiver capabilities. Wireless terminal equipment can also be referred to as a system, subscriber unit, subscriber station, mobile station (MS), remote station, access point (AP), remote terminal, access terminal, user terminal, user agent, subscriber station (SS), customer premises equipment (CPE), terminal, user equipment (UE), mobile terminal (MT), etc.

[0145] By way of example and not limitation, in this embodiment, the terminal device can also be a wearable device. Wearable devices, also known as wearable smart devices or smart wearable devices, are a general term for devices that utilize wearable technology to intelligently design and develop everyday wearables, such as glasses, gloves, watches, clothing, and shoes. Wearable devices are portable devices that are worn directly on the body or integrated into the user's clothing or accessories. Wearable devices are not merely hardware devices, but also achieve powerful functions through software support, data interaction, and cloud interaction. Broadly speaking, wearable smart devices include those that are feature-rich, large in size, and can achieve complete or partial functions without relying on a smartphone, such as smartwatches or smart glasses, as well as those that focus on a specific type of application function and require the use of other devices such as smartphones, such as various smart bracelets, smart helmets, and smart jewelry for vital sign monitoring.

[0146] Terminals can also be drones, robots, devices in device-to-device (D2D) communication, vehicles to everything (V2X) communication, virtual reality (VR) terminal devices, augmented reality (AR) terminal devices, wireless terminals in industrial control, wireless terminals in self-driving, wireless terminals in telemedicine or telehealth services, wireless terminals in smart grids, wireless terminals in transportation safety, wireless terminals in smart cities, wireless terminals in smart homes, etc.

[0147] Furthermore, terminal devices can also be terminal devices in future communication systems beyond the fifth generation (5G) (such as 5G Advanced communication systems) or in future evolved public land mobile networks (PLMNs). For example, 5G Advanced networks can further expand the form and function of 5G communication terminals; 5G Advanced terminals include, but are not limited to, vehicles, cellular network terminals (integrating satellite terminal functions), drones, and Internet of Things (IoT) devices.

[0148] In this embodiment, the terminal device can also obtain artificial intelligence (AI) services provided by the network device. Optionally, the terminal device can also have AI processing capabilities.

[0149] (2) Network equipment: This can be equipment within a wireless network. For example, network equipment can be a RAN node (or device) that connects terminal devices to the wireless network, and can also be called a base station. Currently, some examples of RAN equipment include: base station, evolved NodeB (eNodeB), gNB (gNodeB) in 5G communication systems, transmission reception point (TRP), evolved Node B (eNB), radio network controller (RNC), Node B (NB), home base station (e.g., home evolved Node B, or home Node B, HNB), base band unit (BBU), or wireless fidelity (Wi-Fi) access point (AP), etc. In addition, in a network architecture, network equipment can include central unit (CU) nodes, distributed unit (DU) nodes, or RAN equipment including both CU and DU nodes.

[0150] Optionally, RAN nodes can also be macro base stations, micro base stations, indoor stations, relay nodes, donor nodes, or radio controllers in cloud radio access network (CRAN) scenarios. RAN nodes can also be servers, wearable devices, vehicles, or in-vehicle equipment. For example, the access network equipment in vehicle-to-everything (V2X) technology can be a roadside unit (RSU).

[0151] In another possible scenario, multiple RAN nodes collaborate to assist the terminal in achieving wireless access, with different RAN nodes each implementing some of the base station's functions. For example, RAN nodes can be CUs, DUs, CUs (control plane, CP), CUs (user plane, UP), or radio units (RUs). CUs and DUs can be configured separately or included in the same network element, such as a baseband unit (BBU). RUs can be included in radio equipment or radio units, such as remote radio units (RRUs), active antenna units (AAUs), radio heads (RHs), or remote radio heads (RRHs).

[0152] In different systems, CU (or CU-CP and CU-UP), DU, or RU may have different names, but those skilled in the art will understand their meaning. For example, in an open access network (open RAN, O-RAN, or ORAN) system, CU can also be called O-CU (open CU), DU can also be called O-DU, CU-CP can also be called O-CU-CP, CU-UP can also be called O-CU-UP, and RU can also be called O-RU. For ease of description, this application uses CU, CU-CP, CU-UP, DU, and RU as examples. Any of the units among CU (or CU-CP, CU-UP), DU, and RU in this application can be implemented through software modules, hardware modules, or a combination of software modules and hardware modules.

[0153] Communication between access network devices and terminal devices follows a specific protocol layer structure. This protocol layer may include a control plane protocol layer and a user plane protocol layer. The control plane protocol layer may include at least one of the following: radio resource control (RRC) layer, packet data convergence protocol (PDCP) layer, radio link control (RLC) layer, media access control (MAC) layer, or physical (PHY) layer, etc. The user plane protocol layer may include at least one of the following: service data adaptation protocol (SDAP) layer, PDCP layer, RLC layer, MAC layer, or physical layer, etc.

[0154] The correspondence between network elements and their achievable protocol layer functions in the ORAN system can be found in Table 1 below.

[0155] Table 1

[0156] Network devices can be other devices that provide wireless communication functions for terminal devices. The embodiments of this application do not limit the specific technology or form of the network device. For ease of description, the embodiments of this application are not limited.

[0157] Network equipment may also include core network equipment, such as the Mobility Management Entity (MME), Home Subscriber Server (HSS), Serving Gateway (S-GW), Policy and Charging Rules Function (PCRF), and Public Data Network Gateway (PDN gateway or P-GW) in 4th generation (4G) networks; and access and mobility management function (AMF), user plane function (UPF), or session management function (SMF) in 5G networks. Furthermore, this core network equipment may also include other core network equipment in 5G networks and next-generation networks of 5G networks.

[0158] In this embodiment of the application, the network device may also have network nodes with AI capabilities, which can provide AI services to terminals or other network devices. For example, it may be an AI node, computing node, RAN node with AI capabilities, or core network element with AI capabilities on the network side (access network or core network).

[0159] In this application embodiment, the device for implementing the function of the network device can be the network device itself, or it can be a device capable of supporting the network device in implementing the function, such as a chip system. This device can be disposed within the network device. In the technical solutions provided in this application embodiment, the example of a network device being used to implement the function of the network device is used to describe the technical solutions provided in this application embodiment.

[0160] (3) Configuration and Pre-configuration: In this application, both configuration and pre-configuration are used. Configuration refers to the network device / server sending configuration information or parameter values ​​to the terminal via messages or signaling, so that the terminal can determine communication parameters or resources for transmission based on these values ​​or information. Pre-configuration is similar to configuration; it can be parameter information or parameter values ​​pre-negotiated between the network device / server and the terminal device, parameter information or parameter values ​​specified by standard protocols for use by the base station / network device or terminal device, or parameter information or parameter values ​​pre-stored in the base station / server or terminal device. This application does not limit this.

[0161] Furthermore, these values ​​and parameters can be changed or updated.

[0162] (4) The terms "system" and "network" in the embodiments of this application can be used interchangeably. "Multiple" refers to two or more. "And / or" describes the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, or B exists alone, where A and B can be singular or plural. The character " / " generally indicates that the related objects before and after are in an "or" relationship. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, "at least one of A, B and C" includes A, B, C, AB, AC, BC or ABC. And, unless otherwise specified, the ordinal numbers such as "first" and "second" mentioned in the embodiments of this application are used to distinguish multiple objects and are not used to limit the order, sequence, priority or importance of multiple objects.

[0163] (5) In the embodiments of this application, "send" and "receive" indicate the direction of signal transmission. For example, "send information to XX" can be understood as the destination of the information being XX, which may include sending directly through the air interface or sending indirectly through the air interface by other units or modules. "Receive information from YY" can be understood as the source of the information being YY, which may include receiving directly from YY through the air interface or receiving indirectly from YY through the air interface by other units or modules. "Send" can also be understood as the "output" of the chip interface, and "receive" can also be understood as the "input" of the chip interface.

[0164] In other words, sending and receiving can occur between devices, such as between network devices and terminal devices, or within a device, such as between components, modules, chips, software modules, or hardware modules within the device via buses, wiring, or interfaces.

[0165] It is understandable that information may undergo necessary processing, such as encoding and modulation, between the source and destination, but the destination can understand the valid information from the source. Similar statements in this application can be interpreted in a similar way and will not be elaborated further.

[0166] (6) In the embodiments of this application, "instruction" may include direct instruction and indirect instruction, as well as explicit instruction and implicit instruction. The information indicated by a certain piece of information (as described below, the instruction information) is called the information to be instructed. In the specific implementation process, there are many ways to indicate the information to be instructed, such as, but not limited to, directly indicating the information to be instructed, such as the information to be instructed itself or its index. It can also indirectly indicate the information to be instructed by indicating other information, where there is an association between the other information and the information to be instructed; or it can only indicate a part of the information to be instructed, while the other parts of the information to be instructed are known or pre-agreed upon. For example, the instruction can be implemented by using a pre-agreed (e.g., protocol predefined) arrangement order of various information, thereby reducing the instruction overhead to a certain extent. This application does not limit the specific method of instruction. It is understood that for the sender of the instruction information, the instruction information can be used to indicate the information to be instructed, and for the receiver of the instruction information, the instruction information can be used to determine the information to be instructed.

[0167] In this application, unless otherwise specified, the same or similar parts between the various embodiments can be referred to each other. In the various embodiments of this application, and the various methods / designs / implementations within each embodiment, unless otherwise specified or logically conflicting, the terminology and / or descriptions between different embodiments and between the various methods / designs / implementations within each embodiment are consistent and can be mutually referenced. The technical features in different embodiments and the various methods / designs / implementations within each embodiment can be combined to form new embodiments, methods, or implementations based on their inherent logical relationships. The following descriptions of the embodiments of this application do not constitute a limitation on the scope of protection of this application.

[0168] This application can be applied to long-term evolution (LTE) systems, new radio (NR) systems, or future communication systems beyond 5G. These communication systems include at least one network device and / or at least one terminal device.

[0169] Please refer to Figure 1a, which is a schematic diagram of the architecture of the communication system 1000 used in the embodiments of this application. As shown in Figure 1a, the communication system may include a radio access network (RAN) 100. Optionally, the communication system 1000 may also include a core network 200 and an Internet 300. The RAN 100 includes at least one RAN node (110a and 110b in Figure 1a, collectively referred to as 110) and at least one terminal (120a-120j in Figure 1a, collectively referred to as 120). The RAN 100 may also include other RAN nodes, such as wireless relay devices and / or wireless backhaul devices (not shown in Figure 1a). The terminal 120 is wirelessly connected to the RAN node 110, and the RAN node 110 is wirelessly or wiredly connected to the core network 200. The core network equipment in the core network 200 and the RAN node 110 in the RAN 100 may be independent and different physical devices, or they may be the same physical device integrating the logical functions of the core network equipment and the logical functions of the RAN node. Terminals can be connected to each other, as can RAN nodes, via wired or wireless means.

[0170] Taking the communication system shown in Figure 1a as an example, in addition to performing communication-related services, different devices (including network devices and network devices, network devices and terminal devices, and / or terminal devices and terminal devices) may also perform AI-related services.

[0171] As shown in Figure 1b, taking a network device as a base station as an example, the base station can perform communication-related services and AI-related services with one or more terminal devices, and different terminal devices can also perform communication-related services and AI-related services.

[0172] As shown in Figure 1c, taking terminal devices including televisions and mobile phones as an example, communication-related services and AI-related services can also be performed between televisions and mobile phones.

[0173] The technical solutions provided in this application can be applied to wireless communication systems (such as the systems shown in Figures 1a, 1b, or 1c). For example, AI network elements can be introduced into the communication system provided in this application to realize some or all AI-related operations. AI network elements can also be called AI nodes, AI devices, AI entities, AI modules, AI models, or AI units, etc. The AI ​​network element can be built into a network element within the communication system. For example, the AI ​​network element can be an AI module built into: access network equipment, core network equipment, cloud server, or operation, administration, and maintenance (OAM) to realize AI-related functions. The OAM can act as the network management system for the core network equipment and / or the access network equipment. Alternatively, the AI ​​network element can also be an independently set network element in the communication system. Optionally, the terminal or its built-in chip can also include an AI entity to realize AI-related functions.

[0174] Optionally, in communication systems, AI application cases may include, but are not limited to: channel state information (CSI) feedback enhancement, beam management enhancement, positioning accuracy enhancement, network energy saving, load balancing, and mobility optimization. These will be explained below.

[0175] 1. Enhanced CSI feedback

[0176] Channel quality information (CSI) is the channel attribute of a communication link, reported by the terminal device to the network device. By reporting this information, the terminal device can select an appropriate modulation and coding scheme (MCS) to adapt to changing wireless channels. For example, the terminal device might perform channel estimation based on the received channel state information-reference signal (CSI-RS) and then feed back the CSI-RS to the network device. This information serves as input to the network device's model, enabling AI model training. Applying AI to CSI feedback enhancement can reduce overhead, improve accuracy, and enhance predictive capabilities.

[0177] CSI-RS feedback enhancement may include at least one sub-function, such as: CSI compression, CSI prediction, and CSI-RS configuration signaling reduction. CSI compression may further include CSI compression in at least one domain: spatial, time, and frequency.

[0178] 2. Enhanced Beam Management

[0179] Enhanced beam management primarily aims to discover the strongest transmit / receive beam pairs. AI-based sparse beam prediction can improve accuracy. This can be achieved through both network-side and terminal-side AI sparse beam prediction, based on AI training and inference. Taking terminal-side AI sparse beam prediction as an example, the pre-trained AI model on the terminal device can be provided by the network or pre-stored on the terminal device. During training, the network device scans all possible beams and then reports the transmit beam pattern to the terminal device. Once training is complete, the network device only needs to scan a small subset of beams, and the terminal device then feeds back the inference results. AI-based beam management can achieve beam prediction in, for example, the temporal and / or spatial domains, reducing overhead and latency and improving beam selection accuracy.

[0180] Beam management enhancements may include at least one sub-function, such as: beam scan matrix prediction and optimal beam prediction.

[0181] 3. Enhanced positioning accuracy

[0182] 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: positioning enhancement based on access network devices, positioning enhancement based on positioning management function network elements, and positioning enhancement based on terminal devices.

[0183] 4. Network energy saving

[0184] Network energy conservation can be achieved through cell activation / deactivation, load reduction, coverage improvement, or other RAN setting adjustments. AI technology can be used to optimize energy-saving decisions by leveraging data collected within the RAN network. AI algorithms can predict energy efficiency and load status for the next cycle, which can be used to assist in cell activation / deactivation decisions 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.

[0185] 5. Load balancing

[0186] Load balancing can distribute the load evenly between cells and across different areas within a cell, or transfer some traffic from congested cells, or offload users across a single cell, carrier, or access standard, thereby improving network performance. Using AI models to enhance load balancing performance—such as inputting various measurements and feedback from terminal devices and network nodes, as well as historical data—can provide a higher quality user experience and increase system capacity.

[0187] 6. Mobility Management

[0188] Mobility management is a solution that ensures service continuity for mobile devices by minimizing dropped calls, radio link failures (RLFs), unnecessary handovers, and ping-pong effects. AI can enhance mobility management by, for example, reducing the probability of unexpected events, predicting device location / mobility / performance, and routing traffic.

[0189] It should be understood that the definitions of the above technical terms are merely illustrative. For example, as technology continues to develop, the scope of the above definitions may also change, and the embodiments of this application are not intended to limit the scope.

[0190] For example, an AI function may include multiple AI sub-functions.

[0191] Optionally, AI application cases are also called AI application scenarios or AI functions.

[0192] As described above regarding AI application examples, AI can be widely used to improve network performance in areas such as CSI feedback enhancement, beam management, positioning accuracy enhancement, energy saving, mobility enhancement, and load balancing. AI models can typically be deployed on the network side and / or the terminal device side. The training of AI models relies on the collection of training data, which can come from measurements and feedback from the terminal devices.

[0193] The following is a brief introduction to the concepts that may be involved in this application.

[0194] AI can endow machines with human-like intelligence, for example, allowing them to use computer hardware and software to simulate certain intelligent human behaviors. To achieve artificial intelligence, machine learning methods can be employed. In machine learning, machines learn (or train) a model using training data. This model represents the mapping between inputs and outputs. The learned model can be used for reasoning (or prediction), that is, it can be used to predict the output corresponding to a given input. This output can also be called the reasoning result (or prediction result).

[0195] Machine learning can include supervised learning, unsupervised learning, and reinforcement learning. Unsupervised learning can also be called learning without supervision.

[0196] Supervised learning, based on collected sample values ​​and labels, uses machine learning algorithms to learn the mapping relationship between sample values ​​and labels, and then expresses this learned mapping relationship using an AI model. The process of training the machine learning model is the process of learning this mapping relationship. During training, sample values ​​are input into the model to obtain the model's predicted values, and the model parameters are optimized by calculating the error between the model's predicted values ​​and the sample labels (ideal values). After the mapping relationship is learned, it can be used to predict new sample labels. The mapping relationship learned in supervised learning can include linear or non-linear mappings. Based on the type of label, the learning task can be divided into classification tasks and regression tasks.

[0197] Unsupervised learning relies on collected sample values ​​to discover inherent patterns within the samples themselves. One type of unsupervised learning algorithm uses the samples themselves as supervisory signals, meaning the model learns the mapping relationship from sample to sample; this is called self-supervised learning. During training, model parameters are optimized by calculating the error between the model's predictions and the samples themselves. Self-supervised learning can be used for signal compression and decompression recovery applications; common algorithms include autoencoders and generative adversarial networks.

[0198] Reinforcement learning, unlike supervised learning, is a type of algorithm that learns problem-solving strategies through interaction with the environment. Unlike supervised and unsupervised learning, reinforcement learning problems do not have explicit "correct" action labels. The algorithm needs to interact with the environment to obtain reward signals from the environment, and then adjust its decision actions to obtain a larger reward signal value. For example, in downlink power control, the reinforcement learning model adjusts the downlink transmission power of each user based on the total system throughput feedback from the wireless network, aiming to achieve a higher system throughput. The goal of reinforcement learning is also to learn the mapping relationship between the environment state and a better (e.g., optimal) decision action. However, because the label of the "correct action" cannot be obtained in advance, the network cannot be optimized by calculating the error between the action and the "correct action." Reinforcement learning training is achieved through iterative interaction with the environment.

[0199] Neural networks (NNs) are a specific model in machine learning techniques. According to the general approximation theorem, neural networks can theoretically approximate any continuous function, thus enabling them to learn arbitrary mappings. Traditional communication systems rely on extensive expert knowledge to design communication modules, while deep learning communication systems based on neural networks can automatically discover hidden pattern structures from large datasets, establish mapping relationships between data, and achieve performance superior to traditional modeling methods.

[0200] The idea behind neural networks comes from the neuronal structure of the brain. For example, each neuron performs a weighted summation of its input values ​​and outputs the result through an activation function.

[0201] Figure 2a shows a schematic diagram of a neuron structure. Assume the input to the neuron is x = [x0, x1, ..., x...]. n The weights corresponding to each input are w = [w0, w1, ..., w] n ], where n is a positive integer, w i and x i It can be any possible type, such as a decimal, an integer (e.g., 0, a positive integer, or a negative integer), or a complex number. i As x i The weights are used to assign weights to x. i Weighting is applied. The bias for the weighted sum of the input values ​​is, for example, b. Activation functions can take many forms. Suppose the activation function of a neuron is: y = f(z) = max(0, z), then the output of that neuron is: If the activation function of a neuron is y = f(z) = z, then the output of the neuron is: Here, b can be any possible type, such as a decimal, an integer (e.g., 0, a positive integer, or a negative integer), or a complex number. The activation functions of different neurons in a neural network can be the same or different.

[0202] Furthermore, neural networks generally consist of multiple layers, each of which may include one or more neurons. Increasing the depth and / or width of a neural network can improve its expressive power, providing more powerful information extraction and abstract modeling capabilities for complex systems. The depth of a neural network can refer to the number of layers it includes, and the number of neurons in each layer can be called the width of that layer. In one implementation, a neural network includes an input layer and an output layer. The input layer processes the received input information through neurons and passes the processing result to the output layer, which then obtains the output of the neural network. In another implementation, a neural network includes an input layer, hidden layers, and an output layer. The input layer processes the received input information through neurons and passes the processing result to the hidden layer. The hidden layer calculates the received processing result and passes the calculation result to the output layer or the next adjacent hidden layer, ultimately obtaining the output of the neural network. A neural network may include one hidden layer or multiple sequentially connected hidden layers, without limitation.

[0203] Neural networks, for example, are deep neural networks (DNNs). Depending on how the network is constructed, DNNs can include feedforward neural networks (FNNs), convolutional neural networks (CNNs), and recurrent neural networks (RNNs).

[0204] Figure 2b is a schematic diagram of an FNN network. A characteristic of FNN networks is that neurons in adjacent layers are completely connected pairwise. This characteristic makes FNNs typically require a large amount of storage space, leading to high computational complexity.

[0205] CNNs are neural networks specifically designed to process data with a grid-like structure. For example, time-series data (e.g., discrete sampling along a time axis) and image data (e.g., two-dimensional discrete sampling) can both be considered grid-like data. CNNs do not use all the input information at once for computation; instead, they use a fixed-size window to extract a portion of the information for convolution operations, which significantly reduces the computational cost of model parameters. Furthermore, depending on the type of information extracted by the window (e.g., people and objects in an image represent different types of information), each window can use different convolution kernels, allowing CNNs to better extract features from the input data.

[0206] Recurrent Neural Networks (RNNs) are a type of neural network that utilizes feedback time-series information. The input to an RNN includes the current input value and its own output value from the previous time step. RNNs are suitable for acquiring temporally correlated sequence features, and are applicable to applications such as speech recognition and channel coding / decoding.

[0207] In the model training process described above, a loss function can be defined. The loss function describes the difference between the model's output value and the ideal target value. The loss function can be expressed in various forms, and there are no restrictions on its specific form. The model training process can be viewed as follows: by adjusting some or all of the model's parameters, the value of the loss function is made to be less than a threshold or to meet the target requirement.

[0208] A model can also be called an AI model, a rule, or other names. An AI model can be considered a specific method for implementing AI functions. An AI model represents the mapping relationship or function between the model's input and output. AI functions can include one or more of the following: data collection, model training (or model learning), model information dissemination, model inference (or model reasoning, inference, or prediction, etc.), model monitoring or model validation, or inference result publication, etc. AI functions can also be called AI (related) operations or AI-related functions.

[0209] The implementation process of the neural network will be described below with reference to the accompanying drawings.

[0210] 1. Fully connected neural network, also known as multilayer perceptron (MLP).

[0211] As shown in Figure 2c, an MLP consists of an input layer (left side), an output layer (right side), and multiple hidden layers (middle). Each layer of an MLP contains several nodes, called neurons. Neurons in adjacent layers are connected pairwise.

[0212] Optionally, considering neurons in two adjacent layers, the output h of the next layer's neurons is the weighted sum of all neurons x in the previous layer connected to it, processed by an activation function, and can be expressed as: h = f(wx + b).

[0213] Where w is the weight matrix, b is the bias vector, and f is the activation function.

[0214] Alternatively, the output of the neural network can be recursively expressed as: y = f z (w z f z-1 (…)+b z ).

[0215] Where z is the index of the neural network layer, z is greater than or equal to 1 and z is less than or equal to Z, where Z is the total number of layers in the neural network.

[0216] In other words, a neural network can be understood as a mapping from an input data set to an output data set. Neural networks are typically initialized randomly; the process of obtaining this mapping from random values ​​w and b using existing data is called training the neural network.

[0217] Optionally, the training method involves using a loss function to evaluate the output of the neural network.

[0218] As shown in Figure 2d, the error can be backpropagated, and the neural network parameters (including w and b) can be iteratively optimized using gradient descent until the output of the loss function reaches its minimum value, which is the "better point (e.g., the optimal point)" in Figure 2d. It can be understood that the neural network parameters corresponding to the "better point (e.g., the optimal point)" in Figure 2d can be used as the neural network parameters in the trained AI model information.

[0219] Alternatively, the gradient descent process can be represented as:

[0220] Where θ represents the parameters to be optimized (including w and b), L is the loss function, and η is the learning rate, controlling the step size of gradient descent. This represents the differentiation operation. This indicates taking the derivative of θ with respect to L.

[0221] Alternatively, the backpropagation process can utilize the chain rule for partial derivatives.

[0222] As shown in Figure 2e, the gradient of the parameters in the previous layer can be recursively calculated from the gradient of the parameters in the next layer, and can be expressed as:

[0223] Among them, w ij Let s be the weight of the connection between node j and node i. i The weighted sum of the inputs at node i.

[0224] 2. Federated Learning (FL).

[0225] The concept of federated learning effectively addresses the current challenges in the development of artificial intelligence. While fully protecting user data privacy and security, it enables various edge devices and central servers to collaborate efficiently to complete the model's learning task.

[0226] As shown in Figure 2f, the FL architecture is the most widely used training architecture in the current FL field, and the FedAvg algorithm is the basic algorithm of FL. The FedAvg algorithm flow is roughly as follows:

[0227] (1) Initialize the model to be trained at the center end. And broadcast it to all clients.

[0228] (2) In the t∈[1,T] round, the client k∈[1,K] is based on the local dataset. For the received global model Perform E epochs of training to obtain the local training results. This is then reported to the central node. In the example shown in Figure 2f, the local training results sent by distributed nodes n, k, and m are denoted as G, respectively. n G k G m .

[0229] (3) The central node collects local training results from all (or some) clients. Assume the set of clients uploading local models in round t is... The central server will use the number of samples from the corresponding client as weights to calculate the new global model. The specific update rule is as follows: Then the central end will send the latest version of the global model. The broadcast is sent to all clients for a new round of training.

[0230] (4) Repeat steps (2) and (3) until the model finally converges or the number of training rounds reaches the upper limit.

[0231] Optionally, in addition to reporting the local model, the client can also... It can also train local gradients The central node averages the local gradients reported by all clients and updates the global model based on this average gradient.

[0232] As can be seen in the FL framework, the dataset resides on distributed nodes (such as clients). These distributed nodes collect their local datasets, perform local training, and report the local results (model or gradients) to the central node. The central node itself may not have a dataset; it can be responsible for fusing the training results from the distributed nodes to obtain a global model, which is then distributed back to the distributed nodes.

[0233] 3. Decentralized learning.

[0234] Figure 2g illustrates a fully distributed system without a central node. The design goal f(x) of a decentralized learning system is generally the goal f of each node. i The mean of (x), i.e. Where n is the number of distributed nodes, and x is the parameter to be optimized; in machine learning, x is the parameter of the machine learning model (such as a neural network). Each node utilizes local data and its local target f. i (x) Calculate the local gradient Then it is sent to its communicatively reachable neighboring nodes. Upon receiving the gradient information from its neighbor, any node can update the parameters x of its local model according to the following formula:

[0235] in, This represents the parameters of the local model after the (k+1)th update (k is a natural number) in the i-th node. This represents the parameters of the local model for the i-th node after the k-th update (if k is 0, then it represents...). (where α is the parameter of the local model of the i-th node that is not involved in the update) k N represents the tuning coefficient. i It is the set of neighboring nodes of node i, |N i | represents the number of elements in the set of neighboring nodes of node i, that is, the number of neighboring nodes of node i. Through information interaction between nodes, the decentralized learning system will eventually learn a unified model.

[0236] The technical solution provided in this application can be applied to communication systems (such as the systems shown in Figure 1a, 1b, or 1c), in which different communication devices can transmit communication signals. During the transmission of communication signals, sparse signal estimation instructs the signal receiver to recover or estimate a sparse signal from a limited set of observations. A sparse signal can be represented as a linear combination of a few non-zero or significant elements, while the remainder is zero or close to zero. Through the sparse signal estimation process, the number of observations or measurements required by the signal receiver can be reduced, thereby reducing energy consumption and improving efficiency.

[0237] However, how to improve the processing efficiency of sparse signal estimation is a technical problem that urgently needs to be solved.

[0238] To address the aforementioned problems, this application provides a communication method and related apparatus, which will be described in detail below with reference to the accompanying drawings.

[0239] Please refer to Figure 3, which is a schematic diagram of an implementation of the communication method provided in this application. The method includes the following steps.

[0240] It should be noted that in the following text, Figures 3, 5, and 6 illustrate the method using the first communication device and other communication devices (such as the second communication device) as examples of the execution subjects of this interaction illustration, but this application does not limit the execution subjects of this interaction illustration. For example, the communication device can be a communication equipment, or a chip, baseband chip, modem chip, system-on-chip (SoC) chip containing a modem core, system-in-package (SIP) chip, communication module, chip system, processor, logic module, or software in the communication equipment. Optionally, the communication equipment can be a terminal device or network device (e.g., access network equipment, access network element, core network element, or core network equipment, etc.).

[0241] S301. The second communication device sends first information, and correspondingly, the first communication device receives the first information. The first information is used to indicate channel sparsity.

[0242] S302. The second communication device sends a first signal, and correspondingly, the first communication device receives the first signal.

[0243] It should be noted that the execution order of steps S301 and S302 is not limited here. For example, the first communication device and the second communication device may execute step S301 first and then step S302, or the first communication device and the second communication device may execute step S302 first and then step S301.

[0244] S303. The first communication device sends second information, and correspondingly, the first communication device receives the second information. The second information indicates first estimation information and the first signal; the second information is used for model management of the first model, and the first model is used for channel estimation; the first estimation information is obtained by estimating the channel corresponding to the first signal based on the first information.

[0245] It should be understood that after a signal transmitter sends a signal, the signal is transmitted through a wireless channel, and then the signal receiver can receive the signal. During this process, the wireless channel will inevitably affect the signal sent by the signal transmitter (e.g., multipath propagation, interference, attenuation, etc.), which may result in the signal received by the signal receiver being different from the signal sent by the signal transmitter. Accordingly, the first signal received by the first communication device in step S302 may be different from the first signal sent by the signal transmitter (e.g., the second communication device) in step S302.

[0246] For example, the signal Y received by the signal receiver satisfies: Y = AX + n;

[0247] Where A represents the transmission channel information between the signal sender and the signal receiver, X represents the signal sent by the signal sender, and n represents noise.

[0248] Therefore, the first signal indicated by the second information sent by the first communication device in step S303 is specifically the first signal received (or parsed, or detected) by the first communication device in step S302, such as the "Y" mentioned above.

[0249] In this application, the model (e.g., the first model, the second model described below, etc.) may include a mathematical model, an AI model, a neural network model, an AI neural network model, a machine learning model, or an AI processing model, etc.

[0250] For example, a model (e.g., a first model, a second model described later, etc.) used for channel estimation can be understood as follows: the input of the model may include the signal received by the signal receiver and transmitted through the wireless channel, and the output of the model may include an estimation result of the wireless channel (e.g., second estimation information, etc.). For example, the estimation result may include one or more of the following: the estimated non-zero element index, the estimated non-zero element value, OR, and residual error.

[0251] It should be understood that residual error can indicate the error, difference, or deviation between the signal actually received by the signal receiver and the signal constructed by the signal receiver based on the estimated information obtained from channel estimation.

[0252] In this application, model management may include one or more of the following: model monitoring, model training, model inference, model scheduling, model deployment, model updating, model generation, model switching, or function rollback.

[0253] As an example, when the first communication device deploys the first model, after the first communication device sends the second information in step S303, the recipient of the second information (e.g., the second communication device) can perform model management on the first model deployed on the first communication device based on the second information. The model management may include one or more of the following: model monitoring, model scheduling, model switching, function rollback, or model generation.

[0254] For example, the receiver can determine the performance of the first model based on the second information, thereby enabling model monitoring (e.g., monitoring of model performance) of the first model deployed on the first communication device.

[0255] For example, the receiver can schedule whether the first communication device performs channel estimation processing for one or more other signals based on the second information, thereby achieving model scheduling of the first model deployed on the first communication device.

[0256] For example, if the first communication device deploys two or more models for channel estimation, the receiver can determine the performance of the first model based on the second information, and determine whether to instruct the first communication device to perform channel estimation processing on the channel corresponding to the signal based on other models based on the performance of the first model, thereby realizing model switching of the first model deployed on the first communication device.

[0257] As another example, when the first communication device deploys the second model, after the first communication device sends the second information in step S303, the recipient of the second information (e.g., the second communication device) can perform model management on the first model deployed on the second communication device based on the second information. The model management may include one or more of the following: model monitoring, model scheduling, model switching, function rollback, or model generation.

[0258] For example, the receiver can determine the performance of the first model based on the second information, thereby enabling model monitoring (e.g., monitoring of model performance) of the first model deployed on the second communication device.

[0259] For example, the receiver can determine the performance of the first model based on the second information, and determine whether to perform channel estimation processing on the received signal based on the model, thereby realizing model management of whether to perform function rollback on the first model deployed on the second communication device.

[0260] For example, the receiver can perform model processing (such as fine-tuning, adjustment, and updating) on ​​the first model based on the second information to update the first model deployed on the second communication device.

[0261] It should be understood that when a model is deployed on a communication device (e.g., a first model is deployed on a first communication device, or a first model is deployed on a second communication device, etc.), it can be understood that after the communication device obtains the model parameters of the model, it obtains / generates / constructs the model based on the model parameters. Subsequently, the communication device can perform data processing based on the model. Optionally, the model parameters may include one or more of the following: model hyperparameters, model dataset (including model input data and corresponding label data, etc.), model structure, model weights, and model parameters.

[0262] In this application, channel sparsity can indicate the number of elements to be estimated (or the number of non-zero elements) in the channel information, or the ratio of the number of elements to be estimated (or the number of non-zero elements) in the channel information to the dimension of the channel information.

[0263] Optionally, in step S301, the first information received by the first communication device, in addition to indicating channel sparsity, also indicates at least one of the following: residual error, capability configuration of the first communication device (e.g., the capability configuration indicates a certain capability of the first communication device, including one or more of supported bandwidth, number of antennas, channel estimation algorithm, channel estimation quality or performance, enabling the first communication device to subsequently process the first signal and obtain the first estimation information based on the configuration indicated by the capability configuration), and configuration information of the reference signal. In this way, the first communication device can perform channel estimation based on parameters specified by other communication devices to obtain the first estimation information, and send second information containing the first estimation information, enabling the other communication devices to perform model management on the channel estimation model based on the first estimation information corresponding to the specified parameters, thereby improving model management efficiency.

[0264] Optionally, before receiving the first information, the first communication device may send one or more capability information supported by the first communication device (e.g., at least one of one or more residual errors, one or more capability information, and configuration information of one or more reference signals) so that the receiver (e.g., the second communication device) can determine the capability configuration indicated by the first information based on the received one or more capability information.

[0265] Based on the scheme shown in Figure 3, after receiving the first signal in step S302, the first communication device can estimate the channel corresponding to the first signal based on the first information indicating channel sparsity, thus obtaining first estimation information. Subsequently, in step S303, the first communication device can send second information indicating the first estimation information and the first signal, allowing the receiver of the second information to use it for model management of the first model. The first model is used for channel estimation; that is, the first model can be a channel estimation model. In other words, the second information sent by the first communication device can be used for model management of the channel estimation model. In this way, communication devices deploying the channel estimation model can perform channel estimation on received signals based on the channel estimation model, thereby improving the processing efficiency of sparse signal estimation.

[0266] Furthermore, the first estimation information indicated by the second information sent by the first communication device in step S303 is determined based on the first information indicating channel sparsity. This second information can also be used for model management of the channel estimation model. In this way, the first communication device can perform channel estimation based on the channel sparsity specified by other communication devices to obtain the first estimation information, and send information containing this first estimation information. This allows other communication devices (e.g., the receiver of the second information) to perform model management of the channel estimation model based on the first estimation information corresponding to the specified channel sparsity, thereby improving model management efficiency.

[0267] In one possible implementation of the method shown in Figure 3, before step S301, the method further includes: the first communication device sending third information, which is used to determine the first information in step S301; wherein the third information indicates one or more estimation information of the reference signal, and each estimation information indicates the residual error corresponding to one of the one or more channel sparsities. In other words, the first communication device may also send third information for determining the first information, wherein the third information indicates one or more estimation information of the reference signal, and different estimation information can indicate the residual error corresponding to different channel sparsities. In this way, the receiver of the third information (e.g., the second communication device) can determine the residual error corresponding to one or more channel sparsities through the third information, so that the receiver can instruct the first communication device to obtain estimation information based on one of the channel sparsities through the first information in step S301, and then the receiver can obtain the first estimation information corresponding to the specified channel sparsity through the second information, and perform model management on the channel estimation model based on the first estimation information corresponding to the specified channel sparsity, so as to improve the model management efficiency.

[0268] It should be understood that one or more estimation information of the reference signal can be understood as one or more estimation information obtained by the first communication device after receiving the reference signal and performing one or more channel estimations based on the reference signal.

[0269] Optionally, the model parameters of the first model (e.g., the number of neural network levels and layers included in the first model) can be determined based on the third information. In other words, the third information sent by the first communication device can be used to determine the model parameters of the first model, that is, the recipient of the third information (e.g., the second communication device) can use the third information for the management of the first model (e.g., model update, model generation, etc.), so that the recipient obtains a first model that is compatible with the estimation information obtained by the first communication device from channel estimation of the reference signal.

[0270] Optionally, the model parameters of the first model (e.g., the number of neural network levels and layers included in the first model) are determined based on the first information. In other words, the first information received by the first communication device can be used to determine the model parameters of the first model, that is, the first communication device can use the first information for the management of the first model (e.g., model update, model generation, etc.), so that the first communication device obtains a first model that is compatible with the channel coefficient indicated by the first information.

[0271] In one possible implementation of the method shown in Figure 3, after step S303, the method further includes: the first communication device receiving fourth information, which is used to indicate a second model, the second model being obtained by training the first model. In other words, the receiver of the first information (e.g., the second communication device) can train the first model based on the first information to obtain the second model, and deploy the second model to the first communication device through the fourth information, thereby realizing model training and model deployment.

[0272] In one possible implementation of the method shown in Figure 3, the method further includes: the first communication device sending fifth information indicating the difference between the first estimation information and the second estimation information, wherein the second estimation information is obtained based on the first model. In other words, the first communication device can obtain the second estimation information by estimating the received first signal using the first model, and the first communication device can also obtain the first estimation information by estimating the received first signal using other methods. Subsequently, the fifth information sent by the first communication device is used to indicate the difference between the first and second estimation information, enabling the recipient of the fifth information to obtain the difference based on the fifth information and to perform model management on the first model based on the difference.

[0273] Optionally, when the first model is deployed on the first communication device, the first communication device can obtain the second estimation information based on the first model deployed locally and send the fifth information, so that the second communication device can perform model management on the first model based on the differences indicated by the fifth information.

[0274] Optionally, if the first model is deployed on another communication device (e.g., the second communication device), the second communication device can obtain the second estimation information by receiving information from the other communication device (or by obtaining the first model deployed by itself), so that the second communication device can perform model management on the first model based on the first estimation information and the second estimation information (e.g., the difference between the two), that is, the first communication device may not send the fifth information.

[0275] Optionally, the above differences can be characterized by mathematical calculations of different outputs. For example, the mathematical calculation results may include one or more of the following: difference, mean square error (MSE), normalized mean square error (NMSE), or cosine similarity.

[0276] Optionally, the fifth information and the second information in step S303 above can be carried in the same message / signaling / information to save overhead.

[0277] It should be understood that the first communication device can estimate the received first signal in a way different from the first model to obtain the first estimation information, which will be described below with reference to some implementation examples.

[0278] As an example of implementation, this other approach can be a traditional channel estimation method, including but not limited to channel estimation methods based on orthogonal matching pursuit (OMP) algorithms, Bayesian learning-based algorithms, etc. In this way, the first estimation information can serve as the target value (or expected value, or ground truth, etc.) corresponding to the second estimation information, enabling the receiver of the fifth information to manage the first model based on the difference between the first estimation information indicated by the fifth information and the target value.

[0279] As another implementation example, this other approach can be a processing method for models other than the first model.

[0280] For example, if the model capability of the other model is superior to that of the first model, the model inference accuracy and / or the communication performance (e.g., system performance, link performance) corresponding to the model output of the former may be superior to that of the latter. Accordingly, the first estimated information obtained from the output of the other model can be used as the target value (or expected value, or benchmark truth, etc.) corresponding to the second estimated information, so that the receiver of the fifth information can manage the first model based on the difference between the first estimated information indicated by the fifth information and the target value.

[0281] For example, if the model capability of the other model is inferior to that of the first model, the model inference accuracy and / or the communication performance (e.g., system performance, link performance) corresponding to the model output of the former may be inferior to that of the latter. Accordingly, the second estimated information obtained by the output of the first model can be used as the target value (or expected value, or benchmark truth, etc.) corresponding to the first estimated information obtained by the output of the other model, so that the receiver of the fifth information can realize model management of the other model based on the difference between the first estimated information indicated by the fifth information and the target value.

[0282] In one possible implementation, the first model comprises an N-layer neural network, where the input to the (i+1)th layer is determined based on the output of the ith layer, where N is a positive integer and i ranges from 1 to N-1 (or, the input to the ith layer is determined based on the output of the (i-1)th layer, where N is a positive integer and i ranges from 2 to N). Specifically, the first model may contain N layers of neural networks, where adjacent layers can satisfy the aforementioned correlation, enabling the next layer to acquire the feature information accumulated by the previous layer, thereby improving the processing performance of the next layer.

[0283] As an example, as shown in Figure 4a, taking an N value greater than 2 as an example, the first model can include the first layer of the neural network, the second layer of the neural network, ..., the Nth layer of the neural network in Figure 4a. Furthermore, the input to the next layer of the neural network can be obtained based on part or all of the output of the previous layer, allowing the neural network with a larger layer index to obtain the feature information accumulated by the neural network with a smaller layer index, thereby improving the overall processing performance of the neural network.

[0284] Optionally, any estimation information output by the first model (such as the second estimation information described above) includes: a set of some or all channel estimation results from the N channel estimation results obtained by the N-layer neural network included in the first model; or some or all channel estimation results from the N channel estimation results obtained by the N-layer neural network included in the first model. Specifically, after the first model performs channel estimation processing on the input received signal, any estimation information output can be implemented in the above-mentioned multiple ways to improve the flexibility of the scheme implementation.

[0285] In one possible implementation, the first model includes a P-level neural network, where one level of the P-level neural network comprises the N-layer neural network, and different levels of the neural network contain different neural networks, where P is a positive integer. Specifically, the first model may include a P-level neural network to achieve sparse signal estimation through joint processing of one or more levels of neural networks.

[0286] As an example, as shown in Figure 4b, taking a value of P greater than 2 as an example, the first model may include the first-level neural network, ..., the p-th (where p is an integer greater than 1)-th level neural network, ..., the P-th level neural network shown in Figure 4b. Furthermore, one of the first-level neural networks in the P-th level neural network contains N layers of neural networks (for example, the p-th level neural network contains the N layers of neural networks shown in Figure 4a).

[0287] Optionally, in a P-level neural network, the number of neural network layers contained in different levels of neural networks is N.

[0288] As an example, as shown in Figure 4c, taking a value of P greater than 2, the first model can include the first-level neural network, ..., the p-th (p is a positive integer)-level neural network, ..., the P-th-level neural network shown in Figure 4b. Furthermore, each level of the P-level neural network contains N neural network layers.

[0289] The first-level neural network contains N layers, denoted as 1-1, 1-2, ..., 1-N.

[0290] The p-th level neural network (where p is a positive integer greater than 1) contains N layers, denoted as p-1, p-2, ..., pN.

[0291] The P-th level neural network contains N layers, denoted as P-1, P-2, ..., PN.

[0292] It should be noted that, in the case where the first model includes a P-level neural network, each layer of the neural network, which contains one or more layers of neural networks, can be configured (or pre-configured, set, etc.) with corresponding channel estimation parameters to perform estimation processing for that layer of neural network.

[0293] Optionally, for any layer of neural network, the channel estimation parameters corresponding to that layer can be regarded as part of the model parameters of that layer of neural network, or as parameters input by that neural network through a certain input interface (i.e., the model parameters of that layer of neural network do not include the channel estimation parameters corresponding to that layer of neural network), without any limitation here.

[0294] It should be understood that after a signal transmitter sends a signal, the signal is transmitted through a wireless channel, and then the signal receiver can receive the signal. During this process, the wireless channel will inevitably affect the signal transmitted by the transmitter (e.g., interference, attenuation), which may cause the signal received by the receiver to differ from the signal transmitted by the transmitter. In the above scheme, channel estimation parameters are used to represent the correlation between the signal transmitted by the transmitter and the signal received by the receiver. For example, channel estimation parameters may include one or more of the following: precoding information, spatial codebook, frequency codebook, time codebook, oversampled discrete Fourier transform (DFT) codebook, channel multipath components, or a basis consisting of the direction (angle) / distance (delay) / velocity (Doppler) of the sensing target range.

[0295] Furthermore, in any level of a P-level neural network, the channel estimation parameters of each neural network may be related to one or more layers of neural networks. This will be discussed in more detail below with examples of implementation.

[0296] Method 1: In a P-level neural network, the channel estimation parameters corresponding to one or more layers of neural networks contained in any level of neural network are the same.

[0297] In Method 1, in the P-level neural network included in the first model, the channel estimation parameters corresponding to different layers of the same level of neural network can be the same, so that different layers of the same level of neural network can obtain channel estimation results based on the same channel estimation parameters, thereby simplifying the implementation and improving the model processing efficiency.

[0298] As an example, as shown in Figure 4d, the first model can include the P-level neural network in Figure 4c. In Figure 4d, the channel estimation parameters corresponding to one or more layers of neural networks in any level of neural network are the same, satisfying:

[0299] The channel estimation parameters corresponding to one or more layers of neural networks contained in the first-level neural network are all the same, denoted as A_1.

[0300] The channel estimation parameters corresponding to one or more layers of neural networks contained in the p-th level neural network are all the same, denoted as A_p.

[0301] The channel estimation parameters corresponding to one or more layers of neural networks contained in the P-th level neural network are all the same, denoted as A_P.

[0302] Method 2: In a P-level neural network, the channel estimation parameters corresponding to one or more layers of neural networks contained in the p+1-th level neural network are determined based on the channel estimation parameters corresponding to one or more layers of neural networks contained in the p-th level neural network.

[0303] As an example, as shown in Figure 4e, the first model may include the P-level neural network in Figure 4d. In Figure 4e, the channel estimation parameters corresponding to one or more layers of neural networks contained in any level of neural network are determined based on the channel estimation parameters corresponding to one or more layers of neural networks contained in the previous level (if a previous level exists).

[0304] For example, in Figure 4e, the direction indicated by the black arrows is the direction of transmission of channel estimation parameters of different levels of neural networks, that is, the direction of the arrows between A_1, ..., A_p, ..., A_P in the figure.

[0305] Optionally, the channel estimation parameters corresponding to one or more neural networks included in the (p+1)th level neural network can be determined not only based on the channel estimation parameters corresponding to one or more neural networks included in the p-th level neural network, but also based on at least one of the following:

[0306] ① The channel estimation results output by one or more layers of neural networks contained in the p-th level neural network. In other words, the channel estimation parameters corresponding to the (p+1)-th level neural network can be determined based on the channel estimation results output by the p-th level neural network (e.g., the second output described later). In this way, the channel estimation results output by the upper-level neural network can be used to determine the channel estimation parameters of the lower-level neural network, so that the lower-level neural network can obtain the signal features accumulated by the upper-level neural network during the channel estimation process based on the channel estimation parameters, thereby improving the processing performance of the lower-level neural network.

[0307] ② The attention weights of the attention modules contained in one or more layers of the p-th level neural network (e.g., these attention weights can be represented as softmax(σQK), where softmax represents the normalized exponential operation and σ is a constant coefficient). T represents the dimension of the column vector of K, Q represents the query input to the attention module, and K represents the key input to the attention module. In other words, the channel estimation parameters corresponding to the (p+1)th level neural network can be determined based on the attention weights of the attention modules contained in the p-th level neural network. In this way, the attention weights of the attention modules contained in the upper-level neural network can be used to determine the channel estimation parameters of the lower-level neural network. This enables sparse signal estimation processing while also leveraging the application of attention modules to obtain the gain of the Transformer architecture (e.g., the attention mechanism allows the model to focus on the most relevant or important parts when processing the input, thereby improving the efficiency and accuracy of the model), thus improving the processing efficiency of sparse signal estimation.

[0308] It should be understood that attention modules can implement neural network processing through attention mechanisms. In attention mechanisms, Q represents the query vector, K represents the key vector, and V represents the value vector. For example, the query vector represents the information retrieved from the input data. In the case of self-attention, the query vector can be input data from the attention module and can be used to determine which parts of the input sequence are most important to the current task. The key vector is compared with the query vector to determine the allocation of attention. The key vector, also derived from the input data of the attention module, helps the system identify which parts of the input should be "attented" by the query vector. The degree of matching between the key and query vectors determines the strength of the attention. Furthermore, the value vectors contain actual information and, after being assigned attention weights, are further processed (e.g., weighted summation) to produce the output of the attention mechanism. Value vectors also come from the input data of the attention module, but they can have different representations than the query and key vectors.

[0309] Method 3: In a P-level neural network, the channel estimation parameters corresponding to the j-th layer of the p+1-th level neural network are based on the output of the j-th layer of the p-th level neural network (e.g., the second output s described later). j (j) is determined, where j is a positive integer.

[0310] As an example, as shown in Figure 4f, the first model may include the P-level neural network in Figure 4d. In Figure 4f, the channel estimation parameters corresponding to the j-th layer of the neural network in any level of the neural network are determined based on the output of the j-th layer of the neural network in the previous level (if a previous level exists).

[0311] Optionally, in a P-level neural network, the channel estimation parameters corresponding to different layers of the neural network contained in any level of the neural network can be different.

[0312] For example, the channel estimation parameters corresponding to the first layer of the first-level neural network are denoted as A_1-1 in Figure 4f, the channel estimation parameters corresponding to the second layer of the first-level neural network are denoted as A_1-2 in Figure 4f, ... the channel estimation parameters corresponding to the Nth layer of the first-level neural network are denoted as A_1-N in Figure 4f.

[0313] For example, the channel estimation parameters corresponding to the first layer of the second-level neural network are denoted as A_2-1 in Figure 4f, the channel estimation parameters corresponding to the second layer of the second-level neural network are denoted as A_2-2 in Figure 4f, ... the channel estimation parameters corresponding to the Nth layer of the second-level neural network are denoted as A_2-N in Figure 4f.

[0314] For example, the channel estimation parameters corresponding to the first layer of the neural network contained in the P-1 level neural network are denoted as A_P-1-1 in Figure 4f, the channel estimation parameters corresponding to the second layer of the neural network contained in the P-1 level neural network are denoted as A_P-1-2 in Figure 4f, ... the channel estimation parameters corresponding to the Nth layer of the neural network contained in the P-1 level neural network are denoted as A_P-1-N in Figure 4f.

[0315] For example, the channel estimation parameters corresponding to the first layer of the P-th level neural network are denoted as A_P-1 in Figure 4f, the channel estimation parameters corresponding to the second layer of the P-th level neural network are denoted as A_P-2 in Figure 4f, ... the channel estimation parameters corresponding to the N-th layer of the P-th level neural network are denoted as A_P-N in Figure 4f.

[0316] Furthermore, in Figure 4f, the direction indicated by the black arrows is the direction of transmission of channel estimation parameters for different levels of neural networks.

[0317] For example, between the first and second levels, the arrows pointing from 1-1 to A_2-1 in the diagram, from 1-2 to A_2-2 in the diagram, ..., and from 1-N to A_2-N in the diagram.

[0318] For example, between level P-1 and level P, the arrows pointing from P-1-1 to A_P-1 in the diagram, from P-1-2 to A_P-2 in the diagram, ..., and the arrows pointing from P-1-N to A_P-N in the diagram.

[0319] Optionally, the channel estimation parameters corresponding to the j-th layer of the neural network included in the p+1-th level neural network can be determined not only based on the output of the j-th layer of the neural network included in the p-th level neural network, but also based on at least one of the following:

[0320] ③ The channel estimation parameters output by the j-th layer of the p-th neural network. Refer to the implementation process of Method 2 above.

[0321] ④ The attention weights of the attention module in the j-th layer of the p-th neural network. Refer to the implementation process corresponding to "②" above.

[0322] For example, the channel estimation parameters corresponding to the j-th layer of the neural network included in the p+1-th level neural network. satisfy:

[0323] Among them, s t satisfy: or or This represents a number between -1 and 1, with a gap of 1 / G. s represents the s of the j-th layer of a P-th level neural network. t ,

[0324] In addition, b(s) t ) indicates that for s t The processing, [b(s t )] t＝1,2,… Let t elements be s1, s2, ..., and for different estimated signals or second outputs, the vector b(s) t ( ) Meets at least one of the following:

[0325] s t Indicates the direction of the radial path of the signal transmission path (e.g., s). t =θ): θ lN represents the l-th angle value. rx λ is the number of antennas, d is the antenna spacing, and λ is the wavelength.

[0326] s t The distance of the path of signal transmission (e.g., s) t =R): R m N represents the m-th distance value. f It is the number of subcarriers, Δ f Where c is the subcarrier spacing and c is the speed of light;

[0327] s t This represents the speed (e.g., s) of the scattering object (or signal transmitting device, or signal receiving device) moving along the signal transmission path. t =v): v n This represents the nth distance value. It is a signed number, T s It is the sign interval, f c 'c' is the carrier frequency, and 'c' is the speed of light.

[0328] in, This represents the second output corresponding to the j-th layer of the p-th neural network (for example, if the channel estimation parameters of the j-th layer of the p-th neural network include a codebook, the second output can include the direction / angle values ​​corresponding to that codebook; or if the channel estimation parameters of the j-th layer of the p-th neural network include a basis composed of the direction (angle) / distance (delay) / velocity (Doppler) of the sensing target range, the second output can include the direction / distance / velocity values ​​corresponding to that codebook), G represents the number of uniformly sampled or discretized parameters, ρ represents the learnable parameters, and softmax represents the normalization exponential operation (e.g., softmax(z)). i =exp(z i ) / ∑ k exp(z k )), σ is a constant coefficient (e.g. T is K j (the dimension of the column vector), This represents the parameters output by the (j-1)th layer of the p-th neural network to the j-th layer (i.e., the first output of the feedforward module contained in the (j-1)th layer of the neural network described later). This represents the key input to the attention module in the j-th layer of the p-th neural network, where Δs is a constant that determines s. t The range.

[0329] It is understandable that in Method 2 or Method 3, in the P-level neural network included in the first model, the channel estimation parameters corresponding to two adjacent neural networks can satisfy the above correlation, so that the next level neural network can obtain the feature information accumulated by the previous level neural network, thereby improving the processing performance of the next level neural network.

[0330] Optionally, in any one or more layers of neural networks contained in the P-level neural network, the j-th layer of the neural network satisfies at least one of the following methods A to F, where j is a positive integer. For example, j takes values ​​from 1 to N, that is, the j-th layer of the neural network can be the 1st, 2nd, ..., or Nth layer of the neural network shown in Figure 4a or Figure 4b above, or the j-th layer of the neural network can be any of the neural networks 1-1, 1-2, ... 1-N, p-1, p-2, ... pN, P-1, P-2, ..., or PN shown in any of the figures in Figures 4c to 4f above.

[0331] Method A: The j-th layer of the neural network includes an attention module, and the input of the attention module includes the j-th key K. j and the j-th value V j K j and the V j It is determined based on the channel estimation parameters corresponding to the j-th layer neural network.

[0332] As an example, as shown in Figure 4g, the input of the attention module in the j-th layer of the neural network (i.e., the j-th key K) j and the j-th value V j The value can be determined based on the channel estimation parameters (denoted as A_j in the figure) corresponding to the j-th layer neural network.

[0333] Among them, K j and V j It can be determined in a variety of ways.

[0334] Example 1, K j and V j It can be determined based on the additive parameter (i.e., K). j and V j (This can be obtained by adding several parameters). For example, K j satisfy:

[0335] Where α represents a configurable parameter, and A_j represents the channel estimation parameters of the j-th layer of the neural network. This represents the neural network parameters associated with the key in the j-th layer of the neural network.

[0336] Similarly, V j satisfy:

[0337] Where β represents a configurable parameter. The parameters of the neural network are associated with the values ​​of the j-th layer.

[0338] Optionally, α and β can be equal or unequal; this is not limited here.

[0339] Example 2, K j and V j It can be determined based on the multiplicative parameter (i.e., K). j and V j (This can be obtained by multiplying several parameters). For example, K j satisfy:

[0340] in, This represents the neural network parameters of the j-th layer that are related to K.

[0341] Similarly, V j satisfy:

[0342] in, This represents the neural network parameters of the j-th layer that are related to V.

[0343] Furthermore, in Examples 1 and 2, the processing of the attention module in the j-th layer of the neural network satisfies: Q j+1 / 2 =softmax(σQ) j K j V j ;

[0344] Where softmax represents the normalization exponent operation, Q j+1 / 2 This represents the third output of the attention module in the j-th layer of the neural network, where σ is a constant coefficient (e.g., ...). T is K j (the dimension of the column vector), Q j This represents the parameters output by the (j-1)th layer to the j-th layer of the neural network (i.e., the first output of the feedforward module contained in the (j-1)th layer of the neural network described later).

[0345] Example 3, K j and V j It can be determined based on a multi-head attention mechanism, for example, this multi-head attention mechanism can contain several branches (the number of branches can be 1, 2, 4, 8, 12, 16 or other values), and the calculation of the h-th branch is obtained. satisfy:

[0346] or,

[0347] Where α represents a configurable parameter. This represents the neural network parameters related to K for the h-th branch of the j-th layer neural network.

[0348] Similarly, V j satisfy:

[0349] or,

[0350] Where β represents a configurable parameter. This represents the neural network parameters associated with V for the h-th branch of the j-th layer neural network.

[0351] In Examples 1 and 2, the attention module in the j-th layer of the neural network processes the following:

[0352] Where softmax represents the normalization exponent operation, Q j+1 / 2 This represents the third output of the attention module in the j-th layer of the neural network, where σ is a constant coefficient (e.g., ...). T is K j (the dimension of the column vector), Q j W represents the parameters output by the (j-1)th layer to the j-th layer of the neural network (i.e., the first output of the feedforward module contained in the (j-1)th layer of the neural network described later). o For learnable multi-head fusion weights.

[0353] Method B: The j-th layer of the neural network includes an attention module, and the input of the attention module includes the j-th query Q. j Q j Satisfy: When j takes the value 1, Q j It is determined based on the received signal (e.g., the first signal received by the first communication device mentioned above, the second signal or the third signal received by the first communication device mentioned below); and / or, when j is greater than 1, Q j It is determined based on the first output of the feedforward module contained in the (j-1)th layer of the neural network.

[0354] As an example, as shown in Figure 4h, the first output of the feedforward module in the (j-1)th layer of the neural network can be used as the Q-value of the attention module in the j-th layer of the neural network. j .

[0355] Method C: The j-th layer neural network includes a feedforward module, and the first output of the feedforward module in the j-th layer neural network is used to determine the (j+1)-th query Q input to the attention module in the (j+1)-th layer neural network.j+1 .

[0356] As an example, as shown in Figure 4i, similar to the implementation process in Figure 4h, the first output of the feedforward module contained in the j-th layer neural network can be used as the Q-value of the attention module in the (j+1)-th layer neural network. j+1 .

[0357] Method D: The j-th layer neural network includes an attention module, and the second output of the attention module in the j-th layer neural network is used to determine the channel estimation result.

[0358] As an example, as shown in Figure 4j, the second output of the attention module in the j-th layer neural network can be used as the channel estimation result of the j-th layer neural network. Similarly, the second output of the attention module in the (j-1)-th layer neural network can be used as the channel estimation result of the (j-1)-th layer neural network.

[0359] For example, in the implementation of Example 1 above, the channel estimation result of the j-th layer neural network satisfies: s j =softmax(σQ) j K j )(Φ j +W j );

[0360] Similarly, in the implementation of Example 1 above, the channel estimation result of the j-th layer neural network satisfies: s j =softmax(σQ) j K j )(Φ j *W j );

[0361] Where softmax represents the normalization exponent operation, and σ is a constant coefficient (e.g., ...). T is K j (the dimension of the column vector), Q j K represents the parameters output from the (j-1)th layer to the j-th layer of the neural network (i.e., the first output of the feedforward module contained in the (j-1)th layer of the neural network described above). j Φ represents the key of the input to the attention module contained in the j-th layer of the neural network. j W represents the parameter value corresponding to the channel estimation parameters of the j-th layer neural network. j This represents the neural network parameters of the j-th layer.

[0362] Based on method D, it can be seen that for the P-level neural network contained in the first model, the channel estimation results of the output of one or more layers of neural networks in each level of the neural network can be referred to the above s. j The implementation process.

[0363] As described above, the estimation information output by the first model (such as the second estimation information described above) includes: a set of some or all of the channel estimation results from the N channel estimation results obtained by the N-layer neural network contained in the first model; and some or all of the channel estimation results from the N channel estimation results obtained by the N-layer neural network contained in the first model.

[0364] For example, a communication device (e.g., a first communication device or a second communication device) that has deployed a first model can estimate a second estimation information based on the first model to estimate a first signal received by the first communication device. This second estimation information includes the outputs of one or more layers of neural networks in each level of a P-level neural network. j The resulting set (denoted as Example A); or, the second estimation information includes the outputs of one or more layers of neural networks in each level of the P-level neural network. j (Referred to as Example B).

[0365] The following will provide an exemplary description of model management for the first model, using some implementation examples.

[0366] As an example, in Example A above, taking a value of P of 1 (i.e., the first model contains N layers of neural network models), the set of the second estimated information can be denoted as... (gather It contains elements s1, s2...s j ...s N The target value indicated by the first estimated information (or the expected value, baseline truth, etc. described above) can be denoted as: It can be used to perform model monitoring.

[0367] For example, based on The supervision loss of the model can be determined, and this supervision loss l satisfies:

[0368] Where s[j] represents the set The j-th element contained express The Π(j)th element, Indicates to Sort the data (to obtain a smaller error), min represents starting from... Determine the minimum value. Indicates to Perform the square operation by taking the absolute value. Furthermore, when used for model monitoring, the supervised loss measured based on the above method can be used as a monitoring value for model performance.

[0369] For example, based on The unsupervised loss of the model can be determined, and this unsupervised loss l aux satisfy:

[0370] Where Y represents the received signal (e.g., the first signal), This represents the received signal constructed based on the second estimation information. Indicates to Perform the square operation by taking the absolute value.

[0371] Optionally, during training, the loss function l+ρl is minimized. aux Training the model in a way that uses ρ to represent the weighting factor can update the model parameters of the neural network model, thereby improving the performance of the trained model (e.g., improving the accuracy of channel estimation and / or reducing errors).

[0372] As an example, in Example B above, taking P as 1 (i.e., the first model contains N layers of neural network), the second estimation information can include s1, s2...s j ...s N The target value indicated by the first estimated information (or the expected value, baseline truth, etc. described above) can be denoted as: It can be used to perform model monitoring.

[0373] For example, based on The successive loss function l of the model can be determined. i satisfy:

[0374] Where i represents up to the i-th layer, and s[j] represents s j , express The Π(j)th element (calculated from the first i layers, s) i It is s that includes the first i layers. Indicates to Sort the data (to obtain a smaller error), min represents starting from... Determine the minimum value. Indicates to Perform the square operation on the absolute value. Furthermore, when used for model monitoring, specifying the sparsity or the number / layers of estimated signals and comparing the loss values ​​of those estimated signals with that number or sparsity can serve as a monitoring value for model performance.

[0375] Optionally, during the progressive training of the model, the loss function l is minimized. iTraining the model in this way can update the model parameters of the neural network model, so that the trained model can improve its performance (e.g., improve the accuracy of channel estimation and / or reduce errors).

[0376] Optionally, during model fine-tuning, the loss function l = ∑ i w i l i (w i Model training can be performed in a way that represents the model parameters of a neural network model. This allows for updating of the model parameters, which can improve the performance of the trained model (e.g., improve channel estimation accuracy and / or reduce errors).

[0377] Method E: The j-th layer neural network includes an attention module and a feedforward module. The input of the feedforward module included in the j-th layer neural network includes the third output of the attention module included in the j-th layer neural network, and / or the fourth output of the feedforward module included in the (j-1)-th layer neural network.

[0378] Optionally, the input to the feedforward module included in the j-th layer neural network may also include received signals (e.g., the first signal received by the first communication device described above, the second signal received by the first communication device described below, the third signal, etc.).

[0379] Method F: The j-th layer neural network includes a feedforward module, and the fourth output of the feedforward module contained in the j-th layer neural network is used to determine the fourth output of the feedforward module contained in the (j+1)-th layer neural network.

[0380] As an example, as shown in Figure 4k, the input to the feedforward module of the j-th layer neural network includes the third output of the attention module of the j-th layer neural network and the fourth output of the feedforward module of the (j-1)-th layer neural network. Similarly, the input to the feedforward module of the (j+1)-th layer neural network includes the third output of the attention module of the (j+1)-th layer neural network and the fourth output of the feedforward module of the j-th layer neural network.

[0381] Taking Figure 4k as an example, the feedforward module in a certain layer of the neural network can update (or adjust, iterate, etc.) the Q input to that layer of the neural network, and it can also update (or adjust, iterate, etc.) the fourth output provided by the previous layer of the neural network. The implementation process of the feedforward module in the j-th layer of the neural network will be described below with some implementation examples.

[0382] For example, the fourth output M of the feedforward module contained in the j-th layer of the neural network j+1 satisfy:

[0383] M j+1 =FFN(Mj Q j+1 / 2 ,Y); or M j+1 =FFN(M j Q j+1 / 2 );

[0384] Where FFN represents the processing of the feedforward module neural network, M j Q represents the fourth output of the feedforward module in the (j-1)th layer of the neural network. j+1 / 2 Y represents the third output of the attention module contained in the j-th layer of the neural network, and Y represents the received signal.

[0385] For example, the Q processed by the feedforward module in the j-th layer of the neural network satisfies:

[0386] Q j+1 =YM j+1 ;or,

[0387] Among them, Q j+1 Y represents the first output of the feedforward module in the j-th layer of the neural network, and M represents the received signal. j+1 This represents the fourth output of the feedforward module contained in the j-th layer of the neural network. A represents the weighting coefficients obtained based on Y and the estimated signal. j Let f(M) represent the channel estimation parameters corresponding to the j-th layer of the neural network, β represent constant coefficients or learnable parameters, and f(M) represent the channel estimation parameters corresponding to the j-th layer of the neural network. j+1 ) indicates that for M j+1 Perform post-processing or neural network operations.

[0388] It should be noted that the feedforward module involved in any of the above implementation examples can be implemented using various neural network structures. For example, the feedforward module can be an MLP, CNN, long short-term memory network (LSTM), or other neural network structures.

[0389] Please refer to Figure 5, which is another schematic diagram of the communication method provided in this application, the method including the following steps.

[0390] S501. The second communication device sends a second signal, and correspondingly, the first communication device receives the second signal.

[0391] S502. The first communication device sends a sixth message, and correspondingly, the first communication device receives the sixth message. The sixth message indicates one or more sparsity-corresponding estimation information and the second signal; the sixth message is used for model management of the first model, which is used for channel estimation; the estimation information is obtained by estimating the channel corresponding to the second signal.

[0392] Optionally, the sixth message transmitted by the first communication device includes at least one of the following:

[0393] The first indication information indicates the one or more sparsities;

[0394] The second indication information indicates one or more residual errors corresponding to the one or more sparsities;

[0395] The third instruction information indicates the capability of the first communication device; or...

[0396] The fourth indication information is the configuration information of the reference signal.

[0397] In one possible implementation, the model parameters of the first model are determined based on the sixth information. Specifically, the sixth information sent by the first communication device can be used to determine the model parameters of the first model. That is, the recipient of the sixth information (e.g., the second communication device) can use the sixth information for the management of the first model (e.g., model update, model generation, etc.), so that the recipient obtains a first model that matches the estimation information obtained by the first communication device from channel estimation of the second signal.

[0398] In one possible implementation, the method further includes: the first communication device receiving seventh information, which instructs a second model obtained by training the first model. Specifically, the first communication device may also receive seventh information instructing the second model, which may be obtained by training the first model. In other words, the recipient of the sixth information (e.g., the second communication device) can train the first model to obtain the second model based on the sixth information, and deploy the second model to the first communication device via the seventh information, thereby achieving model training and model deployment.

[0399] Based on the scheme in Figure 5, after receiving the second signal, the first communication device can estimate the channel corresponding to the second signal to obtain estimation information corresponding to one or more sparsities. Subsequently, the first communication device can send sixth information indicating the one or more sparsity-corresponding estimation information and the second signal, so that the receiver of the sixth information can use it for model management of the first model, which is used for channel estimation; that is, the first model can be a channel estimation model. In other words, the sixth information sent by the first communication device can be used for model management of the channel estimation model. In this way, communication equipment deploying the channel estimation model can perform channel estimation on the received signal based on the channel estimation model, thereby improving the processing efficiency of sparse signal estimation.

[0400] Furthermore, the estimation information indicated by the sixth information sent by the first communication device corresponds to one or more sparsities; that is, the estimation information indicated by the sixth information can indicate the estimation information corresponding to different channel sparsities. In this way, the receiver of the sixth information (e.g., the second communication device) can determine the estimation information corresponding to one or more channel sparsities through the sixth information, thereby enabling the receiver to manage the channel estimation model using the estimation information corresponding to one or more channel sparsities, thus improving model management efficiency.

[0401] It should be understood that the first communication device can estimate the received second signal in a manner different from the first model to obtain estimation information corresponding to one or more sparsities. The following will describe this in conjunction with some implementation examples.

[0402] As an example of implementation, this other approach can be a traditional channel estimation method, including but not limited to channel estimation methods based on orthogonal matching pursuit (OMP) algorithms, Bayesian learning-based algorithms, etc.

[0403] As another implementation example, this other approach can be a processing method for models different from the first model. For example, if the model capability of the other model is superior to that of the first model, the former's model inference accuracy and / or the communication performance (e.g., system performance, link performance) corresponding to the model output can be superior to the latter. Accordingly, the estimated information corresponding to one or more sparsities obtained from the output of the other model can be used as the target value (or expected value, or baseline truth, etc.) corresponding to the estimated information output by the first model, so that the receiver of the sixth information can manage the first model based on the difference between the estimated information output by the first model indicated by the sixth information and the target value.

[0404] For example, if the model capability of the other model is inferior to that of the first model, the model inference accuracy and / or the communication performance (e.g., system performance, link performance) corresponding to the model output of the former may be inferior to that of the latter. Accordingly, the estimated information obtained by the output of the first model can be used as the target value (or expected value, or benchmark truth, etc.) corresponding to one or more estimated information output by the other model, so that the recipient of the sixth information can realize model management of the other model based on the difference between one or more estimated information indicated by the sixth information and the target value.

[0405] It should be noted that in the implementation process shown in Figure 5, the implementation process of the first model can refer to the previous description (for example, any implementation method of Figure 3, Figure 4a to Figure 4k).

[0406] Please refer to Figure 6, which is another schematic diagram of the communication method provided in this application, the method including the following steps.

[0407] S601. The fourth communication device sends a third signal, and correspondingly, the third communication device receives the third signal.

[0408] S602. The third communication device processes the third signal based on the first model to obtain the estimation information corresponding to the third signal; wherein the first model is used for channel estimation.

[0409] Optionally, before step S602, the method shown in Figure 6 further includes:

[0410] S600. The fourth communication device sends an eighth message, and correspondingly, the third communication device receives the eighth message, which is used to instruct the first model.

[0411] Alternatively, the third communication device may obtain / determine the first model through pre-configuration to reduce transmission overhead.

[0412] Based on the scheme shown in Figure 6, after receiving the third signal, the third communication device can estimate the channel corresponding to the third signal to obtain the estimated information corresponding to the third signal. Here, the first model is used for channel estimation; that is, the first model can be a channel estimation model. In this way, communication equipment deploying the channel estimation model can perform channel estimation on the received signal based on the channel estimation model, thereby improving the processing efficiency of sparse signal estimation.

[0413] It should be noted that in the implementation process shown in Figure 6, the implementation process of the first model can refer to the previous description (for example, any implementation method of Figure 3, Figure 4a to Figure 4k).

[0414] Referring to Figure 7, this application embodiment provides a communication device 700. This communication device 700 can implement the functions of the first communication device (or the second communication device, the third communication device, or the fourth communication device) in the above method embodiments, and therefore can also achieve the beneficial effects of the above method embodiments. In this application embodiment, the communication device 700 can be the first communication device (or the second communication device, the third communication device, or the fourth communication device), or it can be an integrated circuit or component inside the first communication device (or the second communication device, the third communication device, or the fourth communication device), such as a chip, baseband chip, modem chip, SoC chip (e.g., an SoC chip containing a modem core), SIP chip, communication module, chip system, processor, etc.

[0415] It should be noted that the transceiver unit 702 may include a transmitting unit and a receiving unit, which are used to perform transmitting and receiving respectively.

[0416] In one possible implementation, when the device 700 is used to execute the method performed by the first communication device in the embodiment shown in FIG3, the device 700 includes a processing unit 701 and a transceiver unit 702; the transceiver unit 702 is used to receive first information, which is used to indicate channel sparsity; the transceiver unit 702 is also used to receive a first signal; the processing unit 701 is used to determine second information; the transceiver unit 702 is also used to send the second information, which indicates first estimation information and the first signal; the second information is used for model management of a first model, which is used for channel estimation; the first estimation information is obtained by estimating the channel corresponding to the first signal based on the first information.

[0417] In one possible implementation, when the device 700 is used to execute the method performed by the second communication device in the embodiment shown in FIG3, the device 700 includes a processing unit 701 and a transceiver unit 702; the processing unit 701 is used to determine first information and a first signal; the transceiver unit 702 is used to transmit the first information, which is used to indicate channel sparsity; the transceiver unit 702 is also used to transmit the first signal; the transceiver unit 702 is also used to receive second information, which indicates first estimation information and the first signal; the second information is used for model management of a first model, which is used for channel estimation; the first estimation information is obtained by estimating the channel corresponding to the first signal based on the first information.

[0418] In one possible implementation, when the device 700 is used to execute the method performed by the first communication device in the embodiment shown in FIG5, the device 700 includes a processing unit 701 and a transceiver unit 702; the transceiver unit 702 is used to receive a second signal; the processing unit 701 is used to determine sixth information; the transceiver unit 702 is also used to send the sixth information, which indicates one or more sparsity-corresponding estimation information and the second signal; the sixth information is used for model management of a first model, which is used for channel estimation; the estimation information is obtained by estimating the channel corresponding to the second signal.

[0419] In one possible implementation, when the device 700 is used to execute the method performed by the second communication device in the embodiment shown in FIG5, the device 700 includes a processing unit 701 and a transceiver unit 702; the processing unit 701 is used to determine a second signal; the transceiver unit 702 is used to transmit the second signal; the transceiver unit is also used to receive sixth information, the sixth information indicating one or more sparsity-corresponding estimation information and the second signal; the sixth information is used for model management of a first model, the first model is used for channel estimation; the estimation information is obtained by estimating the channel corresponding to the second signal.

[0420] In one possible implementation, when the device 700 is used to execute the method performed by the third communication device in the embodiment shown in FIG6, the device 700 includes a processing unit 701 and a transceiver unit 702; the transceiver unit 702 is used to receive a third signal; the processing unit 701 is used to process the third signal based on a first model to obtain estimation information corresponding to the third signal; wherein, the first model is used for channel estimation.

[0421] In one possible implementation, when the device 700 is used to execute the method performed by the fourth communication device in the embodiment shown in FIG6, the device 700 includes a processing unit 701 and a transceiver unit 702; the processing unit 701 is used to determine eighth information; the transceiver unit 702 is used to transmit the eighth information, which is used to indicate a first model; wherein the first model is used to process a signal (e.g., a third signal) to obtain estimation information corresponding to the signal; wherein the first model is used for channel estimation.

[0422] It should be noted that the information execution process of the unit of the above-mentioned communication device 700 can be specifically described in the method embodiment shown above in this application, and will not be repeated here.

[0423] Please refer to Figure 8, which is another schematic structural diagram of the communication device 800 provided in this application. The communication device 800 includes a logic circuit 801 and an input / output interface 802. The communication device 800 can be a chip or an integrated circuit.

[0424] In this context, the transceiver unit 702 shown in Figure 7 can be a communication interface, which can be the input / output interface 802 in Figure 8, and the input / output interface 802 can include an input interface and an output interface. Alternatively, the communication interface can also be a transceiver circuit, which can include an input interface circuit and an output interface circuit.

[0425] Optionally, the input / output interface 802 is used to receive first information, which is used to indicate channel sparsity; the input / output interface 802 is also used to receive a first signal; the logic circuit 801 is used to determine second information; the input / output interface 802 is also used to send second information, which indicates first estimation information and the first signal; the second information is used for model management of the first model, which is used for channel estimation; the first estimation information is obtained by estimating the channel corresponding to the first signal based on the first information.

[0426] Optionally, the logic circuit 801 is used to determine the first information and the first signal; the input / output interface 802 is used to send the first information, which is used to indicate channel sparsity; the input / output interface 802 is also used to send the first signal; the input / output interface 802 is also used to receive the second information, which indicates the first estimation information and the first signal; the second information is used for model management of the first model, which is used for channel estimation; the first estimation information is obtained by estimating the channel corresponding to the first signal based on the first information.

[0427] Optionally, the input / output interface 802 is used to receive the second signal; the logic circuit 801 is used to determine the sixth information; the input / output interface 802 is also used to send the sixth information, which indicates one or more sparsity-corresponding estimation information and the second signal; the sixth information is used for model management of the first model, which is used for channel estimation; the estimation information is obtained by estimating the channel corresponding to the second signal.

[0428] Optionally, logic circuit 801 is used to determine the second signal; input / output interface 802 is used to send the second signal; input / output interface 802 is also used to receive sixth information, the sixth information indicating one or more sparsity-corresponding estimation information and the second signal; the sixth information is used for model management of the first model, the first model is used for channel estimation; the estimation information is obtained by estimating the channel corresponding to the second signal.

[0429] Optionally, the input / output interface 802 is used to receive a third signal; the logic circuit 801 is used to process the third signal based on the first model to obtain the estimation information corresponding to the third signal; wherein, the first model is used for channel estimation.

[0430] Optionally, logic circuit 801 is used to determine eighth information; input / output interface 802 is used to send the eighth information, which is used to indicate a first model; wherein the first model is used to process a signal (e.g., a third signal) to obtain estimation information corresponding to the signal; wherein the first model is used for channel estimation.

[0431] The logic circuit 801 and the input / output interface 802 can also perform other steps performed by the first or second communication device in any embodiment and achieve corresponding beneficial effects, which will not be elaborated here.

[0432] In one possible implementation, the processing unit 701 shown in FIG7 can be the logic circuit 801 in FIG8.

[0433] Optionally, the logic circuit 801 can be a processing device, the functions of which can be partially or entirely implemented in software.

[0434] Optionally, the processing apparatus may include a memory and a processor, wherein the memory is used to store a computer program, and the processor reads and executes the computer program stored in the memory to perform the corresponding processing and / or steps in any of the method embodiments.

[0435] Optionally, the processing device may consist of only a processor. A memory for storing computer programs is located outside the processing device, and the processor is connected to the memory via circuitry / wires to read and execute the computer programs stored in the memory. The memory and processor may be integrated together or physically independent of each other.

[0436] Optionally, the processing device may be one or more chips, or one or more integrated circuits. For example, the processing device may be one or more field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), system-on-chips (SoCs), central processing units (CPUs), network processors (NPs), digital signal processors (DSPs), microcontroller units (MCUs), programmable logic devices (PLDs), or other integrated chips, or any combination of the above chips or processors.

[0437] Please refer to Figure 9, which shows the communication device 900 involved in the above embodiments provided in the embodiments of this application. Specifically, the communication device 900 can be the communication device as a terminal device in the above embodiments. The communication device shown in Figure 9 is implemented through a terminal device (or a component in the terminal device).

[0438] The present invention provides a possible logical structure diagram of the communication device 900, which may include, but is not limited to, at least one processor 901 and a communication port 902.

[0439] In Figure 7, the transceiver unit 702 can be a communication interface, which can be the communication port 902 in Figure 9. The communication port 902 can include an input interface and an output interface. Alternatively, the communication port 902 can also be a transceiver circuit, which can include an input interface circuit and an output interface circuit.

[0440] Further optionally, the device may also include at least one of a memory 903 and a bus 904. In the embodiments of this application, the at least one processor 901 is used to control the operation of the communication device 900.

[0441] Furthermore, the processor 901 can be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit, a field-programmable gate array, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. The processor can also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a digital signal processor and a microprocessor, etc. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0442] It should be noted that the communication device 900 shown in Figure 9 can be used to implement the steps implemented by the terminal device in the aforementioned method embodiments and to achieve the corresponding technical effects of the terminal device. The specific implementation of the communication device shown in Figure 9 can be referred to the description in the aforementioned method embodiments, and will not be repeated here.

[0443] Please refer to Figure 10, which is a schematic diagram of the structure of the communication device 1000 involved in the above embodiments provided in the embodiments of this application. The communication device 1000 can specifically be a communication device as a network device in the above embodiments. The communication device shown in Figure 10 is implemented through a network device (or a component in a network device). The structure of the communication device can refer to the structure shown in Figure 10.

[0444] The communication device 1000 includes at least one processor 1011 and at least one network interface 1014. Optionally, the communication device further includes at least one memory 1012, at least one transceiver 1013, and one or more antennas 1015. The processor 1011, memory 1012, transceiver 1013, and network interface 1014 are connected, for example, via a bus. In this embodiment, the connection may include various interfaces, transmission lines, or buses, etc., and this embodiment is not limited thereto. The antenna 1015 is connected to the transceiver 1013. The network interface 1014 enables the communication device to communicate with other communication devices through a communication link. For example, the network interface 1014 may include a network interface between the communication device and core network equipment, such as an S1 interface; the network interface may also include a network interface between the communication device and other communication devices (e.g., other network devices or core network equipment), such as an X2 or Xn interface.

[0445] In this context, the transceiver unit 702 shown in Figure 7 can be a communication interface, which can be the network interface 1014 in Figure 10. The network interface 1014 can include an input interface and an output interface. Alternatively, the network interface 1014 can also be a transceiver circuit, which can include an input interface circuit and an output interface circuit.

[0446] The processor 1011 is primarily used to process communication protocols and communication data, control the entire communication device, execute software programs, and process data from these programs, for example, to support the actions described in the embodiments of the communication device. The communication device may include a baseband processor and a central processing unit (CPU). The baseband processor is primarily used to process communication protocols and communication data, while the CPU is primarily used to control the entire terminal device, execute software programs, and process data from these programs. The processor 1011 in Figure 10 can integrate the functions of both a baseband processor and a CPU. Those skilled in the art will understand that the baseband processor and CPU can also be independent processors interconnected via technologies such as buses. Those skilled in the art will understand that a terminal device can include multiple baseband processors to adapt to different network standards, and multiple CPUs to enhance its processing capabilities. Various components of the terminal device can be connected via various buses. The baseband processor can also be described as a baseband processing circuit or a baseband processing chip. The CPU can also be described as a central processing circuit or a central processing chip. The function of processing communication protocols and communication data can be built into the processor or stored in memory as a software program, which is then executed by the processor to implement the baseband processing function.

[0447] The memory is primarily used to store software programs and data. The memory 1012 can exist independently or be connected to the processor 1011. Optionally, the memory 1012 can be integrated with the processor 1011, for example, integrated within a single chip. The memory 1012 can store program code that executes the technical solutions of the embodiments of this application, and its execution is controlled by the processor 1011. The various types of computer program code being executed can also be considered as drivers for the processor 1011.

[0448] Figure 10 shows only one memory and one processor. In actual terminal devices, there may be multiple processors and multiple memories. Memory can also be called storage medium or storage device, etc. Memory can be a storage element on the same chip as the processor, i.e., an on-chip storage element, or it can be a separate storage element; this application does not limit this.

[0449] Transceiver 1013 can be used to support the reception or transmission of radio frequency (RF) signals between a communication device and a terminal. Transceiver 1013 can be connected to antenna 1015. Transceiver 1013 includes a transmitter Tx and a receiver Rx. Specifically, one or more antennas 1015 can receive RF signals. The receiver Rx of transceiver 1013 is used to receive the RF signals from the antennas, convert the RF signals into digital baseband signals or digital intermediate frequency (IF) signals, and provide the digital baseband signals or IF signals to processor 1011 so that processor 1011 can perform further processing on the digital baseband signals or IF signals, such as demodulation and decoding. In addition, the transmitter Tx in transceiver 1013 is also used to receive modulated digital baseband signals or IF signals from processor 1011, convert the modulated digital baseband signals or IF signals into RF signals, and transmit the RF signals through one or more antennas 1015. Specifically, the receiver Rx can selectively perform one or more stages of downmixing and analog-to-digital conversion on the radio frequency signal to obtain a digital baseband signal or a digital intermediate frequency (IF) signal. The order of these downmixing and IF conversion processes is adjustable. The transmitter Tx can selectively perform one or more stages of upmixing and digital-to-analog conversion on the modulated digital baseband signal or digital IF signal to obtain a radio frequency signal. The order of these upmixing and IF conversion processes is also adjustable. The digital baseband signal and the digital IF signal can be collectively referred to as digital signals.

[0450] The transceiver 1013 can also be called a transceiver unit, transceiver, transceiver device, etc. Optionally, the device in the transceiver unit that performs the receiving function can be regarded as the receiving unit, and the device in the transceiver unit that performs the transmitting function can be regarded as the transmitting unit. That is, the transceiver unit includes a receiving unit and a transmitting unit. The receiving unit can also be called a receiver, input port, receiving circuit, etc., and the transmitting unit can be called a transmitter, transmitter, or transmitting circuit, etc.

[0451] It should be noted that the communication device 1000 shown in Figure 10 can be used to implement the steps implemented by the network device in the aforementioned method embodiments and to achieve the corresponding technical effects of the network device. The specific implementation of the communication device 1000 shown in Figure 10 can be referred to the description in the aforementioned method embodiments, and will not be repeated here.

[0452] Please refer to Figure 11, which is a schematic diagram of the structure of the communication device involved in the above embodiments provided in the embodiments of this application.

[0453] It is understood that the communication device 110 includes, for example, modules, units, elements, circuits, or interfaces, which are appropriately configured together to execute the technical solutions provided in this application. The communication device 110 may be the terminal device or network device described above, or a component (e.g., a chip) within these devices, used to implement the methods described in the following method embodiments. The communication device 110 includes one or more processors 111. The processor 111 may be a general-purpose processor or a dedicated processor, for example, a baseband processor or a central processing unit. The baseband processor can be used to process communication protocols and communication data, and the central processing unit can be used to control the communication device (e.g., a RAN node, terminal, or chip), execute software programs, and process data from the software programs.

[0454] Optionally, in one design, the processor 111 may include a program 113 (sometimes also referred to as code or instructions) that can be executed on the processor 111 to cause the communication device 110 to perform the methods described in the embodiments below. In yet another possible design, the communication device 110 includes circuitry (not shown in FIG11).

[0455] Optionally, the communication device 110 may include one or more memories 112 storing a program 114 (sometimes referred to as code or instructions), which can be run on the processor 111 to cause the communication device 110 to perform the methods described in the above method embodiments.

[0456] Optionally, the processor 111 and / or memory 112 may include AI modules 117 and 118, which are used to implement AI-related functions. The AI ​​modules can be implemented through software, hardware, or a combination of both. For example, the AI ​​module may include a radio intelligence control (RIC) module. For example, the AI ​​module may be a near real-time RIC or a non-real-time RIC.

[0457] Optionally, the processor 111 and / or memory 112 may also store data. The processor and memory may be configured separately or integrated together.

[0458] Optionally, the communication device 110 may further include a transceiver 115 and / or an antenna 116. The processor 111, sometimes referred to as a processing unit, controls the communication device (e.g., a RAN node or terminal). The transceiver 115, sometimes referred to as a transceiver unit, transceiver, transceiver circuit, or transceiver, is used to realize the transmission and reception functions of the communication device through the antenna 116.

[0459] In this context, the processing unit 701 shown in Figure 7 can be a processor 111. The transceiver unit 702 shown in Figure 7 can be a communication interface, which can be the transceiver 115 in Figure 11. The transceiver 115 can include an input interface and an output interface. Alternatively, the transceiver 115 can also be a transceiver circuit, which can include an input interface circuit and an output interface circuit.

[0460] This application also provides a computer-readable storage medium for storing one or more computer-executable instructions. When the computer-executable instructions are executed by a processor, the processor performs the method described in the possible implementations of the first or second communication device in the foregoing embodiments.

[0461] This application also provides a computer program product (or computer program) that, when executed by a processor, executes the method of the first, second, third, or fourth communication device as described above.

[0462] This application also provides a chip system including at least one processor for supporting a communication device in implementing the functions involved in the possible implementations of the communication device described above. Optionally, the chip system further includes an interface circuit that provides program instructions and / or data to the at least one processor. In one possible design, the chip system may further include a memory for storing the program instructions and data necessary for the communication device. The chip system may be composed of chips or may include chips and other discrete devices, wherein the communication device may specifically be the first, second, third, or fourth communication device in the aforementioned method embodiments.

[0463] This application also provides a communication system, which includes a first communication device and a second communication device from any of the above embodiments. Alternatively, the communication system includes a third communication device and / or a fourth communication device from any of the above embodiments.

[0464] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between apparatuses or units through some interfaces, and may be electrical, mechanical, or other forms.

[0465] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0466] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

Claims

A communication method, characterized in that, include: Receive first information, which is used to indicate channel sparsity; Receive the first signal; Send a second message, which indicates the first estimated information and the first signal; The second information is used for model management of the first model, and the first model is used for channel estimation; the first estimation information is obtained by estimating the channel corresponding to the first signal based on the first information. The method according to claim 1, characterized in that, The method further includes: A third message is sent, the third message being used to determine the first message; wherein the third message indicates one or more estimation messages of the reference signal, each estimation message indicating a residual error corresponding to one of the channel sparsities of one or more channel sparsities. The method according to claim 2, characterized in that, The model parameters of the first model are determined based on the third information. The method according to any one of claims 1 to 3, characterized in that, The model parameters of the first model are determined based on the first information. The method according to any one of claims 1 to 4, characterized in that, The model management includes model training and / or model monitoring. The method according to any one of claims 1 to 5, characterized in that, The method further includes: Receive fourth information, which is used to indicate a second model, which is obtained by training the model based on the first model. The method according to any one of claims 1 to 6, characterized in that, The first information is also used to indicate at least one of the following: At least one of the following: residual error, capability information of the first communication device, and configuration information of the reference signal. The method according to any one of claims 1 to 7, characterized in that, The method further includes: A fifth message is sent, indicating the difference between the first estimation information and the second estimation information, wherein the second estimation information is obtained based on the first model. A communication method, characterized in that, include: Send a first message, which is used to indicate the channel sparsity; Send the first signal; Receive second information, the second information indicating the first estimation information and the first signal; The second information is used for model management of the first model, and the first model is used for channel estimation; the first estimation information is obtained by estimating the channel corresponding to the first signal based on the first information. The method according to claim 9, characterized in that, The method further includes: Receive third information, the third information being used to determine the first information; wherein the third information indicates one or more estimation information of the reference signal, each estimation information indicating the residual error corresponding to one of the one or more channel sparsities. The method according to claim 10, characterized in that, The model parameters of the first model are determined based on the third information. The method according to any one of claims 9 to 11, characterized in that, The model parameters of the first model are determined based on the first information. The method according to any one of claims 9 to 12, characterized in that, The model management includes model training and / or model monitoring (the monitoring method varies depending on which side it is deployed on). The method according to any one of claims 9 to 13, characterized in that, The method further includes: Send a fourth message, which is used to instruct a second model, which is obtained by training the model based on the first model. The method according to any one of claims 9 to 14, characterized in that, The first information is also used to indicate at least one of the following: At least one of the following: residual error, capability information of the first communication device, and configuration information of the reference signal. The method according to any one of claims 9 to 15, characterized in that, The method further includes: A fifth piece of information is received, which indicates the difference between the first estimation information and the second estimation information, wherein the second estimation information is obtained based on the first model. A communication method, characterized in that, include: Receive the second signal; Send a sixth message, which indicates one or more estimates of sparsity and the second signal; The sixth information is used for model management of the first model, which is used for channel estimation; the estimation information is obtained by estimating the channel corresponding to the second signal. The method according to claim 17, characterized in that, The sixth piece of information includes at least one of the following: The first indication information indicates the one or more sparsities; The second indication information indicates one or more residual errors corresponding to the one or more sparsities; The third instruction information indicates the capability of the first communication device; or... The fourth indication information is the configuration information of the reference signal. The method according to claim 17 or 18, characterized in that, The model parameters of the first model are determined based on the sixth information. The method according to any one of claims 17 to 19, characterized in that, The method further includes: Receive a seventh message, which is used to indicate a second model, which is obtained by training the model based on the first model. A communication method, characterized in that, include: Send a second signal; Receive sixth information, the sixth information indicating one or more sparsity-corresponding estimation information and the second signal; The sixth information is used for model management of the first model, which is used for channel estimation; the estimation information is obtained by estimating the channel corresponding to the second signal. The method according to claim 21, characterized in that, The sixth piece of information includes at least one of the following: The first indication information indicates the one or more sparsities; The second indication information indicates one or more residual errors corresponding to the one or more sparsities; The third instruction information indicates the capability of the first communication device; or... The fourth indication information is the configuration information of the reference signal. The method according to claim 21 or 22 is characterized in that, The model parameters of the first model are determined based on the sixth information. The method according to any one of claims 21 to 23, characterized in that, The method further includes: Send a seventh message, which is used to indicate a second model, which is obtained by training the model based on the first model. The method according to any one of claims 1 to 24, characterized in that, The first model includes an N-layer neural network. The input of the (i+1)th layer of the N-layer neural network is determined based on the output of the i-th layer of the neural network. N is a positive integer, and i takes values ​​from 1 to N-1. The method according to claim 25, characterized in that, The first model includes a P-level neural network, wherein one level of the P-level neural network includes the N-layer neural network, and different levels of neural networks contain different neural networks, where P is a positive integer. The method according to claim 26, characterized in that, The P-level neural network satisfies any one of the following: In the P-level neural network, the channel estimation parameters corresponding to one or more layers of neural networks contained in any level of neural network are the same; In the P-level neural network, the channel estimation parameters corresponding to one or more layers of neural networks included in the (p+1)-th level neural network are determined based on the channel estimation parameters corresponding to one or more layers of neural networks included in the p-th level neural network; or, In the P-level neural network, the channel estimation parameters corresponding to the j-th layer of the p+1-th level neural network are determined based on the channel estimation parameters corresponding to the j-th layer of the p-th level neural network, where j is a positive integer. The method according to claim 26 or 27 is characterized in that, In any one or more layers of the P-level neural network, the j-th layer satisfies at least one of the following conditions, where j is a positive integer: The j-th layer neural network includes an attention module, and the input of the attention module includes the j-th key K. j and the j-th value V j The K j and the V j It is determined based on the channel estimation parameters corresponding to the j-th layer neural network; The j-th layer neural network includes an attention module, and the input of the attention module includes the j-th query Q. j The Q j Satisfy the following condition: When j is 1, the Q... j It is determined based on the received signal; and / or, when j is greater than 1, the Q... j It is determined based on the first output of the feedforward module contained in the (j-1)th layer of the neural network; The j-th layer neural network includes a feedforward module, and the first output of the feedforward module in the j-th layer neural network is used to determine the (j+1)-th query Q input to the attention module in the (j+1)-th layer neural network. j+1 ; The j-th layer neural network includes an attention module, and the second output of the attention module in the j-th layer neural network is used to determine the channel estimation result; The j-th layer neural network includes an attention module and a feedforward module. The input of the feedforward module in the j-th layer neural network includes the third output of the attention module in the j-th layer neural network, and / or, the fourth output of the feedforward module in the (j-1)-th layer neural network; or, The j-th layer neural network includes a feedforward module, and the fourth output of the feedforward module included in the j-th layer neural network is used to determine the fourth output of the feedforward module included in the (j+1)-th layer neural network. A communication device, characterized in that, It includes a module for performing the method as described in any one of claims 1 to 8, or a module for performing the method as described in any one of claims 9 to 16, or a module for performing the method as described in any one of claims 17 to 20, or a module for performing the method as described in any one of claims 21 to 28. A communication device, characterized in that, It includes at least one processor, said at least one processor being configured to perform the method as claimed in any one of claims 1 to 8, or to perform the method as claimed in any one of claims 9 to 16, or to perform the method as claimed in any one of claims 17 to 20, or to perform the method as claimed in any one of claims 21 to 28. The communication device according to claim 30 is characterized in that, It also includes a memory that stores computer programs or instructions. The communication device according to claim 30 is characterized in that, The communication device is a chip or chip system. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer programs or instructions, which, when executed by a communication device,... Implement the method as described in any one of claims 1 to 8; or, Implement the method as described in any one of claims 9 to 16; or, Implement the method as described in any one of claims 17 to 20; or, Implement the method as described in any one of claims 21 to 28. A computer program product, characterized in that, This includes computer programs or instructions that, when executed by a computer, Implement the method as described in any one of claims 1 to 8; or, Implement the method as described in any one of claims 9 to 16; or, Implement the method as described in any one of claims 17 to 20; or, Implement the method as described in any one of claims 21 to 28.

Images