Channel state information feedback method, device, electronic device and storage medium

Abstract

The present application provides a channel state information feedback method, apparatus, electronic device, and storage medium. The method comprises: obtaining channel matrix information, and determining first channel information based on the channel matrix information, and obtaining second channel information based on the first channel information, wherein the second channel information is obtained based on a pre-trained channel state information prediction model, and using the first channel information and the second channel information to perform channel state information feedback. By introducing a channel state information prediction model based on deep learning and based on the concept of using partial channels for prediction to obtain all channels, the pilot overhead in CSI feedback in FDD mode is significantly reduced while ensuring feedback performance.

Description

Technical Field

[0001] The present application relates to the field of communication technology, and in particular to a channel state information feedback method, device, electronic device and storage medium. Background Art

[0002] Massive Multiple-Input Multiple-Output (MIMO) is considered a key enabling technology for future mobile communication systems due to its significant spatial multiplexing gain, diversity gain, and beamforming capabilities. However, these potential benefits require accurate access to uplink and downlink channel state information (CSI). In frequency-division duplexing (FDD) mode, downlink CSI must be sent to the base station via a feedback link, which consumes a significant amount of uplink transmission resources.

[0003] Among related technologies, deep learning-based CSI feedback technology mainly focuses on how to improve the performance of CSI feedback reconstruction and how to reduce computational complexity, while ignoring the problem of excessive pilot overhead in the initial channel estimation stage. Summary of the Invention

[0004] In view of this, the purpose of the present application is to provide a channel state information feedback method, device, electronic device and storage medium.

[0005] Based on the above objectives, the present application provides a channel state information feedback method, including:

[0006] Channel matrix information is acquired, and first channel information is determined according to the channel matrix information.

[0007] Second channel information is obtained according to the first channel information; wherein the second channel information is obtained based on a pre-trained channel state information prediction model.

[0008] Channel state information feedback is performed using the first channel information and the second channel information.

[0009] Optionally, obtaining second channel information according to the first channel information includes:

[0010] An antenna prediction module based on deep learning and the channel state information prediction model obtains rough channel information according to the first channel information.

[0011] The second channel information is obtained according to the coarse channel information.

[0012] Optionally, obtaining the second channel information according to the coarse channel information includes:

[0013] Third channel information is determined according to the channel matrix information and the first channel information.

[0014] Accurate channel information is obtained according to the rough channel information.

[0015] Obtaining fourth channel information according to the precise channel information;

[0016] The comparison module based on the channel state information prediction model retains the accurate channel information as the second channel information in response to determining that the error between the fourth channel information and the third channel information is less than a threshold.

[0017] Optionally, the method further includes determining an error between the fourth channel information and the third channel information by using the following formula:

[0018]

[0019] Among them, m represents the number of samples, H p represents the third channel information, represents the fourth channel information, i=1, 2, 3, ..., n.

[0020] Optionally, the method further includes:

[0021] In response to determining that the error between the fourth channel information and the third channel information is not less than a threshold, the first channel information is adjusted, or the channel state information prediction model is adjusted.

[0022] Optionally, the method further includes:

[0023] In response to determining that the error between the fourth channel information and the third channel information is less than -25 decibels, retaining the accurate channel information as the second channel information.

[0024] Optionally, the using the first channel information and the second channel information to perform channel state information feedback includes:

[0025] The channel state information feedback is implemented based on a channel state information network using the first channel information and the second channel information.

[0026] Based on the above objectives, the present application also provides a channel state information feedback device, including:

[0027] The acquisition module is configured to acquire channel matrix information.

[0028] The determining module is configured to determine first channel information according to the channel matrix information.

[0029] The prediction module is configured to obtain second channel information according to the first channel information; wherein the second channel information is obtained based on a pre-trained channel state information prediction model.

[0030] The feedback module is configured to use the first channel information and the second channel information to perform channel state information feedback.

[0031] Based on the above purpose, the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method described in any one of the above embodiments is implemented.

[0032] Based on the above purpose, the present application also provides a non-transitory computer-readable storage medium, which stores computer instructions, and the computer instructions are used to enable a computer to execute the method described in any one of the above embodiments.

[0033] From the above, it can be seen that the channel state information feedback method, device, electronic device and storage medium provided in the present application, by introducing a channel state information prediction model based on deep learning, and based on the concept of using partial channels for prediction to obtain all channels, while ensuring feedback performance, significantly reduces the pilot overhead in CSI feedback in FDD mode. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in this application or related technologies, the following briefly introduces the drawings required for use in the embodiments or related technical descriptions. Obviously, the drawings described below are merely embodiments of this application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0035] Figure 1 A schematic diagram shows a framework 100 of an exemplary method for channel state information feedback provided in the present application for a wireless communication system.

[0036] Figure 2 A flow chart of an exemplary method for channel state information feedback provided in the present application is shown.

[0037] Figure 3 A flowchart of an exemplary method for selecting an antenna position provided in this application is shown.

[0038] Figure 4 A schematic diagram showing the channel state information prediction model architecture provided in this application is shown.

[0039] Figure 5 A schematic diagram of a process for predicting channel information based on a pilot interpolation algorithm in a channel state information prediction model provided by the present application is shown.

[0040] Figure 6 A flow chart of an exemplary method for channel state information feedback provided in the present application is shown.

[0041] Figure 7 A schematic diagram of a channel state information feedback device provided by the present application is shown.

[0042] Figure 8 A schematic diagram of an electronic device provided by the present application is shown. DETAILED DESCRIPTION

[0043] In order to make the objectives, technical solutions and advantages of this application more clear, this application is further described in detail below in combination with specific embodiments and with reference to the accompanying drawings.

[0044] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present application should have the usual meanings understood by people with ordinary skills in the field to which this application belongs. The "first", "second" and similar words used in the embodiments of the present application do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0045] As described in the background technology, current deep learning-based CSI feedback technology mainly focuses on how to improve the performance of CSI feedback reconstruction and how to reduce computational complexity, while ignoring the problem of excessive pilot overhead in the initial channel estimation stage.

[0046] This application provides a channel state information feedback method, apparatus, electronic device, and storage medium. This establishes a new CSI feedback framework. Leveraging spatial channel correlation, this framework introduces a novel concept for CSI feedback: using partial antenna channels to predict the full antenna channel. This significantly reduces pilot overhead during CSI feedback. Finally, a comparison module within the channel state information prediction model allows for direct evaluation of the overall prediction model's performance, enabling early assessment of model quality and enabling timely adjustments or replacements.

[0047] Figure 1 A schematic diagram of a framework 100 for a wireless communication system using an exemplary method for channel state information feedback provided in the present application is shown. The wireless communication system may include: a base station (BS) and a user equipment (UE), wherein the user equipment serves as a receiving end to obtain the CSI of the downlink and returns the CSI to the base station through a feedback link. The base station may be a commonly used base station, an evolved base station, or a network device (e.g., a next-generation base station) or a transmitting and receiving point in a 5G system. The user end may be a mobile phone, a tablet computer, a laptop computer, an ultra-mobile personal computer, a netbook or a personal digital assistant, a mobile Internet device, a wearable device or a vehicle-mounted device, etc. The framework 100 may include a CSI prediction model (i.e., a channel state information prediction model) and a CsiNet feedback reconstruction model (i.e., a channel state information network). The channel state information prediction model is a channel state information prediction model based on deep learning, which may include an antenna prediction module and a comparison module based on deep learning.

[0048] Here, we use a simple single-cell downlink massive MIMO FDD system model as an example. Assuming this model operates on an OFDM (Orthogonal Frequency Division Multiplexing) system, the base station is equipped with 32 antennas, 24 of which are used to transmit pilot signals, while the user end is equipped with only a single antenna. After receiving the channel matrix in the space-frequency domain, the user end first converts the channel matrix from the space-frequency domain to the angular delay domain using a two-dimensional discrete Fourier transform (DFT). This method leverages the channel sparsity in the angular delay domain to reduce feedback overhead. The user end then selects the channels of 16 of the 24 received antennas and uses a deep learning-based channel state information prediction model to predict the channels of all 32 antennas. The remaining 8 of the 24 received antennas are then used in subsequent comparison steps.

[0049] After obtaining the complete channel state information for all 32 antennas (i.e., the channel state information for the selected 16 antennas and the predicted 16 antennas), the actual channel values ​​for the eight antennas mentioned above are compared with the predicted channel values ​​for the corresponding eight of the 16 antennas to determine the credibility of the deep learning-based channel state information prediction model. This is known as the comparison module. If the error between the actual and predicted values ​​is small, the prediction model is considered effective and feedback reconstruction can proceed directly. If the error between the actual and predicted values ​​is large, the prediction model needs to be modified promptly or the number of antennas used for channel prediction needs to be adjusted appropriately to improve the prediction model's performance.

[0050] Figure 2 The following is a flow chart showing an exemplary method for channel state information feedback provided in the present application. The method may include the following steps.

[0051] In step S201, channel matrix information is obtained and first channel information is determined based on the channel matrix information. Taking the aforementioned system model as an example, the base station is equipped with 32 antennas, or 32 channels, 24 of which are used to transmit pilot signals. The channel information is arranged in a matrix. After receiving the channel information for all 24 antennas, the user terminal first selects the positions of some of the antennas.

[0052] In some embodiments, as Figure 3 As shown, assuming that there are 16 antennas to be selected, a special case can be used to select antenna positions, namely, using the channel information at odd-numbered positions at equal intervals to predict the channel information at even-numbered positions. In other words, in a channel matrix with 24 antennas, the channel information of 16 antennas at odd positions (i.e., the first channel information) is selected to predict the channel information at the remaining even positions (i.e., the second channel information described below). The channel information of the remaining 8 antennas in the channel matrix (i.e., the third channel information described below) is then used in subsequent steps to compare and verify the model.

[0053] In step S203, second channel information is obtained based on the first channel information; wherein the second channel information is obtained based on a pre-trained channel state information prediction model. The model architecture is as follows Figure 4As shown, to obtain complete channel information for 32 antennas, a two-stage training method is proposed for a deep learning-based channel state information prediction model. In the first stage, a coarse channel estimate at even positions can be performed using other interpolation algorithms such as binary B-spline interpolation, radial basis functions, or binary linear interpolation. In the second stage, a super-resolution reconstruction neural network such as SRResNet (for image super-resolution) or SRCNN (convolutional neural network for image super-resolution) can be used to accurately estimate the channel matrix to obtain accurate antenna channel prediction values. SRResNet is a high-resolution reconstruction network that contains 16 residual blocks, each of which contains two 3×3 convolutional layers (Conv). The first convolutional layer is followed by a batch normalization (BN) layer and a PReLU (Parametric Rectified Linear Unit) activation function, while the second convolutional layer is only followed by BN. Figure 4 In the figure, Conv represents convolution, BN stands for batch normalization, PReLu is a parameterized activation function, Elementwise Sum is a summation operation, and K9N64S1 is the parameterized explanation of the convolution operation. K is the convolution kernel size (9×9), N is the number of output channels (64), and S is the stride size (1). This method, based on a channel state information prediction model, uses partial channels for prediction to obtain the full channel, significantly reducing pilot overhead in the CSI feedback process.

[0054] In some embodiments, obtaining the second channel information based on the first channel information further includes: obtaining rough channel information based on the first channel information by an antenna prediction module based on deep learning based on the channel state information prediction model; and obtaining the second channel information based on the rough channel information. In the first stage, if Figure 5 As shown, the channel values ​​(i.e., the first channel information) of the 16 predicted antennas equidistantly placed at odd positions in the angle domain are adopted. At this time, drawing on the principle of pilot interpolation in channel estimation, the binary B-spline interpolation algorithm is used to perform coarse channel estimation at even positions (i.e., obtain coarse channel information).

[0055] In some embodiments, obtaining the second channel information based on the coarse channel information further includes: determining third channel information based on the channel matrix information and the first channel information; obtaining precise channel information based on the coarse channel information; obtaining fourth channel information based on the precise channel information; and retaining the precise channel information as the second channel information in response to determining that the error between the fourth channel information and the third channel information is less than a threshold value based on a comparison module of the channel state information prediction model. In the second stage, the interpolated channel matrix will be regarded as a low-resolution image (i.e., coarse channel information). At this time, the high-resolution image reconstruction idea in image processing is used to send the matrix to the partial architecture of SRResNet for accurate estimation of the channel matrix, so as to obtain accurate predicted channel values ​​(i.e., obtain precise channel information), and then obtain all channel values ​​of the 32 antennas (i.e., first channel information and second channel information).

[0056] In some embodiments, based on the comparison module of the channel state information prediction model, in response to determining that the error between the fourth channel information and the third channel information is less than a threshold, the precise channel information is retained as the second channel information. For the comparison module in the channel state information prediction model, the performance of the prediction model is evaluated by calculating the mean square error between the true channel values ​​of the remaining 8 antennas (i.e., the third channel information) and the antenna channel values ​​of the number of antennas corresponding to the 8 antennas in the predicted channel values ​​(i.e., the fourth channel information), so that the quality of the model can be known as early as possible so as to adjust or replace the model in time. Wherein, the error between the fourth channel information and the third channel information is determined by the following formula:

[0057]

[0058] Among them, m represents the number of samples, H p represents the third channel information, represents the fourth channel information, i=1, 2, 3, ..., n.

[0059] In some embodiments, in response to determining that the error between the fourth channel information and the third channel information is less than -25 decibels, the accurate channel information is retained as the second channel information. If the error between the actual channel value and the predicted channel value is less than -25 decibels, the channel state information prediction model is considered effective and subsequent feedback reconstruction can be performed directly.

[0060] In some embodiments, obtaining the second channel information based on the rough channel information may further include: adjusting the first channel information or adjusting the channel state information prediction model in response to determining that the error between the fourth channel information and the third channel information is not less than a threshold. If the error between the actual channel value and the predicted channel value is large, it is necessary to promptly change the channel state information prediction model or appropriately adjust the number of antennas used for channel prediction (i.e., adjust the first channel information) to improve the performance of the channel state information prediction model. Changing the channel state information prediction model can adjust the parameters in the model and can also adjust the structure of the model.

[0061] In some embodiments, the quality of the channel state information prediction model can be judged by designing a comparison module of the channel state information prediction model at the base station. Figure 6 As shown in the figure, the user end selects 16 antenna channels from the received 24 antenna channels for deep learning-based channel prediction and encodes the channels for 32 antennas (the predicted 16 antennas and the selected 16 antennas) for feedback. Simultaneously, the channels for the remaining 8 antennas are encoded and fed back to the user end. After receiving the compressed codewords for the predicted 32 antenna channels and the 8 comparison antenna channels from the user end, the base station decodes them and calculates the mean squared error between the decoded predicted values ​​and the actual decoded values ​​for the 8 comparison antennas to determine the quality of the deep learning-based antenna prediction model.

[0062] In step S205, channel state information feedback is performed using the first channel information and the second channel information.

[0063] In some embodiments, the using the first channel information and the second channel information to perform channel state information feedback further includes: using the first channel information and the second channel information and implementing the channel state information feedback based on a channel state information network. Figure 2 As shown in the figure, the CsiNet feedback reconstruction model consists of two parts: an encoder and a decoder. The encoder is deployed at the user end and is used to perform CSI compression for 32 antennas (i.e., the first channel information and the second channel information). The decoder is deployed at the base station end and is used to reconstruct the CSI. Finally, the reconstruction accuracy of the reconstructed CSI is calculated with the original data set at different compression ratios to test the performance of the framework 100. After the feedback reconstruction of the complete 32 antennas, the NMSE (Normalized Root Mean Square Error) of the feedback reconstructed channel and the original channel is calculated to determine the performance guarantee capability of the framework 100 when reducing the pilot overhead.

[0064] It should be noted that the method of the embodiment of the present application can be performed by a single device, such as a computer or server. The method of this embodiment can also be applied in a distributed scenario and performed by multiple devices working together. In such a distributed scenario, one of the multiple devices may only perform one or more steps of the method of the embodiment of the present application, and the multiple devices will interact with each other to complete the method.

[0065] It should be noted that the above description is limited to some embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in an order different from that described in the above embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0066] Based on the same technical concept, corresponding to any of the above-mentioned embodiment methods, the present application also provides a channel state information feedback device.

[0067] refer to Figure 7 , the channel state information feedback device includes:

[0068] The acquisition module 701 is configured to acquire channel matrix information.

[0069] The determination module 702 is configured to determine first channel information according to the channel matrix information.

[0070] The prediction module 703 is configured to obtain second channel information according to the first channel information; wherein the second channel information is obtained based on a pre-trained channel state information prediction model.

[0071] The feedback module 704 is configured to perform channel state information feedback using the first channel information and the second channel information.

[0072] For the convenience of description, the above devices are described as being divided into various modules according to their functions. Of course, when implementing this application, the functions of each module can be implemented in the same or multiple software and / or hardware.

[0073] The apparatus of the above embodiment is used to implement the corresponding channel state information feedback method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be described in detail here.

[0074] Based on the same technical concept, corresponding to any of the above-mentioned embodiments and methods, the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and runnable on the processor, wherein when the processor executes the program, the channel state information feedback method as described in any of the above embodiments is implemented.

[0075] Figure 8 10 is a schematic diagram showing a more specific hardware structure of an electronic device provided in this embodiment. The device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are communicatively connected to each other within the device via the bus 1050.

[0076] The processor 1010 can be implemented using a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0077] The memory 1020 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage devices, dynamic storage devices, etc. The memory 1020 can store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 1020 and is called and executed by the processor 1010.

[0078] The input / output interface 1030 is used to connect an input / output module to implement information input and output. The input / output module can be configured as a component in the device (not shown in the figure) or can be externally connected to the device to provide corresponding functions. Input devices may include a keyboard, mouse, touch screen, microphone, various sensors, etc., and output devices may include a display, speaker, vibrator, indicator light, etc.

[0079] The communication interface 1040 is used to connect to a communication module (not shown) to enable communication between the device and other devices. The communication module can communicate via a wired method (such as USB, network cable, etc.) or a wireless method (such as mobile network, WiFi, Bluetooth, etc.).

[0080] The bus 1050 comprises a pathway for transmitting information between the various components of the device (eg, the processor 1010 , the memory 1020 , the input / output interface 1030 , and the communication interface 1040 ).

[0081] It should be noted that although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040, and the bus 1050, in a specific implementation, the device may also include other components necessary for normal operation. In addition, it will be understood by those skilled in the art that the above device may only include the components necessary to implement the embodiments of this specification, and does not necessarily include all the components shown in the figure.

[0082] The electronic device of the above embodiment is used to implement the corresponding channel state information feedback method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be described in detail here.

[0083] Based on the same technical concept, corresponding to any of the above-mentioned embodiment methods, the present application also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the channel state information feedback method described in any of the above embodiments.

[0084] The computer-readable media of this embodiment include permanent and non-permanent, removable and non-removable media, and information storage can be achieved by any method or technology. Information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device.

[0085] The computer instructions stored in the storage medium of the above embodiment are used to enable the computer to execute the channel state information feedback method described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0086] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present application (including the claims) is limited to these examples. Within the scope of the present application, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the embodiments of the present application as described above, which are not provided in detail for the sake of simplicity.

[0087] In addition, for simplicity of description and discussion, and in order not to make the embodiment of the application difficult to understand, the known power supply / ground connection with integrated circuit (IC) chip and other components may or may not be shown in the accompanying drawings provided. In addition, the device can be shown in the form of a block diagram to avoid making the embodiment of the application difficult to understand, and this also takes into account the following fact, that is, the details of the embodiment of these block diagram devices are highly dependent on the platform to be implemented in the embodiment of the application (that is, these details should be fully within the scope of understanding of those skilled in the art). When specific details (for example, circuit) are set forth to describe exemplary embodiments of the application, it will be apparent to those skilled in the art that the embodiment of the application can be implemented without these specific details or when these specific details are changed. Therefore, these descriptions should be considered to be illustrative rather than restrictive.

[0088] Although the present invention has been described in conjunction with specific embodiments thereof, many alternatives, modifications, and variations of these embodiments will be apparent to those skilled in the art based on the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may utilize the embodiments discussed.

[0089] The embodiments of the present application are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present application should be included in the scope of protection of this application.

Claims

1. A channel state information feedback method, characterized in that: include: Acquiring channel matrix information and determining first channel information based on the channel matrix information, including: selecting channel information at odd positions in the channel matrix information; Obtaining second channel information based on the first channel information, including: using channel information at odd positions to predict channel information at even positions at equal intervals; wherein the second channel information is obtained based on a pre-trained channel state information prediction model; Channel state information feedback is performed using the first channel information and the second channel information.

2. The method according to claim 1, characterized in that The obtaining second channel information according to the first channel information includes: An antenna prediction module based on deep learning of the channel state information prediction model obtains rough channel information according to the first channel information; The second channel information is obtained according to the coarse channel information.

3. The method according to claim 2, characterized in that The obtaining the second channel information according to the coarse channel information includes: determining third channel information according to the channel matrix information and the first channel information; Obtaining precise channel information based on the rough channel information; Obtaining fourth channel information according to the precise channel information; The comparison module based on the channel state information prediction model retains the accurate channel information as the second channel information in response to determining that the error between the fourth channel information and the third channel information is less than a threshold.

4. The method according to claim 3, characterized in that The method further includes determining an error between the fourth channel information and the third channel information by using the following formula: Among them, m represents the number of samples, H p represents the third channel information, represents the fourth channel information, i=1, 2, 3, ..., n.

5. The method according to claim 3, characterized in that The method further comprises: In response to determining that the error between the fourth channel information and the third channel information is not less than a threshold, the first channel information is adjusted, or the channel state information prediction model is adjusted.

6. The method according to claim 3, characterized in that The method further comprises: In response to determining that the error between the fourth channel information and the third channel information is less than -25 decibels, retaining the accurate channel information as the second channel information.

7. The method according to claim 1, characterized in that The using the first channel information and the second channel information to perform channel state information feedback includes: The channel state information feedback is implemented based on a channel state information network using the first channel information and the second channel information.

8. A channel state information feedback device, characterized in that: include: An acquisition module, configured to acquire channel matrix information; A determination module is configured to determine first channel information according to the channel matrix information, including: selecting channel information at an odd position in the channel matrix information; a prediction module configured to obtain second channel information based on the first channel information, comprising: using channel information at odd positions to predict channel information at even positions at equal intervals; wherein the second channel information is obtained based on a pre-trained channel state information prediction model; The feedback module is configured to use the first channel information and the second channel information to perform channel state information feedback.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the method according to any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to enable a computer to execute the method according to any one of claims 1 to 7.

Images