Communication method and communication device

Abstract

The invention discloses a communication method and a communication device, in the method, network equipment realizes port dimension reduction based on a dimension reduction model to measure an equivalent downlink channel, and terminal equipment feeds back a compression feedback quantity of equivalent channel information obtained based on a compression model matched with the dimension reduction model to the network equipment; a network device obtains compressed equivalent downlink channel information reconstructed based on a reconstruction model matched with a dimensionality reduction model. In the method, the dimensionality reduction model, a compression model and the reconstruction model are matched models after joint training. According to the method, the compressibility of the equivalent channel information can be improved, and the feedback overhead is further reduced. The model involved in the invention can be an AI model, a neural network model, a machine learning model and the like.

Description

Technical Field

[0001] This application relates to the field of communications, and more particularly, to a communication method and a communication device. Background Art

[0002] In a communication system, a network device can implement measuring equivalent downlink channels under multiple physical antennas by using a small number of downlink reference signals through port dimension reduction. A terminal device measures the received downlink reference signals to obtain downlink channel information, and performs compression processing on the obtained downlink channel information based on an artificial intelligence (AI) model. Then, the compressed information is fed back to the network device through uplink control information (UCI). Currently, the dimension reduction weights used for port dimension reduction can be calculated through an orthogonal codebook or orthogonal decomposition. However, it is found that based on this dimension reduction weight, the compressibility of the measured equivalent channel information is not high. Therefore, how to further reduce the feedback overhead has become an urgent problem to be solved. Summary of the Invention

[0003] This application provides a communication method and a communication device, which can further reduce the feedback overhead of downlink channel information.

[0004] In a first aspect, a communication method is provided. This method can be executed by a network device, or can also be executed by a component (such as a chip or a circuit) of the network device, and this is not limited.

[0005] The method includes: performing digital precoding on M reference signals corresponding to M antenna ports based on dimension reduction weights, where the dimension reduction weights are used to measure equivalent downlink channels of N physical antennas by using the M reference signals, N is a positive integer, M is a positive integer less than N, and the dimension reduction weights are determined based on a dimension reduction model; sending the M reference signals; sending first indication information, where the first indication information indicates a compression model, and the compression model matches the dimension reduction model; receiving a first sequence, where the first sequence is a compressed feedback quantity obtained by using first downlink channel information as an input to the compression model, and the first downlink channel information is obtained based on the measurement of the M reference signals; obtaining second downlink channel information, where the second downlink channel information is information obtained by using the first sequence as an input to a reconstruction model, and the reconstruction model matches the dimension reduction model, and the second downlink channel is used to determine the downlink channel information of the N physical antennas.

[0006] It can be understood that the optimization objective of the current dimensionality reduction method only minimizes the energy loss of the equivalent channel after dimensionality reduction. Such a method does not take into account the compressibility of the channel after dimensionality reduction. In this application, the dimensionality reduction model and the compression model are matching models. That is to say, the dimensionality reduction method and the compression method in this application are jointly optimized. While achieving port dimensionality reduction, the compressibility of the channel after dimensionality reduction is also considered. Therefore, compared with the current downlink channel information feedback scheme, the above technical solution can make the compressibility of the equivalent downlink channel information higher, thus achieving the effect of smaller feedback overhead.

[0007] In some implementation manners of the first aspect, the method further includes: inputting the first sequence into a reconstruction model for decompression to obtain second downlink channel information.

[0008] In a second aspect, a communication method is provided. This method can be executed by a terminal device, or can also be executed by a component (such as a chip or a circuit) of the terminal device, and this is not limited.

[0009] The method includes: receiving M reference signals corresponding to M antenna ports, where M is a positive integer; obtaining first downlink channel information based on the M reference signals; receiving first indication information, where the first indication information indicates a compression model, and the compression model matches the dimensionality reduction model. The dimensionality reduction model is used to obtain dimensionality reduction weights, and the dimensionality reduction weights are used to measure the equivalent downlink channel of N physical antennas by using the M reference signals, where N is a positive integer and M is a positive integer less than N; sending a first sequence, where the first sequence is a compression feedback amount obtained by taking the first downlink channel information as the input of the compression model. The first sequence is used as the input of a reconstruction model to be decompressed to obtain second downlink channel information, and the reconstruction model matches the dimensionality reduction model. The second downlink channel is used to determine the downlink channel information of N physical antennas.

[0010] For the beneficial effects of the second aspect, refer to the description of the first aspect, and details are not described here again.

[0011] In some implementation manners of the first aspect or the second aspect, the method further includes: inputting the first downlink channel information into a compression model for compression to obtain a first sequence.

[0012] In some implementation manners of the first aspect or the second aspect, the dimensionality reduction weights are output information obtained by taking channel prior information as the input of the dimensionality reduction model.

[0013] In some implementation manners of the first aspect or the second aspect, the channel prior information is determined based on at least one of the following information: currently measured uplink channel information, historical uplink channel information, historical downlink channel information, or sensed channel information.

[0014] In certain implementations of the first aspect or the second aspect, the dimensionality reduction model, the compression model, and the reconstruction model are all related to the number of antenna ports M, the number of physical antennas N, and the channel prior information.

[0015] In certain implementations of the first aspect or the second aspect, the first downlink channel information indicates the response of the equivalent downlink channel, and the first sequence is the compressed feedback amount of the response.

[0016] In certain implementations of the first aspect or the second aspect, the first downlink channel information indicates the precoding matrix corresponding to the response of the equivalent downlink channel, and the first sequence is the compressed feedback amount of the precoding matrix information corresponding to the response.

[0017] In certain implementations of the first aspect or the second aspect, the downlink channel information of N physical antennas includes the downlink precoding matrix of N physical antennas and the downlink channel response of N physical antennas.

[0018] In certain implementations of the first aspect, obtain the downlink precoding matrix of N physical antennas, and the downlink precoding matrix of N physical antennas is determined based on the dimensionality reduction weights and the second downlink channel information.

[0019] In a third aspect, a communication method is provided. This method can be executed by a first device, or alternatively, it can be executed by a component (such as a chip or a circuit) of the first device, and there is no limitation in this regard. For example, the first device can be a network device or an AI network element #1, such as a third-party network element #1.

[0020] The method includes: based on the dimensionality reduction model, obtaining dimensionality reduction weights, which are used to perform digital precoding on the M reference signals corresponding to M antenna ports to measure the equivalent downlink channel of N physical antennas, where N is a positive integer and M is a positive integer less than N; inputting the first sequence into the reconstruction model, where the first sequence is the compressed feedback amount obtained by using the first downlink channel information as the input of the compression model, the first downlink channel information is obtained by measuring the M reference signals, and the compression model, the reconstruction model, and the dimensionality reduction model are matched; outputting the second downlink channel information, and the second downlink channel is used to determine the downlink channel information of N physical antennas.

[0021] In the above technical solution, the dimensionality reduction model, the compression model, and the reconstruction model are models that are matched after joint training. This method can increase the compressibility of the equivalent downlink channel information and reduce the feedback overhead.

[0022] In a fourth aspect, a communication method is provided. This method can be executed by a second device, or alternatively, it can be executed by a component (such as a chip or a circuit) of the second device, and there is no limitation in this regard. For example, the second device can be a terminal device or an AI network element #2, such as a third-party network element #2.

[0023] The method includes: inputting first downlink channel information into a compression model, where the first downlink channel information is obtained by measuring M reference signals, and the M reference signals are reference signals obtained by digitally precoding the reference signals corresponding to M antenna ports based on reduced-dimensional weights, and the reduced-dimensional weights are used to measure the equivalent downlink channel of N physical antennas by using the M reference signals, N is a positive integer, M is a positive integer less than N, the reduced-dimensional weights are determined based on a reduced-dimensional model, and the compression model and the reduced-dimensional model match; outputting a first sequence, where the first sequence is used as an input to a reconstruction model to be decompressed to obtain second downlink channel information, the reconstruction model is associated with the reduced-dimensional model, and the second downlink channel is used to determine the downlink channel information of N physical antennas, and the reconstruction model is related to the reduced-dimensional model.

[0024] For the beneficial effects of the fourth aspect, refer to the description of the third aspect and will not be elaborated here.

[0025] In some implementation manners of the third aspect or the fourth aspect, the reduced-dimensional weights are output information obtained by using channel prior information as an input to the reduced-dimensional model.

[0026] In some implementation manners of the third aspect or the fourth aspect, the channel prior information is determined based on at least one of the following information: currently measured uplink channel information, historical uplink channel information, historical downlink channel information, or sensed channel information.

[0027] In some implementation manners of the third aspect or the fourth aspect, the reduced-dimensional model, the compression model, and the reconstruction model are all related to the number of antenna ports M, the number of physical antennas N, and the channel prior information.

[0028] In some implementation manners of the third aspect or the fourth aspect, the first downlink channel information indicates the response of the equivalent downlink channel, and the first sequence is the compressed feedback amount of the response.

[0029] In some implementation manners of the third aspect or the fourth aspect, the first downlink channel information indicates the precoding matrix corresponding to the response of the equivalent downlink channel, and the first sequence is the compressed feedback amount of the precoding matrix information corresponding to the response.

[0030] In some implementation manners of the third aspect or the fourth aspect, the downlink channel information of N physical antennas includes the downlink precoding matrix of N physical antennas and the downlink channel response of N physical antennas.

[0031] In a fifth aspect, a method for model training is provided. This method can be executed by a first device, or can also be executed by a component (such as a chip or a circuit) of the first device, and this is not limited. For example, the first device can be a network device or an AI network element #1, such as a third-party network element #1.

[0032] The method includes: obtaining a training data set, where the training data set includes channel prior information and first downlink channel information, and the first downlink channel information is the true downlink channel information of N physical antennas, and N is a positive integer; sending second downlink channel information, where the second downlink channel information is determined based on the first downlink channel information and a dimension reduction weight, and the dimension reduction weight is the output information obtained by taking the channel prior information as the input of a dimension reduction model, and the dimension reduction weight is used to perform digital precoding on M reference signals corresponding to M antenna ports to measure the equivalent downlink channel of N physical antennas, and M is a positive integer less than N; receiving a first sequence, where the first sequence is the output information obtained by taking the second downlink channel information as the input of a compression model; sending first gradient information, where the first gradient information is the input-side gradient information of a reconstruction model, and the first gradient information is used to update the network parameters of the compression model, and the first gradient information is determined based on first error information and the reconstruction model, and the first error information is determined based on third downlink channel information and the first downlink channel information, and the third downlink channel information is the output information obtained by taking the first sequence as the input of the reconstruction model or is calculated based on the output information of the reconstruction model; receiving second gradient information, where the second gradient information is the input-side gradient information of the compression model, and the second gradient information is determined based on the first gradient information and the compression model; updating the model parameters of the reconstruction model based on the first error information, and updating the model parameters of the dimension reduction model based on the second gradient information.

[0033] The current model training method only supports the training of two models at both ends, while the above technical solution can support the joint training of multiple models at both ends. Based on the method of the above technical solution, the dimension reduction model, the compression model, and the reconstruction model can be jointly trained and optimized, and the effect of reducing the channel information feedback overhead can be achieved.

[0034] In some implementation manners of the fifth aspect, when the termination condition of model training is satisfied, the model training is terminated. For example, the termination condition of model training can be: the error indicated by the first error information is less than or equal to a first threshold, or the number of model training rounds is equal to a second threshold.

[0035] In a sixth aspect, a method for model training is provided. This method can be executed by a second device, or can also be executed by components (such as chips or circuits) of the second device, and this is not limited. For example, the second device can be a terminal device or an AI network element #2, such as a third-party network element #2.

[0036] The method includes: receiving second downlink channel information, which is determined based on first downlink channel information and a dimension reduction weight value. The first downlink channel information is the true downlink channel information of N physical antennas, and the dimension reduction weight value is the output information obtained by taking channel prior information as the input of a dimension reduction model. The dimension reduction weight value is used to perform digital precoding on M reference signals corresponding to M antenna ports to measure the equivalent downlink channel of N physical antennas. N is a positive integer, and M is a positive integer less than N. The first downlink channel information and the channel prior information are a training data set for model training; sending a first sequence, where the first sequence is the output information obtained by taking the information of the second downlink channel as the input of a compression model; receiving first gradient information, which is the input-side gradient information of a reconstruction model. The first gradient information is used to update the model parameters of the compression model and is determined based on first error information and the reconstruction model. The first error information is determined based on third downlink channel information and the first downlink channel information. The third downlink channel information is the output information obtained by taking the first sequence as the input of the reconstruction model or is calculated based on the output information of the reconstruction model. The first error information is used to update the model parameters of the reconstruction model; sending second gradient information, which is the input-side gradient information of the compression model and is determined based on the first gradient information and the compression model; updating the model parameters of the compression model based on the first gradient information.

[0037] For the beneficial effects of the sixth aspect, refer to the description of the fifth aspect and will not be elaborated here.

[0038] In some implementation manners of the fifth aspect or the sixth aspect, the channel prior information is determined based on at least one of the following information: currently measured uplink channel information, historical uplink channel information, historical downlink channel information, or sensed channel information.

[0039] In some implementation manners of the fifth aspect or the sixth aspect, the first downlink channel information is the true downlink channel response of N physical antennas.

[0040] In some implementation manners of the fifth aspect or the sixth aspect, the second downlink channel information indicates a downlink channel response, which is determined based on the first downlink channel information and the dimension reduction weight value, or the second downlink channel information indicates the weight value corresponding to the downlink channel response.

[0041] In some implementation manners of the fifth aspect or the sixth aspect, the third downlink channel information is the weight value of the reconstructed N physical antennas, and the first error information indicates the error between the third downlink channel information and the true weight value of the N physical antennas. The true weight value of the N physical antennas is determined based on the first downlink channel information.

[0042] In a seventh aspect, a method for model training is provided. This method can be executed by a first device, or alternatively, by a component (such as a chip or a circuit) of the first device, without limitation in this regard. By way of example, the first device can be a network device, a terminal device, or an AI network element #1, such as a third-party network element #1.

[0043] The method includes: obtaining a first training data set, where the first training data set includes channel prior information and first downlink channel information, and the first downlink channel information is the true downlink channel information of N physical antennas, and N is a positive integer; jointly training a dimensionality reduction model, a compression model, and a reconstruction model. Among them, the input of the dimensionality reduction model is the channel prior information, and the output is a dimensionality reduction weight value, where the dimensionality reduction weight value is used for digital precoding of M reference signals corresponding to M antenna ports to measure the equivalent downlink channel of N physical antennas, and M is a positive integer less than N. The input of the compression model is second downlink channel information, and the output is a first sequence, where the second downlink channel information is determined based on the first downlink channel information and the dimensionality reduction weight value. The input of the reconstruction model is the first sequence, and the output is third downlink channel information; updating the network device parameters of the dimensionality reduction model, the compression model, and the reconstruction model based on first error information, where the first error information is determined based on the third downlink channel information and the first downlink channel information.

[0044] In the above technical solution, the difference from the joint training methods in the fifth and sixth aspects is that in this method, three models are jointly trained on one side (i.e., the first device) side.

[0045] In some implementation manners of the seventh aspect, when the termination condition for model training is met, the model training is terminated. By way of example, the termination condition for model training can be: the error indicated by the first error information is less than or equal to a first threshold, or the number of model training rounds is equal to a second threshold.

[0046] In an eighth aspect, a method for model training is provided. This method can be executed by a second device, or alternatively, by a component (such as a chip or a circuit) of the second device, without limitation in this regard. Among them, the second device can be a device different from the first device. By way of example, the second device can be a network device, a terminal device, or an AI network element #2, such as a third-party network element #2.

[0047] The method includes: receiving a second training data set, where the second training data set includes the output information of the dimensionality reduction model and the output information of the compression model obtained by using the first training data set again after the dimensionality reduction model, the compression model, and the reconstruction model complete joint training based on the first training data set. The first training data set includes channel prior information and first downlink channel information. In one round of joint training corresponding to the dimensionality reduction model, the compression model, and the reconstruction model, the input of the dimensionality reduction model is the channel prior information, and the output is the dimensionality reduction weight value, where the dimensionality reduction weight value is used for digital precoding of M reference signals corresponding to M antenna ports to measure the equivalent downlink channel of N physical antennas, and M is a positive integer less than N. The input of the compression model is the second downlink channel information, and the output is a first sequence, where the second downlink channel information is determined based on the first downlink channel information and the dimensionality reduction weight value. The input of the reconstruction model is the first sequence, and the output is the third downlink channel information; training a first model based on the second training data set to obtain a second model.

[0048] In some implementations of the seventh aspect or the eighth aspect, the method further includes: sending the compression model corresponding to after completing model training; or, sending the second training data set, where the second training data set includes the output information of the dimensionality reduction model and the output information of the compression model obtained by using the first training data set again after completing model training.

[0049] In the above technical solution, if model A among the three models will be deployed on another device side (for example, the second device side), the first device may send the trained model A or the training data set corresponding to model A to the second device, and the second device completes model deployment or performs secondary training based on the received model A or training data set.

[0050] In some implementations of the seventh aspect or the eighth aspect, the channel prior information is determined based on at least one of the following information: currently measured uplink channel information, historical uplink channel information, historical downlink channel information, and sensed channel information.

[0051] In some implementations of the seventh aspect or the eighth aspect, the first downlink channel information is the true downlink channel response of N physical antennas.

[0052] In some implementations of the seventh aspect or the eighth aspect, the second downlink channel information indicates the downlink channel response, where the downlink channel response is determined based on the first downlink channel information and the dimensionality reduction weight value, or the second downlink channel information indicates the weight value corresponding to the downlink channel response.

[0053] In certain implementations of the seventh or eighth aspect, the third downlink channel information is the weights of the reconstructed N physical antennas, the first error information indicates the error between the third downlink channel information and the true weights of the N physical antennas, and the true weights of the N physical antennas are determined based on the first downlink channel information.

[0054] In a ninth aspect, a communication device is provided, which is used to execute the method provided in the first aspect above. Specifically, the device may include units and / or modules for executing the method in any aspect or any possible implementation manner in the first aspect, such as a processing unit and / or a communication unit.

[0055] In one implementation, the device is a network device. When the device is a network device, the communication unit may be a transceiver circuit; the processing unit may be a processing circuit.

[0056] Exemplarily, the transceiver circuit may be a transceiver, an input / output interface, or an input / output circuit.

[0057] Exemplarily, the processing circuit may be one or more processors, or may also be all or part of the circuits in one or more processors.

[0058] In another implementation, the device is a chip, a chip system, or a circuit used in a network device. When the device is a chip, a chip system, or a circuit used in a terminal device, the communication unit may be the transceiver circuit on the chip, the chip system, or the circuit; the processing unit may be a processing circuit.

[0059] Exemplarily, the transceiver circuit may be a transceiver, an input / output interface, an input / output circuit, a pin, or a related circuit.

[0060] Exemplarily, the processing circuit may be a logic circuit or a processor.

[0061] In a tenth aspect, a communication device is provided, which is used to execute the method provided in the second aspect above. Specifically, the device may include units and / or modules for executing the method in any aspect or any possible implementation manner in the second aspect, such as a processing unit and / or a communication unit.

[0062] In one implementation, the device is a terminal device. When the device is a terminal device, the communication unit may be a transceiver circuit; the processing unit may be a processing circuit.

[0063] In another implementation, the device is a chip, a chip system, or a circuit used in a terminal device. When the device is a chip, a chip system, or a circuit used in a terminal device, the communication unit may be the transceiver circuit on the chip, the chip system, or the circuit; the processing unit may be a processing circuit.

[0064] For examples of the transceiver circuit and the processing circuit, refer to the description of the ninth aspect, which will not be elaborated here.

[0065] The eleventh aspect provides a communication device for performing the method provided in the third aspect or the fifth aspect or the seventh aspect above. Specifically, the device may include units and / or modules for performing the method in any one of the third aspect or the fifth aspect or the seventh aspect or any possible implementation manner in the third aspect or the fifth aspect or the seventh aspect, such as a processing unit and / or a communication unit.

[0066] In one implementation manner, the device is a first device. When the device is a first device, the communication unit may be a transceiver circuit; the processing unit may be a processing circuit.

[0067] In another implementation manner, the device is a chip, a chip system or a circuit for the first device. When the device is a chip, a chip system or a circuit for the first device, the communication unit may be a transceiver circuit on the chip, the chip system or the circuit; the processing unit may be a processing circuit.

[0068] For examples of the transceiver circuit and the processing circuit, refer to the description of the ninth aspect, which will not be elaborated here.

[0069] The twelfth aspect provides a communication device for performing the method provided in the fourth aspect or the sixth aspect or the eighth aspect above. Specifically, the device may include units and / or modules for performing the method in any one of the fourth aspect or the sixth aspect or the eighth aspect or any possible implementation manner in the fourth aspect or the sixth aspect or the eighth aspect, such as a processing unit and / or a communication unit.

[0070] In one implementation manner, the device is a second device. When the device is a second device, the communication unit may be a transceiver circuit; the processing unit may be a processing circuit.

[0071] In another implementation manner, the device is a chip, a chip system or a circuit for the second device. When the device is a chip, a chip system or a circuit for the terminal device, the communication unit may be a transceiver circuit on the chip, the chip system or the circuit; the processing unit may be a processing circuit.

[0072] For examples of the transceiver circuit and the processing circuit, refer to the description of the ninth aspect, which will not be elaborated here.

[0073] In a thirteenth aspect, a communication device is provided. The device includes at least one processing circuit coupled to at least one memory. The at least one memory is configured to store computer programs or instructions, and the at least one processing circuit is configured to call and run the computer programs or instructions from the at least one memory, such that the communication device executes the method in any one of the possible implementations of the first aspect, the third aspect, the fifth aspect, the seventh aspect, or any one of the possible implementations of the first aspect, the third aspect, the fifth aspect, the seventh aspect.

[0074] In a fourteenth aspect, a communication device is provided. The device includes at least one processing circuit coupled to at least one memory. The at least one memory is configured to store computer programs or instructions, and the at least one processing circuit is configured to call and run the computer programs or instructions from the at least one memory, such that the communication device executes the method in any one of the second aspect, the fourth aspect, the sixth aspect, the eighth aspect, and any one of the possible implementations of the second aspect, the fourth aspect, the sixth aspect, the eighth aspect.

[0075] In a fifteenth aspect, a processing circuit is provided for executing the methods provided in the above aspects.

[0076] For operations such as sending and obtaining / receiving involved in the processing circuit, if there is no special description, or if it does not conflict with its actual role or internal logic in the relevant description, it can be understood as operations such as outputting, receiving, and inputting by the processing circuit, or it can also be understood as sending and receiving operations performed by the radio frequency circuit and the antenna. This application does not make any limitations in this regard.

[0077] In a sixteenth aspect, a computer-readable storage medium is provided. The computer-readable storage medium stores program code for a device to execute. The program code includes the method for executing any one of the first aspect to the eighth aspect or any one of the possible implementations of the first aspect to the eighth aspect.

[0078] In a seventeenth aspect, a computer program product including instructions is provided. When the computer program product runs on a computer, it causes the computer to execute the method in any one of the first aspect to the eighth aspect and any one of the possible implementations of the first aspect to the eighth aspect.

[0079] In an eighteenth aspect, a chip is provided. The chip includes a processing circuit and a communication interface. The processing circuit reads instructions stored on a memory through the communication interface and executes the method in any one of the first aspect to the eighth aspect or any one of the possible implementations of the first aspect to the sixth aspect.

[0080] Optionally, as an implementation, the chip further includes a memory in which computer programs or instructions are stored. The processing circuit is configured to execute the computer programs or instructions stored on the memory. When the computer programs or instructions are executed, the processing circuit is configured to execute the methods in any one of the first aspect to the eighth aspect or any possible implementation manner of the first aspect to the eighth aspect.

[0081] Exemplarily, the processing circuit described in the above aspects may be one or more processors, or may also be all or part of the circuits in one or more processors.

[0082] In a nineteenth aspect, a communication system is provided, which includes the communication devices shown in the thirteenth aspect and the fourteenth aspect. Description of the Drawings

[0083] Figure 1 It is a schematic diagram of a possible application framework in the communication system.

[0084] Figure 2 It is a schematic diagram of a possible application framework in the communication system.

[0085] Figure 3 It is a schematic diagram of a communication system applicable to the embodiments of the present application.

[0086] Figure 4 It is a schematic diagram of another communication system applicable to the embodiments of the present application.

[0087] Figure 5 It is a schematic block diagram of an autoencoder.

[0088] Figure 6 It is a schematic diagram of the forward propagation and backward propagation of a neural network.

[0089] Figure 7 It is a schematic diagram of a CSI feedback process.

[0090] Figure 8 It is a schematic flowchart of a communication method 800 proposed by the present application.

[0091] Figure 9 It is a schematic flowchart of a model training method 900 proposed by the present application.

[0092] Figure 10 It is a schematic flowchart of a model training method 1000 proposed by the present application.

[0093] Figure 11 It is a comparison chart of the simulation performance corresponding to the model dimensionality reduction and compression method provided by the present application and the dimensionality reduction and compression method of the existing orthogonal codebook.

[0094] Figure 12 It is a schematic block diagram of the communication device 1200 provided by an embodiment of the present application.

[0095] Figure 13 It is a schematic block diagram of the communication device 1300 provided by an embodiment of the present application. Detailed implementation manners

[0096] Next, the technical solutions in the embodiments of the present application will be described with reference to the accompanying drawings.

[0097] Before introducing the embodiments of the present application, the following points are first explained.

[0098] 1. In the description of the embodiments of the present application, unless otherwise specified, the meaning of "a plurality" is two or more.

[0099] 2. In each embodiment of the present application, if there is no special description and logical conflict, the terms and / or descriptions between different embodiments are consistent and can be referenced to each other. The technical features in different embodiments can be combined to form new embodiments according to their internal logical relationships.

[0100] 3. The various numerical numbers involved in the present application are only for the convenience of description and do not limit the scope of the present application. The size of the serial numbers involved in the present application does not mean the order of execution. The execution order of each process should be determined by its function and internal logic. For example, the terms "first", "second", "third", "fourth" and other various term numbers in the specification, claims and drawings of the present application (if any) are used to distinguish similar objects and do not limit the size, content, order, timing, priority or importance of multiple objects. For example, the first information and the second information do not represent differences in the amount of information, content, priority or importance, etc.

[0101] 4. The terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily limit to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0102] 5. In each embodiment of the present application, "network element A sends information A to network element B" can be understood as that the destination of information A or the intermediate network element in the transmission path between the destination and the source is network element B, which may include sending information to network element B directly or indirectly. "Network element B receives information A from network element A" can be understood as that the source of information A or the intermediate network element in the transmission path between the source and the destination is network element A, which may include receiving information from network element A directly or indirectly. Necessary processing may be performed on the information between the source and the destination of the information transmission, such as format conversion, etc., but the destination can understand the valid information from the source. Similar expressions in the present application can be understood similarly and will not be elaborated here.

[0103] In other words, the sending and receiving can be carried out between devices, for example, between terminal device #1 and terminal device #2, or can be carried out within a device, for example, sending or receiving between components, modules, chips, software modules or hardware modules within a device through a bus, a trace or an interface.

[0104] 6. In the embodiments of the present application, the indication includes direct indication (also known as explicit indication) and implicit indication. Among them, directly indicating information A means including the information A; implicitly indicating information A means indicating information A through the correspondence between information A and information B and directly indicating information B. Among them, the correspondence between information A and information B can be predefined, pre-stored, pre-burned, or pre-configured.

[0105] 7. In the embodiments of the present application, information C is used for the determination of information D, which includes both the case where information D is determined only based on information C and the case where it is determined based on information C and other information. In addition, for the case where information C is used for the determination of information D, there may also be an indirect determination situation, for example, information D is determined based on information E, and information E is determined based on information C.

[0106] 8. The "storage" or "saving" involved in the embodiments of the present application may refer to saving in one or more memories. The one or more memories may be set separately, or may be integrated in an encoder or a decoder, a processor, or a communication device. The one or more memories may also be partly set separately and partly integrated in a decoder, a processor, or a communication device. The type of the memory may be any form of storage medium, which is not limited in the present application.

[0107] 9. The "protocol" involved in the embodiments of the present application may refer to the standard protocols in the communication field. For example, it may include the fourth-generation (4G) network / fifth-generation (5G) network protocols, the new radio (NR) protocol, and the relevant protocols applied to future communication systems. The present application does not make any limitations in this regard.

[0108] 10. The arrows or boxes shown as dotted lines in the schematic diagrams in the attached drawings of the specification of the present application represent optional steps or optional modules.

[0109] The technical solutions provided by the present application can be applied to various communication systems. For example: fifth-generation (5G) or new radio (NR) systems, long-term evolution (LTE) systems, LTE frequency division duplex (FDD) systems, LTE time division duplex (TDD) systems, wireless local area network (WLAN) systems, satellite communication systems, future communication systems, such as sixth-generation (6G) mobile communication systems, or integrated systems of multiple systems, etc. The technical solutions provided by the present application can also be applied to device-to-device (D2D) communication, vehicle-to-everything (V2X) communication, machine-to-machine (M2M) communication, machine type communication (MTC), and Internet of Things (IoT) communication systems or other communication systems.

[0110] The technical solutions provided by this application can be applied to various communication systems, such as: the fifth-generation (5G) or new radio (NR) system, the long-term evolution (LTE) system, the LTE frequency division duplex (FDD) system, the LTE time division duplex (TDD) system, the wireless local area network (WLAN) system, the satellite communication system, future communication systems, such as the sixth-generation (6G) mobile communication system, or a fusion system of multiple systems, etc. The technical solutions provided by this application can also be applied to device-to-device (D2D) communication, vehicle-to-everything (V2X) communication, machine-to-machine (M2M) communication, machine type communication (MTC), and the Internet of Things (IoT) communication system or other communication systems.

[0111] A device in a communication system can send a signal to another device or receive a signal from another device. The signal can include information, signaling, data, etc. Herein, the device can also be replaced with an entity, a network entity, a network element, a communication device, a communication module, a node, a communication node, etc. In this disclosure, the description is made by taking the device as an example. For example, a communication system can include at least one terminal device and at least one network device. The network device can send a downlink signal to the terminal device, and / or the terminal device can send an uplink signal to the network device. It can be understood that the terminal device in this application can be replaced with a first device, and the network device can be replaced with a second device, and the two execute the corresponding communication methods in this disclosure.

[0112] In the embodiments of this application, the terminal device can also be referred to as a user equipment (UE), an access terminal, a user unit, a user station, a mobile station, a mobile platform, a remote station, a remote terminal, a mobile device, a user terminal, a terminal, a wireless communication device, a user agent, or a user device.

[0113] The terminal device can be a device that provides voice / data. For example, it can be a handheld device with wireless connection function, a vehicle-mounted device, etc. Currently, some examples of terminals are: mobile phone, tablet computer, laptop computer, palmtop computer, mobile internet device (MID), wearable device, virtual reality (VR) device, augmented reality (AR) device, wireless terminal in industrial control, wireless terminal in self-driving, wireless terminal in remote medical surgery, wireless terminal in smart grid, wireless terminal in transportation safety, wireless terminal in smart city, wireless terminal in smart home, cellular phone, cordless phone, session initiation protocol (SIP) phone, wireless local loop (WLL) station, personal digital assistant (PDA), handheld device with wireless communication function, computing device or other processing device connected to a wireless modem, wearable device, terminal device in a 5G network or terminal device in a future evolved public land mobile network (PLMN), etc. The embodiments of the present application are not limited thereto.

[0114] By way of example and not limitation, in the embodiments of the present application, the terminal device can also be a wearable device. A wearable device can also be referred to as a wearable intelligent device, which is a general term for devices developed by applying wearable technology to the intelligent design of daily wear, such as glasses, gloves, watches, clothing, shoes, etc. A wearable device is a portable device that is either directly worn on the body or integrated into the user's clothes or accessories. A wearable device is not just a hardware device, but more importantly, it realizes powerful functions through software support, data interaction, and cloud interaction. Broadly speaking, wearable intelligent devices include those with complete functions and large sizes that can achieve complete or partial functions without relying on a smartphone, such as smart watches or smart glasses, etc., and those that only focus on a certain type of application function and need to cooperate with other devices such as smartphones, such as various smart bracelets and smart jewelry for physical sign monitoring.

[0115] In the embodiments of the present application, the device for implementing the functions of the terminal device may be the terminal device itself, or a device capable of supporting the terminal device to implement such functions, such as a chip system. This device may be installed in the terminal device or used in combination with the terminal device. In the embodiments of the present application, the chip system may be composed of chips, or may include chips and other discrete devices. In the embodiments of the present application, only the case where the device for implementing the functions of the terminal device is the terminal device is used as an example for illustration, which does not limit the solutions of the embodiments of the present application.

[0116] The network device in the embodiments of the present application can be a device for communicating with a terminal device. This network device can also be referred to as an access network device or a radio access network device. For example, the network device can be a base station. The network device in the embodiments of the present application can refer to a radio access network (RAN) node (or device) that connects a terminal device to a wireless network. The base station can generally cover various names as follows, or be replaced with the following names, such as: Node B, evolved Node B (eNB), next generation Node B (gNB), relay station, access point, transmitting and receiving point (TRP), transmitting point (TP), master station, secondary station, multi-mode radio (MSR) node, home base station, network controller, access node, wireless node, access point (AP), transmission node, transceiver node, baseband unit (BBU), remote radio unit (RRU), active antenna unit (AAU), remote radio head (RRH), central unit (CU), distributed unit (DU), radio unit (RU), positioning node, etc. The base station can be a macro base station, micro base station, relay node, donor node or the like, or a combination thereof. The base station can also refer to a communication module, a modem or a chip used in the foregoing devices or apparatuses. The base station can also be a mobile switching center and a device that undertakes the function of a base station in D2D, V2X, M2M communications, a network-side device in a 6G network, a device that undertakes the function of a base station in a future communication system, etc. The base station can support networks with the same or different access technologies. Optionally, the RAN node can also be a server, a wearable device, a vehicle or an in-vehicle device, etc. For example, the access network device in vehicle to everything (V2X) technology can be a road side unit (RSU). The embodiments of the present application do not limit the specific technologies and specific device forms adopted by the network device.

[0117] In some deployments, the network device mentioned in the embodiments of the present application may be a device including a CU, or a DU, or a device including a CU and a DU, or a control plane CU node (Central Unit-Control Plane, CU-CP) and a user plane CU node (Central Unit-User Plane, CU-UP) and a DU node. For example, the network device may include a gNB-CU-CP, a gNB-CU-UP, and a gNB-DU.

[0118] In some deployments, multiple RAN nodes cooperate to assist the terminal in achieving wireless access, and different RAN nodes respectively implement some functions of the base station. For example, the RAN node may be a CU, a DU, a CU-CP, a CU-UP, or an RU, etc. The CU and the DU may be separately provided, or may also be included in the same network element, such as a BBU. The RU may be included in a radio frequency device or a radio frequency unit, such as included in an RRU, an AAU, or an RRH.

[0119] The RAN node may support one or more types of fronthaul interfaces. Different fronthaul interfaces respectively correspond to DUs and RUs with different functions. If the fronthaul interface between the DU and the RU is a Common Public Radio Interface (CPRI), the DU is configured to implement one or more of the baseband functions, and the RU is configured to implement one or more of the radio frequency functions. If the fronthaul interface between the DU and the RU is another interface, compared with the CPRI, some of the downlink and / or uplink baseband functions, for example, for the downlink, one or more of precoding, digital beamforming (BF), or inverse fast Fourier transform (IFFT) / cyclic prefix (CP) addition, are moved from the DU to the RU for implementation. For the uplink, one or more of digital beamforming (BF), or fast Fourier transform (FFT) / CP removal, are moved from the DU to the RU for implementation. In a possible implementation manner, this interface may be an Enhanced Common Public Radio Interface (eCPRI). In the eCPRI architecture, the splitting method between the DU and the RU is different, corresponding to different categories (Cat) of eCPRI, such as eCPRI Cat A, B, C, D, E, F.

[0120] Taking eCPRI Cat A as an example, for downlink transmission, with layer mapping as the division, the DU is configured to implement layer mapping and one or more functions before it (i.e., one or more of encoding, rate matching, scrambling, modulation, layer mapping), while other functions after layer mapping (e.g., one or more of resource element (RE) mapping, BF, or IFFT / adding CP) are moved to the RU for implementation. For uplink transmission, with de-RE mapping as the division, the DU is configured to implement demapping and one or more functions before it (i.e., one or more of decoding, derate matching, descrambling, demodulation, inverse discrete Fourier transform (IDFT), channel equalization, de-RE mapping), while other functions after demapping (e.g., one or more of digital BF or FFT / removing CP) are moved to the RU for implementation. It can be understood that for the function descriptions of the DU and RU corresponding to various types of eCPRI, reference can be made to the eCPRI protocol and will not be elaborated here.

[0121] In a possible design, the processing unit in the BBU for implementing baseband functions is called the baseband high (BBH) unit, and the processing unit in the RRU / AAU / RRH for implementing baseband functions is called the baseband low (BBL) unit.

[0122] In different systems, the CU (or CU-CP and CU-UP), DU, or RU may also have different names, but those skilled in the art can understand their meanings. For example, in an open-radio access network (O-RAN / ORAN) system, the CU can also be called the O-CU (open CU), the DU can also be called the O-DU, the CU-CP can also be called the O-CU-CP, the CU-UP can also be called the O-CU-UP, and the RU can also be called the O-RU. Any unit among the CU (or CU-CP, CU-UP), DU, and RU in this application can be implemented by a software module, a hardware module, or a combination of a software module and a hardware module.

[0123] In the embodiments of this application, the device for implementing the functions of the network device can be the network device; it can also be a device capable of supporting the network device to implement this function, such as a chip system, a hardware circuit, a software module, or a combination of a hardware circuit and a software module. This device can be installed in the network device or used in matching with the network device. In the embodiments of this application, only the case where the device for implementing the functions of the network device is the network device is taken as an example for illustration, which does not limit the solutions of the embodiments of this application.

[0124] Network devices and / or terminal devices can be deployed on land, including indoors or outdoors, handheld or vehicle-mounted; they can also be deployed on water; or they can be deployed on airplanes, balloons, and satellites in the air. In the embodiments of the present application, the scenarios where the network devices and terminal devices are located are not limited. In addition, the terminal devices and network devices can be hardware devices, or software functions running on dedicated hardware or software functions running on general hardware. For example, they are virtualized functions instantiated on a platform (such as a cloud platform), or entities including dedicated or general hardware devices and software functions. The present application does not limit the specific forms of the terminal devices and network devices.

[0125] In a wireless communication network, such as in a mobile communication network, the services supported by the network are becoming increasingly diverse, so the requirements to be met are becoming increasingly diverse. For example, the network needs to be able to support ultra-high speeds, ultra-low latency, and / or ultra-large connections. This feature makes network planning, network configuration, and / or resource scheduling increasingly complex. In addition, due to the increasing power of the network, such as supporting higher and higher frequencies, supporting high-order multiple input multiple output (MIMO) technology, supporting beamforming, and / or supporting new technologies such as beam management, network energy saving has become a popular research topic. These new requirements, new scenarios, and new features have brought unprecedented challenges to network planning, operation and maintenance, and efficient operation. To meet this challenge, artificial intelligence technology can be introduced into the wireless communication network to achieve network intelligence.

[0126] To support AI technology in the wireless network, AI nodes may also be introduced into the network.

[0127] Optionally, the AI nodes can be deployed at one or more of the following positions in the communication system: access network devices, terminal devices, or core network devices, etc. Or, the AI nodes can also be deployed separately. For example, they can be deployed at a position outside any of the above devices, such as in the host of an over the top (OTT) system or a cloud server. The AI nodes can communicate with other devices in the communication system. Other devices can be, for example, one or more of the following: network devices, terminal devices, or network elements of the core network.

[0128] It can be understood that the present application does not limit the number of AI nodes. For example, when there are multiple AI nodes, the multiple AI nodes can be divided based on functions, such as different AI nodes being responsible for different functions.

[0129] It can also be understood that the AI nodes can be independent devices respectively, or can be integrated into the same device to implement different functions, or can be network elements in a hardware device, or can be software functions running on dedicated hardware, or can be virtualized functions instantiated on a platform (such as a cloud platform). The present application does not limit the specific form of the above AI nodes. Among them, the AI nodes can be AI network elements, AI entities or AI modules.

[0130] Figure 1 It is a schematic diagram of a possible application framework in a communication system. As Figure 1 shown, the network elements in the communication system are connected through interfaces (such as NG, Xn), or through the air interface. One or more AI modules are provided in one or more of these network element nodes, such as core network devices, access network nodes (RAN nodes), terminals, or OAM. (For clarity, Figure 1 only 1 is shown). The access network node can be a separate RAN node, or can include multiple RAN nodes. For example, it includes a CU and a DU. One or more AI modules can also be provided in the CU and / or the DU. Optionally, the CU can also be split into a CU-CP and a CU-UP. One or more AI models are provided in the CU-CP and / or the CU-UP.

[0131] The AI module is used to implement the corresponding AI function. The AI modules deployed in different network elements can be the same or different. According to different parameter configurations of the model of the AI module, the AI module can implement different functions. The model of the AI module can be configured based on one or more of the following parameters: structural parameters (such as at least one of the number of neural network layers, the width of the neural network, the connection relationship between layers, the weights of neurons, the activation function of neurons, or the bias in the activation function), input parameters (such as the type and / or dimension of the input parameters), or output parameters (such as the type and / or dimension of the output parameters). Among them, the bias in the activation function can also be referred to as the bias of the neural network.

[0132] An AI module can have one or more models. One model can infer an output, and the output includes one parameter or multiple parameters. The learning process, training process, or inference process of different models can be deployed in different nodes or devices, or can be deployed in the same node or device.

[0133] Figure 2 It is a schematic diagram of a possible application framework in a communication system. As Figure 2 shown, the communication system includes a RAN intelligent controller (RIC). For example, the RIC can be Figure 1The AI modules 117 and 118 shown are used to implement AI-related functions. The RIC includes a near-real time RIC and a non-real time RIC. Among them, the non-real time RIC mainly processes non-real time information, such as data that is not sensitive to latency, and the latency of this data can be at the second level. The real time RIC mainly processes near-real time information, such as data that is relatively sensitive to latency, and the latency of this data is in the order of dozens of milliseconds.

[0134] The near-real time RIC is used for model training and inference. For example, it is used to train an AI model and perform inference using this AI model. The near-real time RIC can obtain network-side and / or terminal-side information from RAN nodes (such as CU, CU-CP, CU-UP, DU, and / or RU) and / or terminals. This information can be used as training data or inference data. Optionally, the near-real time RIC can submit the inference result to the RAN node and / or the terminal. Optionally, the inference result can be exchanged between the CU and the DU, and / or between the DU and the RU. For example, the near-real time RIC submits the inference result to the DU, and the DU sends it to the RU.

[0135] The non-real time RIC is also used for model training and inference. For example, it is used to train an AI model and perform inference using this model. The non-real time RIC can obtain network-side and / or terminal-side information from RAN nodes (such as CU, CU-CP, CU-UP, DU, and / or RU) and / or terminals. This information can be used as training data or inference data, and the inference result can be submitted to the RAN node and / or the terminal. Optionally, the inference result can be exchanged between the CU and the DU, and / or between the DU and the RU. For example, the non-real time RIC submits the inference result to the DU, and the DU sends it to the RU.

[0136] The near-real time RIC and the non-real time RIC can also be separately set as a network element. Optionally, the near-real time RIC and the non-real time RIC can also be part of other devices. For example, the near-real time RIC is set in a RAN node (such as in a CU or a DU), and the non-real time RIC is set in an OAM, a cloud server, a core network device, or other network devices.

[0137] Figure 3 is a schematic diagram of a communication system applicable to the communication method of the embodiments of the present application. As Figure 3 shown, the communication system 100 may include at least one network device, such as Figure 3 the network device 110 shown; the communication system 100 may also include at least one terminal device, such as Figure 3The terminal devices 120 and 130 shown. The network device 110 and the terminal devices (such as terminal devices 120 and 130) can communicate via a wireless link. Between the communication devices in this communication system, for example, between the network device 110 and the terminal device 120, communication can be carried out through multi-antenna technology.

[0138] Figure 4 It is a schematic diagram of another communication system applicable to the communication method of the embodiments of the present application. Compared with Figure 3 the communication system 100 shown, Figure 4 the communication system 200 shown further includes an AI network element 140. The AI network element 140 is used to perform AI-related operations, for example, constructing a training data set or training an AI model, etc.

[0139] In a possible implementation manner, the network device 110 can send data related to the training of the AI model to the AI network element 140, and the AI network element 140 constructs a training data set and trains the AI model. For example, the data related to the training of the AI model can include the data reported by the terminal device. The AI network element 140 can send the results of the operations related to the AI model to the network device 110, and forward them to the terminal device through the network device 110. For example, the results of the operations related to the AI model can include at least one of the following: the AI model that has completed training, the evaluation result or test result of the model, etc. Exemplarily, a part of the AI model that has completed training can be deployed on the network device 110, and another part can be deployed on the terminal device. Alternatively, the AI model that has completed training can be deployed on the network device 110. Or, the AI model that has completed training can be deployed on the terminal device.

[0140] It can be understood that Figure 4 only taking the direct connection between the AI network element 140 and the network device 110 as an example for illustration, in other scenarios, the AI network element 140 can also be connected to the terminal device. Or, the AI network element 140 can be connected to both the network device 110 and the terminal device at the same time. Or, the AI network element 140 can also be connected to the network device 110 through a third-party network element. The embodiments of the present application do not limit the connection relationship between the AI network element and other network elements.

[0141] The AI network element 140 can also be set as a module in the network device and / or the terminal device, for example, set in Figure 3 the network device 110 or the terminal device shown.

[0142] It should be noted that Figure 3 and Figure 4 are only simplified schematic diagrams for easy understanding. For example, the communication system can also include other devices, such as wireless relay devices and / or wireless backhaul devices, etc.Figure 3 and Figure 4 are not drawn. In practical applications, the communication system may include multiple network devices or multiple terminal devices. The embodiments of the present application do not limit the number of network devices and terminal devices included in the communication system.

[0143] To facilitate the understanding of the solutions of the embodiments of the present application, the following terms that may be involved in the embodiments of the present application are explained.

[0144] (1) AI model:

[0145] An AI model is an algorithm or computer program that can implement AI functions. The AI model represents the mapping relationship between the input and output of the model. The type of the AI model can be a neural network, a linear regression model, a decision tree model, a support vector machine (SVM), a Bayesian network, a Q-learning model, or other machine learning (ML) models.

[0146] (2) Two-sided model:

[0147] The two-sided model can also be called a bilateral model, a collaborative model, a dual model, or a two-side model, etc. The two-sided model refers to a model composed of multiple sub-models combined together. The multiple sub-models that make up the model need to match each other. The multiple sub-models can be deployed in different nodes.

[0148] The sub-models involved in the embodiments of the present application include an encoder for compressing information and a decoder for restoring the compressed information. The encoder and the decoder are used in combination, and it can be understood that the encoder and the decoder are a set of matching models. An encoder can include one or more AI models, and the decoder that matches the encoder also includes one or more AI models. The number of AI models included in the matching encoder and decoder is the same and they correspond one by one. By way of example, the encoder and the decoder can be deployed in a terminal device and a network device respectively.

[0149] In one possible design, a set of matching encoder and decoder can specifically be two parts of the same auto-encoder (AE). An auto-encoder is an unsupervised learning neural network. Its characteristic is that the input data is used as the labeled data. Therefore, the auto-encoder can also be understood as a self-supervised learning neural network. The auto-encoder can be used for data compression and restoration. The AE model with the encoder and the decoder deployed in different nodes is a typical bilateral model.

[0150] Figure 5 is a schematic block diagram of an auto-encoder. AsFigure 5 As shown, the encoder in the autoencoder can compress (encode) data A to obtain data B; the decoder in the autoencoder can decompress (decode) data B to recover data A. Or it can be understood that the decoder is the inverse operation of the encoder.

[0151] Alternatively, the AI model in the embodiments of this application can be a single-ended model, and this AI model can be deployed on a terminal device or a network device.

[0152] (3) Neural network (NN):

[0153] A neural network is a specific implementation form of AI or machine learning. According to the universal approximation theorem, a neural network can theoretically approximate any continuous function, so that the neural network has the ability to learn any mapping.

[0154] A neural network can be composed of neural units, and a neural unit can refer to an operation unit with xs and an intercept of 1 as inputs. A neural network is a network formed by connecting many of the above single neural units together, that is, the output of one neural unit can be the input of another neural unit. The input of each neural unit can be connected to the local receptive field of the previous layer to extract the features of the local receptive field, and the local receptive field can be a region composed of several neural units.

[0155] Taking the type of the AI model as a neural network as an example, the AI model involved in this disclosure can be a deep neural network (DNN). According to the construction method of the network, DNN can include a feedforward neural network (FNN), a convolutional neural network (CNN), a recurrent neural network (RNN), etc.

[0156] (4) Channel state information (CSI):

[0157] In a communication system (such as an LTE communication system or an NR communication system, etc.), the network device needs to determine resources for scheduling the downlink data channel of the terminal device, modulation and coding scheme (MCS), and precoding and other configurations based on CSI. It can be understood that CSI belongs to a type of channel information and is information that can reflect channel characteristics and channel quality.

[0158] CSI measurement refers to the receiver solving the channel information based on the reference signal sent by the transmitter, that is, estimating the channel information using channel estimation methods. Exemplarily, the reference signal may include one or more of channel state information reference signal (CSI-RS), synchronization signal / physical broadcast channel block (SSB), sounding reference signal (SRS), or demodulation reference signal (DMRS), etc. One or more of CSI-RS, SSB, and DMRS, etc. can be used to measure downlink CSI. SRS and / or DMRS, etc. can be used to measure uplink CSI.

[0159] Taking the FDD communication scenario as an example, in the FDD communication scenario, since the uplink and downlink channels do not have reciprocity or it is impossible to guarantee the reciprocity of the uplink and downlink channels, the network device usually sends downlink reference signals to the terminal device, and the terminal device performs channel measurement and interference measurement based on the received downlink reference signals to estimate the downlink CSI. The terminal device generates a CSI report according to the protocol-predefined method or the method configured by the network device, and feeds it back to the network device so that it can obtain the downlink CSI.

[0160] Exemplarily, CSI may include at least one of the following: channel quality indication (CQI), precoding matrix indicator (PMI), rank indicator (RI), CSI-RS resource indicator (CRI), layer indicator (LI), reference signal receiving power (RSRP), or signal to interference plus noise ratio (SINR), etc. The signal to interference plus noise ratio can also be referred to as the signal-to-interference-and-noise ratio.

[0161] Among them, RI is used to indicate the number of layers of the downlink transmission recommended by the terminal device, CQI is used to indicate the modulation and coding scheme that the terminal device judges can be supported by the current channel conditions, and PMI is used to indicate the precoding recommended by the terminal device. The number of layers of the precoding indicated by PMI corresponds to RI.

[0162] It can be understood that the RI, CQI, PMI, etc. indicated in the above CSI report are only recommended values of the terminal device, and the network device can perform downlink transmission according to some or all of the information indicated in the CSI report. Alternatively, the network device may also not perform downlink transmission according to the information indicated in the CSI report.

[0163] (5) Weight:

[0164] The precoding matrix / vector required for the network device to perform precoding on the network device side during downlink transmission. In this application, the weight is divided into two weights. One weight is the outer layer weight (which can also be referred to as the dimension reduction weight in the following text), and the dimension reduction weight is used for the network device side to perform port dimension reduction during downlink channel information measurement; the other weight is the inner layer weight. After the terminal device measures the equivalent channel after port dimension reduction and completes the compression feedback of the channel information, the network device obtains the weight after restoring the compressed feedback amount. The network device can determine the final precoding matrix during downlink transmission based on the outer layer weight and the inner layer weight. By way of example, the network device calculates the final precoding matrix C used during downlink transmission based on the two-level codebook method. Specifically, the outer layer weight is A and the inner layer weight is B, and the final precoding matrix C is determined by C = A * B, where * represents multiplication.

[0165] It can be understood that the weight and the precoding matrix in this application can be replaced with each other.

[0166] (6) Port dimension reduction:

[0167] As the scale of the base station side array increases, the dimension of the CSI measured and fed back by the terminal side increases, resulting in an increase in the burden of measurement resource overhead and feedback overhead. Therefore, when the network device performs downlink channel information measurement, the network device performs digital precoding on the reference signals of a small number of antenna ports through the outer layer weight, so that the network device can measure the channels under a large number of physical antennas by using a small number of reference channel resources (corresponding to the number of antenna ports). The dimension reduction in this application all refers to port dimension reduction.

[0168] (7) Antenna port:

[0169] The antenna port is a logical concept, and there is no direct correspondence between an antenna port and a physical antenna. The antenna port is usually associated with the reference signal, and its meaning can be understood as a transceiver interface on the channel experienced by the reference signal. For low frequencies, one antenna port may correspond to one or more physical antennas (one physical antenna can refer to a digital port or an antenna element), and these physical antennas jointly transmit the reference signal, and the receiving end can regard them as a whole without distinguishing these physical antennas.

[0170] (8) Design of the AI model:

[0171] The design of the AI model mainly includes a data collection phase (such as collecting training data and / or inference data), a model training phase, and a model inference phase. Further, it may also include an inference result application phase.

[0172] It can be understood that a network element with artificial intelligence capabilities may be included in the communication system. The above-mentioned phases related to the AI model design can be executed by one or more network elements with artificial intelligence capabilities. In a possible design, the AI function (such as an AI module or an AI entity) can be configured in an existing network element in the communication system to implement AI-related operations, such as the training and / or inference of the AI model. For example, the existing network element can be a network device or a terminal device, etc. Or in another possible design, an independent network element can also be introduced in the communication system to execute AI-related operations, such as training the AI model. This independent network element can be called an AI network element (i.e., an AI entity or an AI node), and the embodiments of this application do not limit this name. Exemplarily, this AI network element can be directly connected to the network device in the communication system, or can be indirectly connected to the network device through a third-party network element. Among them, the third-party network element can be a core network element such as an authentication management function (AMF) network element, a user plane function (UPF) network element, an operation administration and maintenance (OAM), a cloud server, or other network elements, without limitation. Exemplarily, this independent AI network element can be deployed on one or more of the network device side, the terminal device side, or the core network side. Optionally, it can be deployed on a server, such as a cloud server, or an over-the-top (OTT) device. Exemplarily, as Figure 4 shown, an AI network element 140 is introduced in the communication system. Optionally, the server or the OTT device can be a third-party network element. In addition, the third-party network element can also be referred to as a third-party device, that is, a device different from the network device or the terminal device.

[0173] (9) Training data set and inference data:

[0174] The training dataset is used for the training of the AI model. The training dataset may include the input of the AI model, or may include the input and target output of the AI model. Among them, the training dataset includes one or more training data, and the training data may include the training samples input to the AI model, or may include the target output of the AI model. Among them, the target output may also be referred to as a label, sample label, or labeled sample. The label is the true value. In the field of machine learning, the true value (ground truth) usually refers to the data that is considered accurate or real data.

[0175] In the field of communication, the training dataset may include simulation data collected through a simulation platform, or may include experimental data collected in an experimental scenario, or may also include measured data collected in an actual communication network. Due to differences in the geographical environment and channel conditions where the data is generated, for example, differences in indoor, outdoor, moving speed, frequency band, or antenna configuration, etc., when obtaining data, the collected data can be classified. For example, data with the same channel propagation environment and antenna configuration can be grouped into one category.

[0176] Model training essentially means learning certain features from the training data. In the process of training an AI model (such as a neural network model), because it is hoped that the output of the AI model is as close as possible to the value that is truly desired to be predicted, the weight vector of each layer of the AI model can be updated by comparing the predicted value of the current network with the truly desired target value and then according to the difference between the two (of course, there is usually an initialization process before the first update, that is, parameters are pre-configured for each layer in the AI model). For example, if the predicted value of the network is high, the weight vector is adjusted to make it predict lower, and continuous adjustment is made until the AI model can predict the truly desired target value or a value very close to the truly desired target value. Therefore, it is necessary to pre-define "how to compare the difference between the predicted value and the target value", which is the loss function or objective function. They are important equations used to measure the difference between the predicted value and the target value. Among them, taking the loss function as an example, the higher the output value (loss) of the loss function, the greater the difference. Then the training of the AI model becomes a process of minimizing this loss as much as possible, making the value of the loss function less than the threshold, or making the value of the loss function meet the target requirements. For example, if the AI model is a neural network, adjusting the model parameters of the neural network includes adjusting at least one of the following parameters: the number of layers of the neural network, width, weights of neurons, or parameters in the activation function of neurons.

[0177] The inference data can be used as the input of the trained AI model for the inference of the AI model. During the model inference process, when the inference data is input into the AI model, the corresponding output can be obtained, which is the inference result.

[0178] (10) Forward Propagation and Backward Propagation:

[0179] Figure 6 It is a schematic diagram of the forward propagation and backward propagation of a neural network. During the training process of an AI model, there are forward propagation and backward propagation of data streams. As Figure 6 shown, the AI module refers to a certain neural network layer or an AI model. Among them, forward propagation means that the neural network passes the intermediate calculation results of the previous neural network layer or the previous AI model to the next neural network layer or the next AI model to complete the intermediate calculation of the final output result under this parameter. Backward propagation is a method for the neural network to calculate gradients. That is, after calculating the error between the final output result and the label (i.e., the true result), in order to calculate the update gradient of the network parameters to be trained, the neural network will calculate the gradients layer by layer from back to front through backward propagation. The principle is similar to the chain rule in calculus, and the error of the latter layer is calculated and propagated to the previous layer, and the previous layer calculates the gradient of this network layer based on the current error.

[0180] When AI technology is introduced into a wireless communication network, a CSI feedback method based on an AI model is generated. The terminal device uses the AI model to compress and feedback the CSI, and the network device uses the AI model to recover the compressed CSI. What is transmitted in the CSI feedback based on the AI model is a sequence (such as a bit sequence). Compared with traditional CSI, the overhead of CSI feedback is lower.

[0181] The following combines Figure 7 to introduce the CSI feedback process based on the AI model after current port dimension reduction. In this process, the number of physical antennas is N, and the number of antenna ports after dimension reduction is M, where M is less than N, and both M and N are positive integers. This process includes the following steps.

[0182] P100: The terminal device sends an uplink reference signal (such as SRS) to measure the uplink channel information, and the network device receives the uplink reference signal to obtain the uplink channel information at this moment.

[0183] P101: The network device receives the uplink channel information and calculates the dimension reduction weight w0 based on the received uplink channel information (where the dimension of w0 is N×M). For example, based on this uplink channel information, that is, the uplink channel response, and the historical uplink channel response, calculate the second-order statistic Z of the uplink channel information, and then calculate the dimension reduction weight w0 based on this second-order statistic Z through the following method. After obtaining the dimension reduction weight w0, the network device performs port dimension reduction based on the dimension reduction weight, that is, digitally precodes the reference signals on M ports using the dimension reduction weight to measure the response h of the equivalent downlink channel under N physical antennas using M downlink reference signals (such as CSI-RS) resources.DL = H DL * w0, where H DL is the response of the true downlink channel under N physical antennas (where the spatial dimension of H DL is N). Two currently used methods for calculating the reduced-dimensional weight are exemplarily given below.

[0184] In one method, the network device calculates the projection energy of the second-order statistic Z under the discrete Fourier transform (DFT) codebook, and selects the first M DFT bases {l0, l1,..., l M-1}(where l i is a 1×N vector) with the maximum energy projection as the reduced-dimensional weight w0.

[0185] In another method, the network device performs a singular value decomposition (SVD) on the second-order statistic Z, and selects the first M eigen-directions {l0, l1,..., l M-1}(where l i is a 1×N vector) as the reduced-dimensional weight w0.

[0186] The starting point of the above methods is to minimize the loss of the energy of the reduced-dimensional equivalent downlink channel compared to the energy of the true downlink channel. Among them, the energy of the equivalent downlink channel is determined based on the response h DL of the equivalent downlink channel, and the energy of the true downlink channel is determined based on the response H DL of the true downlink channel. Exemplarily, the relationship between the energy E and the response H is E = ||H||2, where ||H||2 represents the 2-norm of the matrix H.

[0187] P102: The terminal device measures M downlink reference signals to obtain the corresponding equivalent channel information H DL w0. After the measurement is completed, the terminal device sends a downlink measurement report to the network device, and the report includes CSI.

[0188] The CSI feedback method based on the AI model is that the network device or the terminal device can design and train an autoencoder to replace the codebook in the 3rd generation partnership project (3GPP) R16 protocol for CSI compression in the spatial-frequency double domain. For example, the compression amount between the output of the encoder and the input of the decoder in the autoencoder is the compression amount of the precoding matrix of the equivalent downlink channel, and this compression amount is used to replace the feedback of the precoding matrix in the feedback process to reduce the feedback overhead.

[0189] P103: The network device receives the CSI in the downlink measurement report, obtains the compressed feedback amount of the equivalent channel precoding matrix therefrom, and reconstructs the downlink precoding matrix by recovering the inner layer weights v (i.e., the equivalent channel precoding matrix before compression) through the following method.

[0190] Since the compressed feedback amount is obtained by the terminal device through compression using the encoder in the autoencoder, the network device inputs the compressed feedback amount into the decoder of the autoencoder, and the decoder outputs to obtain the inner layer weights v. Finally, based on the structure of the two-level codebook, the network device obtains the downlink precoding matrix V under the final N physical antennas through the formula V = v * w0.

[0191] Since the above-mentioned reduced-dimensional weights are calculated by means of an orthogonal codebook or orthogonal decomposition, only considering the minimum energy loss of the equivalent channel after dimension reduction, which is independent of the compression scheme, resulting in low compressibility of the measured equivalent channel information, how to further reduce the feedback overhead has become an urgent problem to be solved.

[0192] In view of this, the present application proposes a communication method that can effectively solve the above technical problems. The following is a detailed description of this communication method.

[0193] Figure 8 It is a schematic flowchart of a communication method 800 proposed by the present application. The method includes the following steps.

[0194] S810, the network device performs digital precoding on the M reference signals corresponding to the M antenna ports based on the reduced-dimensional weights. The reduced-dimensional weights are used to measure the equivalent downlink channel of the N physical antennas by using the M reference signals. N is a positive integer, M is a positive integer less than N, and the reduced-dimensional weights are determined based on a dimension reduction model.

[0195] Optionally, the model in the present application is an AI model, a neural network model, a machine learning model, etc., and the present application does not make any restrictions.

[0196] Among them, the network device performs digital precoding on the M reference signals corresponding to the M antenna ports based on the reduced-dimensional weights, that is, the network device realizes port dimension reduction based on the reduced-dimensional weights, and the dimension of the reduced-dimensional weights is N * M. Exemplarily, the reference signal can be CSI-RS or SSB or DMRS, and the present application does not make any restrictions.

[0197] Optionally, the reduced-dimensional weights are the output information obtained by using the channel prior information as the input of the dimension reduction model. Exemplarily, the channel prior information is determined based on at least one of the following information: the currently measured uplink channel information, historical uplink channel information, historical downlink channel information, or, sensed channel information. Exemplarily, the channel prior information can be the currently measured uplink channel information, or it can be calculated based on the previously measured uplink channel information and historical uplink channel information.

[0198] Optionally, the dimensionality reduction model can be located in a network device or an AI network element, such as a third-party network element, as a module. By way of example, the AI network element can be a core network element such as an AMF network element or a UPF network element, or can also be an OAM, a cloud server, an OTT, or other network elements, without limitation.

[0199] In Example 1, if the dimensionality reduction model is located in the network device, the network device can input the channel prior information into the dimensionality reduction model to obtain the dimensionality reduction weights (i.e., the output of the dimensionality reduction model).

[0200] In Example 2, if the dimensionality reduction model is located in an AI network element, such as a third-party network element, the network device can send the channel prior information and the information of the dimensionality reduction model to the AI network element. The AI network element inputs the channel prior information into the dimensionality reduction model to obtain the dimensionality reduction weights. Then, the AI network element sends the dimensionality reduction weights to the network device.

[0201] S820: The network device sends M reference signals to the terminal device. Correspondingly, the terminal device receives the M reference signals from the network device.

[0202] S830: The terminal device obtains first downlink channel information based on the M reference signals.

[0203] It can be understood that this application does not specifically limit the type of information specifically indicated by the first downlink channel information. Some examples are given below.

[0204] By way of example, the first downlink channel information indicates the response of the equivalent downlink channel. If H DL is the true downlink channel response under N physical antennas, the spatial domain dimension is N, and w0 is the dimensionality reduction weight, then the response of the equivalent downlink channel obtained by the network device measuring M reference signals is h DL = H DL * w0, and the spatial domain dimension is M.

[0205] By way of example, the first downlink channel information indicates the precoding matrix (which can also be called the weight) corresponding to the response of the equivalent downlink channel. By way of example, if the downlink channel response obtained by the network device measuring M reference signals is the above h DL , then the weight v0 under the equivalent channel h DL can be obtained through SVD decomposition, and the spatial domain dimension is M.

[0206] S840: The network device sends first indication information, and the first indication information indicates a compression model, and the compression model matches the dimensionality reduction model. Correspondingly, the terminal device receives the first indication information.

[0207] It can be understood that after the terminal device obtains the first downlink channel information, it feeds back the obtained first downlink channel information to the network device. However, considering the feedback overhead, the terminal device does not directly feed back such a high-dimensional tensor to the network device, but sends the compressed feedback amount after compressing the first downlink channel information through a compression model to the network device. Since a dimensionality reduction model is introduced in this method and the dimensionality reduction method corresponding to the dimensionality reduction model is associated with the compression method, the network device indicates a compression model matching the dimensionality reduction model to the terminal device in order to obtain better compression performance.

[0208] Furthermore, in a scenario where dimensionality reduction and compression are not associated (such as a scenario where port dimensionality reduction is not performed, or a scenario where the dimensionality reduction weights are orthogonal), since the terminal device only needs to feed back the precoding matrix of the equivalent downlink channel to complete the reconstruction of the real channel precoding matrix, the input of the compression model is consistent with the output of the reconstruction model. At this time, the compression model is used for compressing the downlink channel information, and the reconstruction model is used for decompressing the output information of the compression model. It is sufficient that the reconstruction model matches the compression model; in a scenario where dimensionality reduction and compression are associated (such as the solution proposed in this application), the dimensionality reduction weights are not restricted to be orthogonal, and the compression method, reconstruction method, and dimensionality reduction method are strongly correlated. The input of the compression model is inconsistent with the output of the reconstruction model. Therefore, the operating principles of the compression model and the reconstruction model are essentially different from those in the scenario where dimensionality reduction and compression are not associated. Thus, in the embodiments of this application, a compression model matching the dimensionality reduction model is indicated through the first indication information, and this compression model is used for feature extraction and compression of the inner-layer weights of the first downlink channel information, which can not only improve the compression performance but also further ensure the decompression performance of the reconstruction model.

[0209] Exemplarily, the dimensionality reduction model, compression model, and reconstruction model in the embodiments of this application are all related to the number of antenna ports M, the number of physical antennas N, and the channel prior information.

[0210] The following gives several possible specific implementation manners of the first indication information.

[0211] Implementation manner 1: The first indication information indicates the identifier (ID) of the compression model.

[0212] Implementation manner 2: The first indication information indicates the dimensionality reduction model ID and / or the reconstruction model ID.

[0213] For example, after the joint training of the compression model, dimensionality reduction model, and reconstruction model is completed, the compression model ID, dimensionality reduction model ID, and reconstruction model ID can be predefined and the three model IDs can be associated. Then, the network device can implicitly indicate the compression model by indicating the dimensionality reduction model ID and / or the reconstruction model ID.

[0214] Implementation manner 3: The first indication information indicates the dimensionality reduction method or the compression method.

[0215] It can be understood that a dimensionality reduction model can be determined based on the dimensionality reduction method, and a compression model can be determined based on the dimensionality reduction model. Therefore, the network device can implicitly indicate the compression model by indicating the dimensionality reduction method.

[0216] Similarly, a compression model can be determined based on the compression method. Therefore, the network device can implicitly indicate the compression model by indicating the compression method.

[0217] It should be noted that the dimensionality reduction model, the compression model, and the reconstruction model are named based on the main functions of each model in this application. It can be understood that this application does not make specific limitations on the names of the three models. The three models may also have other functions besides the functions corresponding to their names. For example, the compression model can also implement a quantization function, and / or the reconstruction model can also implement a dequantization function. This application does not make limitations.

[0218] S850, the terminal device sends a first sequence to the network device. The first sequence is a compression feedback quantity obtained by using the first downlink channel information as the input of the compression model. Correspondingly, the network device receives the first sequence from the terminal device.

[0219] Optionally, the compression model can be located as a module in the terminal device or an AI network element, such as a third-party network element.

[0220] Example 1, if the compression model is located in the terminal device, the terminal device can input the first downlink channel information into the compression model to obtain the first sequence (i.e., the output of the compression model).

[0221] Example 2, if the compression model is located in the AI network element, the terminal device can send the first downlink channel information and the information of the compression model to the AI network element. The AI network element inputs the first downlink channel information into the indicated compression model to obtain the first sequence. Then, the AI network element sends the first sequence to the terminal device.

[0222] S860, the network device obtains second downlink channel information. The second downlink channel information is the information obtained by using the first sequence as the input of the reconstruction model. The reconstruction model matches the dimensionality reduction model. The second downlink channel is used to determine the downlink channel information of N physical antennas.

[0223] Optionally, the compression model can be an encoder. Correspondingly, the reconstruction model can be a decoder.

[0224] Optionally, the reconstruction model can be located as a module in the network device or an AI network element, such as a third-party network element.

[0225] Example 1, if the reconstruction model is located in the network device, the network device can input the first sequence into the reconstruction model to obtain the second downlink channel information (i.e., the output of the reconstruction model).

[0226] In Example 2, if the reconstruction model is located in the AI network element, the network device may send the first downlink channel information and the information of the reconstruction model to the AI network element. The AI network element inputs the first sequence into the reconstruction model to obtain the second downlink channel information. Then, the AI network element sends the second downlink channel information to the network device.

[0227] Optionally, if the dimensionality reduction model, the compression model, and the reconstruction model are located in an AI network element, such as a third-party network element, the third-party network elements where the respective models are located may be the same or different, and this application does not limit this. In a possible implementation manner, the dimensionality reduction model and the reconstruction model may be located in the third-party network element #1, and the compression model may be located in the third-party network element #2. For the inputs and outputs corresponding to the respective models, refer to the above description and will not be elaborated here.

[0228] Optionally, the second downlink channel is used to determine the downlink channel information of N physical antennas. Specifically, it may be that the second downlink channel information is used to determine the downlink precoding matrix V of N physical antennas. The possible second downlink channel information will be illustrated by examples below.

[0229] In Example 1, the input of the reconstruction model (i.e., the first sequence) is the compression amount of the response h of the equivalent downlink channel DL and the output (i.e., the second downlink channel information) may be the channel response h DL '.

[0230] In Example 2, the input of the reconstruction model (i.e., the first sequence) is the compression amount of the precoding matrix v0 corresponding to the response of the equivalent downlink channel, and the output (i.e., the second downlink channel information) is the inner layer weight v.

[0231] Based on the above Example 1 and Example 2, if the network device determines V, for example, in a possible implementation manner, the network device calculates V based on h DL ' in Example 1 and the dimensionality reduction weight w0, or calculates V based on v in Example 2 and the dimensionality reduction weight w0. For example, the network device calculates V through V = v * w0 based on the existing two-level codebook structure (i.e., the structure of the aforementioned outer layer weight and inner layer weight); in another implementation manner, the network device may determine V through Model #1. For example, the network device inputs the dimensionality reduction weight w0 and the second downlink channel information (h' DL or v) into Model #1 to calculate V.

[0232] It can be understood that Model #1 is a model with the function of calculating V. Specifically, input the specific data of the input parameters required for Model #1 to calculate V into Model #1 to obtain the output data of Model #1, that is, obtain the corresponding V. For example, Model #1 is an AI model or a neural network model or a machine learning model, without limitation.

[0233] Example 3. The input to the reconstruction model (i.e., the first sequence) is h in Example 1 DL The compression amount and the reduced-dimensional weight w0, or the input is the compression amount and the reduced-dimensional weight w0 of v0 in Example 2, and the output of the reconstruction model (i.e., the second downlink channel information) is V.

[0234] Optionally, the second downlink channel is used to determine the downlink channel information of N physical antennas. Specifically, it can be: the second downlink channel information is used to determine the downlink channel response H of N physical antennas DL . The following gives examples of possible second downlink channel information.

[0235] Example 1. The input to the reconstruction model (i.e., the first sequence) is the response h of the equivalent downlink channel DL The compression amount, and the output (i.e., the second downlink channel information) can be the channel response h DL '.

[0236] Based on the above Example 1, if the network device needs to determine H DL , then, for example, in a possible implementation, the network device can calculate H based on h DL ' and the reduced-dimensional weight w0, for example, the network device calculates H through H DL = w0 * h DL '; in another implementation, the network device can determine H through model #2 DL , for example, the network device inputs the reduced-dimensional weight w0 and h DL ' to model #2 and calculates H DL . DL '. DL .

[0237] It can be understood that model #2 is a model with the function of calculating H DL . Specifically, the specific data of the input parameters required for model #2 to calculate H DL are input into model #2 to obtain the output data of model #2, that is, the corresponding H DL . For example, model #2 is an AI model or a neural network model or a machine learning model, without limitation.

[0238] Example 2. The input to the reconstruction model (i.e., the first sequence) is h DL ' and the reduced-dimensional weight w0 in Example 1, and the output is (i.e., the second downlink channel information) H.

[0239] It can be seen that Figure 7In the method shown, port dimension reduction is achieved through an orthogonal codebook or orthogonal decomposition. Considering only from the perspective of minimizing the equivalent channel energy loss, such a method does not take into account the compressibility of the channel after dimension reduction, resulting in poor compressibility of the equivalent channel information. In the method proposed in this application, the dimension reduction model, compression model, and reconstruction model are matching models. That is to say, the dimension reduction method and compression method in this application are jointly optimized, and the compressibility of the channel after dimension reduction is considered while achieving port dimension reduction. Therefore, the method proposed in this application can improve the compressibility of the equivalent channel information compared to Figure 7 the method shown, thereby achieving the effect of further reducing the feedback overhead.

[0240] Based on the above description, in order to achieve the above effect of reducing the feedback overhead and ensuring the decoding (also known as decompression) performance, the dimension reduction model, compression model, and reconstruction model are matching models. Therefore, before applying the corresponding models in practice, the three models can be jointly trained to obtain the three matching models. The joint training process of the three models is described in detail below.

[0241] Figure 9 is a schematic flowchart of a model training method 900 proposed in this application. It can be understood that this method 900 can be independently applied or combined with the foregoing Figure 8 corresponding embodiments, which is not limited herein. In this method, the first device and the second device can deploy the model locally, and transfer the intermediate layer calculation results and gradients in real time to complete the training of the bilateral model. Among them, the dimension reduction model and reconstruction model are deployed on the first device side, and the compression model is deployed on the second device side. Optionally, the first device can be a network device or an AI network element #1, such as a third-party network element #1, and the second device can be a terminal device or an AI network element #2, such as a third-party network element #2. This method includes the following steps.

[0242] S910, the first device obtains a training data set, which includes channel prior information and downlink channel information #1, and the downlink channel information #1 is the real downlink channel information of N physical antennas, where N is a positive integer.

[0243] Before jointly training the three models, the network device collects training data. The training data for training the AI model includes training samples and sample labels. Among them, the channel prior information can be regarded as training samples, and the downlink channel information #1 can be regarded as sample labels (i.e., true values). For the description of the channel prior information, refer to the description in S810, which will not be elaborated here.

[0244] For ease of description, in the following steps, the channel prior information is the uplink channel response H UL , and the downlink channel information #1 is the real downlink channel response H of N physical antennas DLFor example, during model training, H UL is used as the input to the dimensionality reduction model, and H DL is used to calculate the input v0 of the compression model (the result after performing the SVD operation on the equivalent downlink channel response H DL w0) and the calculation of the final label V (i.e., the precoding matrix of N physical antennas).

[0245] For example, the network device can obtain the uplink channel response H DL through measurement, and complete the collection of the corresponding downlink channel response H DL through the existing air interface process. However, it should be noted that when collecting the downlink channel response, each downlink channel response H DL sample should be indicated by the network device to correspond to each uplink channel response H UL sample. For example, it is required that H UL and H DL are located in the same or adjacent bandwidths (the frequency domain position interval is less than K RBs), and the time domain interval between the two should be less than t time slots. That is, the channel responses reflected by H UL and H DL have reciprocity that meets the requirements, such as uplink-downlink reciprocity in the time-delay angle domain.

[0246] S920, the first device sends downlink channel information #2 to the second device. The downlink channel information #2 is determined based on the downlink channel information #1 and the dimensionality reduction weight. The dimensionality reduction weight is the output information obtained by using the channel prior information as the input to the dimensionality reduction model. The dimensionality reduction weight is used to perform digital precoding on the M reference signals corresponding to the M antenna ports to measure the equivalent downlink channel of the N physical antennas, where M is a positive integer less than N. Correspondingly, the second device receives the downlink channel information #2 from the first device.

[0247] Optionally, before S920, the method further includes: the first device inputs the uplink channel response H UL into the dimensionality reduction model, and outputs the dimensionality reduction weight w0. Then, the first device determines the downlink channel information #2 (i.e., the equivalent downlink channel information) based on the dimensionality reduction weight w0.

[0248] For example, the downlink channel information #2 is the response H DL w0 of the equivalent downlink channel;

[0249] For example, the downlink channel information #2 is the equivalent channel weight v0 after performing SVD on the response H DL w0 of the equivalent downlink channel.

[0250] S930, the second device sends a first sequence to the first device, where the first sequence is the output information obtained by using the downlink channel information #2 as the input of the compression model. Correspondingly, the first device receives the first sequence from the second device.

[0251] Optionally, before S930, the method further includes: the second device inputs the downlink channel information #2 into the compression model and outputs the first sequence.

[0252] Exemplarily, the downlink channel information #2 is the response H of the equivalent downlink channel DL w0, then the first sequence can be the compressed feedback amount of the response H of the equivalent downlink channel DL w0, or the compressed feedback amount of the equivalent channel weight v0, which is not limited.

[0253] Among them, the downlink channel information #2 and the first sequence can be regarded as the intermediate calculation results of the forward propagation.

[0254] S940, the first device sends first gradient information to the second device, where the first gradient information is the input-side gradient information of the reconstruction model. Among them, the first gradient information is used to update the model parameters of the compression model, and the first gradient information is determined based on the first error information and the reconstruction model. The first error information is determined based on the downlink channel information #3 and the downlink channel information #1. The downlink channel information #3 is the output information obtained by using the first sequence as the input of the reconstruction model or is calculated based on the output information of the reconstruction model. The first error information is used to update the model parameters of the reconstruction model. Correspondingly, the second device receives the first gradient information from the first device.

[0255] Optionally, before S940, the method further includes: the first device inputs the first sequence into the reconstruction model to obtain the output information of the reconstruction model (the output information is the downlink channel information #3 or is used to determine the downlink channel information #3). After that, the first device determines the first error information based on the downlink channel information #1 and the downlink channel information #3. Then, the first device calculates the gradient information (i.e., the first gradient information) of the input side of the reconstruction model through backpropagation based on the first error information and the reconstruction model.

[0256] Among them, the first error information is determined based on the downlink channel information #3 and the downlink channel information #1, which can be understood as follows: the downlink channel information #3 is the actual output information of the reconstruction model in the current training round obtained based on the channel prior information in the training dataset or the information calculated based on the actual output information. For example, the downlink channel information #3 is the channel response (or channel weight) actually obtained during the training process. Correspondingly, the first device can determine the expected channel response (or channel weight) based on the downlink information #1 in the training dataset. Then, the first device can determine the first error information based on the difference between the expected channel response (or channel weight) and the actually obtained (or channel weight).

[0257] The following is an example with specific parameters. For example, if the first sequence is the compressed feedback amount of the equivalent channel weight v0, the first device inputs v0 into the reconstruction model, obtains the output information of the reconstruction model, and determines the downlink weights of N physical antennas based on the output information of the reconstruction model (i.e., the downlink channel information #3). Among them, the output information of the reconstruction AI model can be itself, or the output information of the reconstruction AI model can also be the inner layer weight v. Then, the first device can calculate based on the inner layer weight v and the dimension reduction weight w0 After that, the first device performs SVD on the true downlink channel response H DL (i.e., the downlink channel information #1) to obtain the weights V of N physical antennas of the true value, and calculates the error between V and (i.e., the first error information). Based on the calculated error, the first device calculates the input-side gradient of the reconstruction model (i.e., the first gradient information) through backpropagation and transmits it to the second device.

[0258] S950, the second device sends the second gradient information to the first device. The second gradient information is the input-side gradient information of the compression model, which is determined based on the first gradient information and the compression model. The second gradient information is used to update the model parameters of the dimension reduction model. Correspondingly, the first device receives the second gradient information from the second device.

[0259] Before S950, the method further includes: the second device calculates the input-side gradient information (i.e., the second gradient information) of the compression model through backpropagation based on the first gradient information and the compression model.

[0260] S960, the first device updates the model parameters of the reconstruction model based on the first error information and updates the model parameters of the dimension reduction model based on the second gradient information; the second device updates the model parameters of the compression model based on the first gradient information.

[0261] Exemplarily, the model parameters may include at least one of the following parameters: the weights of neurons, or biases, etc.

[0262] Optionally, the method further includes: S970, when the termination condition for model training is met, the first device determines to terminate the model training.

[0263] Exemplarily, the termination condition for model training may be: the error indicated by the first error information is less than or equal to the first threshold, or the number of model training rounds is equal to the second threshold.

[0264] The above S920 to S960 can be regarded as one round of model training. After each round of model training ends, both the first device and the second device need to update the model parameters of the models deployed locally by themselves. After the update, continue with the next round of model training until the model training is terminated.

[0265] Exemplarily, after the first device determines to terminate the model training, the first device may send a relevant instruction to the second device to inform the second device to terminate the model training.

[0266] Exemplarily, the above threshold (the first threshold or the second threshold) may be predefined, or set by the first device itself, or indicated by other devices (such as the second device), which is not limited in this application.

[0267] It can be understood that Figure 8In the dimensionality reduction model, compression model, and reconstruction model described in the method 800 shown, they can be the dimensionality reduction model, compression model, and reconstruction model obtained after completing model training in the method 900. By way of example, if the first device and the second device are AI network element #1 and AI network element #2, such as third-party network element #1 and third-party network element #2, and the dimensionality reduction model and the reconstruction model will be deployed on the network device side, and the compression model will be deployed on the terminal device side, then the first device will send the trained (i.e., completed training) dimensionality reduction model and reconstruction model, or the training data set corresponding to the trained model (dimensionality reduction model and reconstruction model) to the network device, and the second device will send the trained compression model or the training data set corresponding to the trained compression model to the terminal device. Here, an example is given where the second device sends the training data of the compression model to the terminal device. After the above three models complete training to obtain the three trained models, the second device may not directly send the trained compression model, but send the training data set #1 corresponding to the trained compression model to the terminal device, and the terminal device trains the corresponding compression model based on the received training data set #1. The training data set #1 corresponding to the trained compression model may include information A and information B. Among them, information A is the output information obtained by using the channel prior information in the training data set in S910 as the input of the trained dimensionality reduction model, and information B is the output information obtained by using information A as the input of the trained compression model. That is, information A and information B in the training data set #1 are the input information and output information of the trained compression model.

[0268] The model training method 900 provided in this application has been described in detail above. Traditional model training methods only support the training of two models at both ends, while the method 900 supports the joint training of multiple models at both ends. Based on the method 900, the dimensionality reduction model, compression model, and reconstruction model can be jointly trained and optimized, and the effect of further reducing the channel information feedback overhead can be achieved.

[0269] Another model training method proposed in this application will be introduced below. The main difference from the method 900 is that in the method 900, three models are jointly trained at both ends, and in this method, three models are jointly trained on one side (for example, the first device). After that, if one or more of the three models, such as model A, will be deployed on the other device side (for example, the second device side), then the first device can send the trained model A or the training data set corresponding to the trained model A to the second device, and the second device completes model deployment or performs secondary training based on the received model A or training data set. The following will be combined Figure 10 for detailed description.

[0270] Figure 10FIG. 0 is a schematic flowchart of a method 1000 for model training proposed in this application. In this method, the first device can deploy the model locally, and transfer the intermediate layer calculation results and gradients in real time to complete the training of the unilateral model. Among them, the dimensionality reduction model, the compression model, and the reconstruction model are all deployed on the first device side. Optionally, the first device can be a terminal device, a network device, or an AI network element #1, such as a third-party network element #1. This method includes the following steps.

[0271] S1010, the first device obtains a first training data set, the first training data set includes channel prior information and downlink channel information #1, and the downlink channel information #1 is the real downlink channel information of N physical antennas, and N is a positive integer.

[0272] For the description of S1010, refer to the description in S910, which will not be elaborated here.

[0273] It can be understood that the training methods of the unilateral joint training of method 1000 and the double-end joint training of method 900 are the same. For the more specific training process, refer to the description in method 900.

[0274] S1020, the first device performs joint training on the dimensionality reduction model, the compression model, and the reconstruction model. Among them, the input of the dimensionality reduction model is the channel prior information, and the output is the dimensionality reduction weight value. The dimensionality reduction weight value is used to perform digital precoding on M reference signals corresponding to M antenna ports to measure the equivalent downlink channel of N physical antennas. M is a positive integer less than N. The input of the compression model is the downlink channel information #2, and the output is the first sequence. The downlink channel information #2 is determined based on the downlink channel information #1 and the dimensionality reduction weight value. The input of the reconstruction model is the first sequence, and the output is the downlink channel information #3.

[0275] Exemplarily, the channel prior information can be the uplink channel response H UL , and the downlink channel information #1 can be the real downlink channel response H of N physical antennas DL , for the input and output of each model and the example, refer to the description in S920 to S950, which will not be elaborated here.

[0276] S1030, the first device updates the network device parameters of the dimensionality reduction model, the compression model, and the reconstruction model based on the first error information. The first error information is determined based on the downlink channel information #3 and the downlink channel information #1.

[0277] It can be understood that in this method, the first device obtains the first gradient information and the second gradient information described in S940 and S950. Further, the first device updates the model parameters of the reconstruction model based on the first error information, updates the model parameters of the dimensionality reduction model based on the second gradient information, and updates the model parameters of the compression model based on the first gradient information.

[0278] For the first error information, related examples of the first error information, the first gradient information, and the second gradient information, refer to the descriptions in S940 to S950, which will not be elaborated here.

[0279] Optionally, the method further includes: S1040, when the termination condition for model training is satisfied, the first device determines to terminate the model training.

[0280] For examples of possible termination conditions for model training, refer to the descriptions in S970, which will not be elaborated here.

[0281] The above S1020 and S1030 can be regarded as one round of model training. After each round of model training ends, the first device needs to update the model parameters of the 3 models deployed locally. After the update, continue with the next round of model training until the model training is terminated.

[0282] In a possible scenario, if the trained dimensionality reduction model and reconstruction model will be deployed on the network device side, and the trained compression model will be deployed on the terminal device side, and the above first device is a network device, then when the network device completes the joint training of the 3 models, the network device can send the information corresponding to (1) or (2) to the terminal device:

[0283] (1) The network device sends the trained compression model to the terminal device. Correspondingly, the terminal device receives the trained compression model from the network device.

[0284] Example 1, the terminal device directly uses the received compression model as the compression model in method 800.

[0285] Example 2, the terminal device can use the received compression model after correction as the compression model in method 800.

[0286] It can be understood that the correction in Example 2 can be to retrain with some data sets adding the own characteristics of the terminal device, and / or modify the structure of the received compression model and then retrain.

[0287] (2) The network device sends the second training data set to the terminal device. The second training data set is the training data set composed of the input and output of the trained compression model. Correspondingly, the terminal device receives the second training data set from the network device.

[0288] It can be understood that the second training data set includes information C and information D, where information C is the output information obtained by using the channel prior information in the first training data set as the input of the trained dimensionality reduction model, and information D is the output information obtained by using information C as the input of the trained compression model.

[0289] In Example 1, the terminal device uses the compression model trained based on the second training dataset as the compression model in Method 800.

[0290] In Example 2, the compression model trained by the terminal device based on the second training dataset is corrected and used as the compression model in Method 800.

[0291] In another possible scenario, if the trained dimensionality reduction model and reconstruction model are deployed on the network device side, and the trained compression model is deployed on the terminal device side, and the above-mentioned first device is an AI network element, such as a third-party network element, the third-party network element can send the trained models or the corresponding training datasets to the network device and the terminal device respectively, which will not be described one by one here.

[0292] It should be noted that in Method 900 and Method 1000, the single-sided and double-sided joint training processes are described in detail with three models as examples. The training method proposed in this application is also applicable to the joint training process of Y (Y>3) models.

[0293] The above two model training methods provided by this application have been described in detail. Next, in combination with Figure 11 a simulation performance comparison chart between the dimensionality reduction method based on this application and the dimensionality reduction method of the traditional orthogonal codebook is given.

[0294] Figure 11 It is a performance comparison chart obtained by using the model dimensionality reduction and compression method (Method 1) proposed in this application and the dimensionality reduction and compression method (Method 2) under the traditional orthogonal codebook to determine the dimensionality reduction weights. Among them, the dimensionality reduction weights are used to measure the equivalent downlink channel of 128 physical antennas by using 32 reference signals (corresponding to 32 antenna ports), and the AI model structures used for compression in both schemes are the same. Figure 11 The horizontal axis of represents the feedback overhead, and the vertical axis represents the square generalized cosine similarity (SGCS) (SGCS can also be simply understood as the accuracy of the downlink precoding matrix of the reconstructed 128 physical antennas). Among them, SGCS satisfies the following formula where K is the rank (rank of the channel), N f is the number of frequency domain units (a frequency domain unit can be one or more subcarriers or RBs, without limitation), E{·} represents taking the average value of the samples in the brackets, is the downlink precoding matrix of the reconstructed 128 physical antennas, V is the label, that is, V is the true value of the downlink precoding matrix of the 128 physical antennas, (.) H is the conjugate transpose of the vector, ||.|| represents taking the modulus of the vector, and ||.|| represents taking the absolute value of the vector. From Figure 11It can be seen that, compared with the dimensionality reduction scheme under the orthogonal codebook, the scheme proposed in this application has about a 70% reduction in feedback overhead under the same performance (i.e., when the corresponding values on the vertical axis are the same).

[0295] It can be understood that the scope of application of the above method is not limited to port dimensionality reduction, and it can also be applied to dimensionality reduction scenarios in other dimensions. For example, it can be frequency-domain dimensionality reduction or analog beam-domain dimensionality reduction. This application does not make any limitations in this regard.

[0296] It can be understood that the magnitudes of the sequence numbers of the above processes do not imply the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of this application.

[0297] It can also be understood that in some of the above embodiments, devices in the existing network architecture are mainly used as examples for illustrative purposes. It can be understood that this application does not limit the specific form of the devices. For example, devices that can achieve the same functions in the future are applicable to the embodiments of this application.

[0298] It can be understood that in each of the above method embodiments, the methods and operations implemented by the devices (such as the above network devices, terminal devices, first devices, second devices, etc.) can also be implemented by components of the devices (such as chips or circuits).

[0299] Above, in combination with Figures 1 to 11 The method provided by the embodiments of this application has been described in detail. The above method has been mainly introduced from the perspective of the interaction between network devices and terminal devices (or first devices and second devices). It can be understood that network devices and terminal devices (or first devices and second devices) include the corresponding hardware structures and / or software modules for implementing the above functions.

[0300] Those skilled in the art should be able to realize that, in combination with the units and algorithm steps of the examples described in the embodiments disclosed in this article, this application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a certain function is executed in the form of hardware or computer software driving the hardware depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0301] Next, in combination with Figure 12 and Figure 13Describe the communication device provided in the embodiments of the present application. It can be understood that the descriptions of the device embodiments correspond to those of the method embodiments. Therefore, for the content not described in detail, reference can be made to the above method embodiments. For the sake of brevity, some content will not be repeated. The embodiments of the present application can divide the functions of the device according to the above method examples. For example, each function module can be divided corresponding to each function, or two or more functions can be integrated into one processing module. The above integrated module can be implemented in the form of hardware or in the form of a software function module. It should be noted that the division of modules in the embodiments of the present application is illustrative, only a logical function division, and there can be other division methods in actual implementation. The following takes the example of dividing each function module corresponding to each function for illustration.

[0302] The above provides a detailed description of the method provided in the present application. The following introduces the communication device provided in the present application. In one possible implementation, the device is used to implement the steps or processes corresponding to the terminal device in the above method embodiments. In another possible implementation, the device is used to implement the steps or processes corresponding to the network device in the above method embodiments. In yet another possible implementation, the device is used to implement the steps or processes corresponding to the first device in the above method embodiments. By way of example, the first device can be a network device or an AI network element #1, such as a third-party network element #1. In yet another possible implementation, the device is used to implement the steps or processes corresponding to the second device in the above method embodiments. By way of example, the second device can be a terminal device or an AI network element #2, such as a third-party network element #2. In yet another possible implementation, the device is used to implement the steps or processes corresponding to the network device and the AI network element that deploys the dimensionality reduction model and the reconstruction model in the above method 800. In yet another possible implementation, the device is used to implement the steps or processes corresponding to the terminal device and the AI network element that deploys the compression model in the above method embodiments. That is, the first device can be a network device and an AI network element #1, and / or the second device is a terminal device and an AI network element #2.

[0303] Figure 12 It is a schematic block diagram of the communication device 1200 provided in the embodiments of the present application. As Figure 12 shown, the device 1200 may include a communication unit 1210 and a processing unit 1220. The communication unit 1210 can communicate with the outside, and the processing unit 1220 is used for data processing. The communication unit 1210 can also be referred to as a communication interface or a transceiver unit.

[0304] In a possible design, the apparatus 1200 may implement the steps or processes corresponding to those performed by the network device in the above method embodiments. Among them, the processing unit 1220 is configured to perform the operations related to the processing of the network device in the above method embodiments, and the communication unit 1210 is configured to perform the operations related to the transmission of the network device in the above method embodiments.

[0305] In another possible design, the apparatus 1200 may implement the steps or processes corresponding to those performed by the terminal device in the above method embodiments. Among them, the communication unit 1210 is configured to perform the operations related to the reception of the terminal device in the above method embodiments, and the processing unit 1220 is configured to perform the operations related to the processing of the terminal device in the above method embodiments.

[0306] In yet another possible design, the apparatus 1200 may implement the steps or processes corresponding to those performed by the first device in the above method embodiments. Among them, the communication unit 1210 is configured to perform the operations related to the reception of the first device in the above method embodiments, and the processing unit 1220 is configured to perform the operations related to the processing of the first device in the above method embodiments.

[0307] In yet another possible design, the apparatus 1200 may implement the steps or processes corresponding to those performed by the second device in the above method embodiments. Among them, the communication unit 1210 is configured to perform the operations related to the reception of the second device in the above method embodiments, and the processing unit 1220 is configured to perform the operations related to the processing of the second device in the above method embodiments.

[0308] It can be understood that the apparatus 1200 here is embodied in the form of functional units. The term "unit" here may refer to an application specific integrated circuit (ASIC), an electronic circuit, a processor (such as a shared processor, a dedicated processor, or a group of processors, etc.) for executing one or more software or firmware programs, a memory, a combined logic circuit, and / or other suitable components that support the described functions.

[0309] The apparatus 1200 in each of the above solutions has the function of implementing the corresponding steps performed by the device in the above method. The function may be implemented by hardware or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above functions; for example, the communication unit may be replaced by a transceiver (for example, the sending unit in the communication unit may be replaced by a transmitter, and the receiving unit in the communication unit may be replaced by a receiver), and other units, such as the processing unit, etc., may be replaced by a processor to respectively perform the transceiver operations and related processing operations in each method embodiment.

[0310] In addition, the above communication unit may also be a transceiver circuit (for example, it may include a receiving circuit and a transmitting circuit), and the processing unit may be a processing circuit. In the embodiments of the present application, Figure 12 the device in may be the device in the foregoing embodiments, or may be a chip or a chip system, for example: a system on chip (SoC). Among them, the communication unit may be an input / output circuit or a communication interface; the processing unit is a processor, a microprocessor or an integrated circuit integrated on the chip. There is no limitation here.

[0311] Figure 13 FIG. 1300 is a schematic block diagram of a communication device 1300 provided by an embodiment of the present application. The device 1300 includes a processing circuit 1310 and a transceiver circuit 1320. Among them, the processing circuit 1310 and the transceiver circuit 1320 communicate with each other through an internal connection path, and the processing circuit 1310 is used to execute instructions to control the transceiver circuit 1320 to send signals and / or receive signals.

[0312] Optionally, the device 1300 may further include a memory 1330, and the memory 1330 communicates with the processing circuit 1310 and the transceiver circuit 1320 through an internal connection path. The memory 1330 is used to store instructions, and the processing circuit 1310 may execute the instructions stored in the memory 1330.

[0313] Exemplarily, the processing circuit 1310 may be one or more processors, or may also be all or part of the circuits in one or more processors.

[0314] Exemplarily, the transceiver circuit 1320 may be a transceiver, an interface circuit or an input / output circuit.

[0315] In one possible implementation, the device 1300 is used to implement each process and step corresponding to the network device in the foregoing method embodiments. In another possible implementation, the device 1300 is used to implement each process and step corresponding to the terminal device in the foregoing method embodiments. In yet another possible implementation, the device 1300 is used to implement each process and step corresponding to the first device in the foregoing method embodiments. In yet another possible implementation, the device 1300 is used to implement each process and step corresponding to the second device in the foregoing method embodiments.

[0316] It can be understood that the device 1300 can specifically be the device in the above embodiments, or can be a chip or a chip system. Correspondingly, the transceiver circuit 1320 can be the transceiver circuit of the chip, which is not limited herein. Specifically, the device 1300 can be used to execute each step and / or process corresponding to the device in the above method embodiments. Optionally, the memory 1330 can include a read-only memory and a random access memory, and provide instructions and data to the processing circuit. A part of the memory can also include a non-volatile random access memory. For example, the memory can also store information about the device type. The processing circuit 1310 can be used to execute the instructions stored in the memory, and when the processing circuit 1310 executes the instructions stored in the memory, the processing circuit 1310 is used to execute each step and / or process of the above method embodiment corresponding to the device.

[0317] In the implementation process, each step of the above method can be completed by the integrated logic circuit in the hardware of the processor or the instructions in the form of software. The steps of the method disclosed in combination with the embodiments of the present application can be directly embodied as being executed and completed by the hardware processor, or can be executed and completed by the combination of the hardware and software modules in the processor. The software module can be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and combines its hardware to complete the steps of the above method. To avoid repetition, it will not be described in detail here.

[0318] Exemplarily, the processor in the embodiments of the present application may be an integrated circuit chip with signal processing capabilities. During implementation, the steps of the above method embodiments may be completed by the integrated logic circuit in the hardware of the processor or instructions in software form. The above-mentioned processor may be a general-purpose processor, a digital signal processing (DSP), an ASIC, a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The processor in the embodiments of the present application may implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the embodiments of the present application may be directly embodied as being executed and completed by the hardware decoding processor, or executed and completed by a combination of the hardware and software modules in the decoding processor. The software module may be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. This storage medium is located in the memory, and the processor reads the information in the memory and combines its hardware to complete the steps of the above method.

[0319] It can be understood that the memory in the embodiments of the present application can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchlink DRAM (SLDRAM), and direct rambus RAM (DR RAM). It should be noted that the memory of the systems and methods described herein is intended to include but not be limited to these and any other suitable types of memory.

[0320] It should be noted that when the processor is a general-purpose processor, a DSP, an ASIC, an FPGA, or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components, the memory (storage module) can be integrated in the processor.

[0321] In addition, the present application also provides a computer-readable storage medium, in which computer instructions are stored. When the computer instructions are run on a computer, the operations and / or processes performed by the network device, the terminal device, the first device, or the second device in the method embodiments of the present application are executed.

[0322] The present application also provides a computer program product, which includes computer program code or instructions. When the computer program code or instructions are run on a computer, the operations and / or processes performed by the network device, the terminal device, the first device, or the second device in the method embodiments of the present application are executed.

[0323] In addition, the present application further provides a chip, which includes a processing circuit. A memory for storing a computer program is provided separately from the chip, and the processing circuit is configured to execute the computer program stored in the memory, so that the operations and / or processes performed by the network device, the terminal device, the first device, or the second device in any of the method embodiments are executed.

[0324] Further, the chip may further include a communication interface. The communication interface may be an input / output interface or an interface circuit, etc. Further, the chip may further include a memory.

[0325] In addition, the present application further provides a communication system, including the network device and the terminal device in the embodiments of the present application, and / or the first device and the second device.

[0326] It should also be noted that the memory described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0327] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or by a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application. Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein. In several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms. The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place, or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment. In addition, in each embodiment of this application, the functional units can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0328] If the above functions are implemented in the form of software function units and sold or used as independent products, they 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 to the prior art, or a part of this 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 enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of this application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical discs that can store program codes.

[0329] It will be understood that the "embodiments" referred to throughout the specification mean that specific features, structures, or characteristics related to the embodiments are included in at least one embodiment of the present application. Thus, the various embodiments mentioned throughout the specification do not necessarily refer to the same embodiment. In addition, these specific features, structures, or characteristics may be combined in one or more embodiments in any suitable manner.

[0330] It will also be understood that in the present application, "when", "if", and "in case" all refer to the fact that under certain objective circumstances, the network element will perform corresponding processing, which does not limit the time, and does not require the network element to have a judgment action when implemented, nor does it mean the existence of other limitations.

[0331] It will further be understood that in the various embodiments of the present application, "B corresponding to A" means that B is associated with A, and B can be determined based on A. However, it will also be understood that determining B based on A does not mean that B is determined solely based on A, and B can also be determined based on A and / or other information.

[0332] As described above, the above are only specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed in the present application can easily think of changes or substitutions, which should all be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A communication method, characterized in that, Comprising: Performing digital precoding on M reference signals corresponding to M antenna ports based on dimension-reduced weights, where the dimension-reduced weights are used to measure the equivalent downlink channel of N physical antennas by using the M reference signals, N is a positive integer, M is a positive integer less than N, and the dimension-reduced weights are determined based on a dimension-reduction model; Transmitting the M reference signals; Transmitting first indication information, where the first indication information indicates a compression model, and the compression model matches the dimension-reduction model; Receiving a first sequence, where the first sequence is a compressed feedback quantity obtained by taking first downlink channel information as an input of the compression model, and the first downlink channel information is based on measurements of the M reference signals; Obtaining second downlink channel information, where the second downlink channel information is information obtained by taking the first sequence as an input of a reconstruction model, the reconstruction model matches the dimension-reduction model, and the second downlink channel is used to determine the downlink channel information of the N physical antennas.

2. A communication method, characterized in that, Comprising: Receiving M reference signals corresponding to M antenna ports, where M is a positive integer; Obtaining first downlink channel information based on the M reference signals; Receiving first indication information, where the first indication information indicates a compression model, the compression model matches a dimension-reduction model, the dimension-reduction model is used to obtain dimension-reduced weights, and the dimension-reduced weights are used to measure the equivalent downlink channel of N physical antennas by using the M reference signals, N is a positive integer, and M is a positive integer less than N; Transmitting a first sequence, where the first sequence is a compressed feedback quantity obtained by taking the first downlink channel information as an input of the compression model, and the first sequence is used as an input of a reconstruction model to be decompressed to obtain second downlink channel information, the reconstruction model matches the dimension-reduction model, and the second downlink channel is used to determine the downlink channel information of the N physical antennas.

3. The method according to claim 1 or 2, characterized in that, The dimension-reduced weights are output information obtained by taking channel prior information as an input of the dimension-reduction model.

4. The method according to claim 3, characterized in that, The channel prior information is determined based on at least one of the following information: currently measured uplink channel information, historical uplink channel information, historical downlink channel information, or sensed channel information.

5. The method according to claim 3 or 4, characterized in that, The dimension-reduction model, the compression model, and the reconstruction model are all related to the number of antenna ports M, the number of physical antennas N, and the channel prior information.

6. The method according to any one of claims 1 to 5, characterized in that The first downlink channel information indicates the response of the equivalent downlink channel, and the first sequence is a compressed feedback quantity of the response.

7. The method according to any one of claims 1 to 5, characterized in that The first downlink channel information indicates the precoding matrix corresponding to the response of the equivalent downlink channel, and the first sequence is a compressed feedback quantity of the precoding matrix information corresponding to the response.

8. The method according to any one of claims 1 to 7, characterized in that, The downlink channel information of the N physical antennas includes the downlink precoding matrix of the N physical antennas.

9. A model training method, characterized in that, Comprising: Obtain a training data set, where the training data set includes channel prior information and first downlink channel information, and the first downlink channel information is the true downlink channel information of N physical antennas, and N is a positive integer; Transmit second downlink channel information, where the second downlink channel information is determined based on the first downlink channel information and a dimension reduction weight value. Among them, the dimension reduction weight value is the output information obtained by taking the channel prior information as the input of a dimension reduction model. The dimension reduction weight value is used to perform digital precoding on M reference signals corresponding to M antenna ports to measure the equivalent downlink channel of the N physical antennas, and M is a positive integer less than N; Receive a first sequence, where the first sequence is the output information obtained by taking the second downlink channel information as the input of a compression model; Transmit first gradient information, where the first gradient information is the input-side gradient information of a reconstruction model. Among them, the first gradient information is used to update the network parameters of the compression model, and the first gradient information is determined based on first error information and the reconstruction model. The first error information is determined based on third downlink channel information and the first downlink channel information. The third downlink channel information is the output information obtained by taking the first sequence as the input of the reconstruction model or is calculated based on the output information of the reconstruction model; Receive second gradient information, where the second gradient information is the input-side gradient information of the compression model, and the second gradient information is determined based on the first gradient information and the compression model; Update the model parameters of the reconstruction model based on the first error information, and update the model parameters of the dimension reduction model based on the second gradient information.

10. The method according to claim 9, wherein The method further includes: Terminate model training when the termination condition for model training is satisfied.

11. A model training method, characterized in that, It includes: Receive second downlink channel information, where the second downlink channel information is determined based on first downlink channel information and a dimension reduction weight value. The first downlink channel information is the true downlink channel information of N physical antennas, and the dimension reduction weight value is the output information obtained by taking channel prior information as the input of a dimension reduction model. The dimension reduction weight value is used to perform digital precoding on M reference signals corresponding to M antenna ports to measure the equivalent downlink channel of the N physical antennas. N is a positive integer, M is a positive integer less than N, and the first downlink channel information and the channel prior information are the training data set for model training; Transmit a first sequence, where the first sequence is the output information obtained by taking the information of the second downlink channel as the input of a compression model; Receive first gradient information, where the first gradient information is the input - side gradient information of the reconstruction model. The first gradient information is used to update the model parameters of the compression model, and the first gradient information is determined based on the first error information and the reconstruction model. The first error information is determined based on the third downlink channel information and the first downlink channel information. The third downlink channel information is the output information obtained by using the first sequence as the input of the reconstruction model or is calculated based on the output information of the reconstruction model. The first error information is used to update the model parameters of the reconstruction model; Send second gradient information, where the second gradient information is the input - side gradient information of the compression model, and the second gradient information is determined based on the first gradient information and the compression model; Update the model parameters of the compression model based on the first gradient information.

12. The method according to any one of claims 9 to 11, characterized in that, The channel prior information is determined based on at least one of the following information: currently measured uplink channel information, historical uplink channel information, historical downlink channel information, or sensed channel information.

13. The method according to any one of claims 9 to 12, characterized in that, The first downlink channel information is the true downlink channel response of the N physical antennas.

14. The method according to claim 13, wherein, The second downlink channel information indicates the downlink channel response, and the downlink channel response is determined based on the first downlink channel information and the weight reduction value, or, The second downlink channel information indicates the weight value corresponding to the downlink channel response.

15. The method according to claim 14, characterized in that, The third downlink channel information is the reconstructed weights of the N physical antennas. The first error information indicates the error between the third downlink channel information and the true weights of the N physical antennas, and the true weights of the N physical antennas are determined based on the first downlink channel information.

16. A communication device, characterized in that, Comprising a module or unit for performing the method according to any one of claims 1, 3 to 8, or comprising a module or unit for performing the method according to any one of claims 2 to 8, or comprising a module or unit for performing the method according to any one of claims 9, 10, 12 to 15, or comprising a module or unit for performing the method according to any one of claims 11 to 15.

Images