Method, device, storage medium and chip for acquiring channel quality

Abstract

The present disclosure relates to a method, device, storage medium and chip for acquiring channel quality. The method comprises: receiving, by a terminal device, a pilot signal transmitted by a network device through a downlink channel; acquiring a first channel matrix according to the pilot signal; compressing the first channel matrix according to a channel state information (CSI) compression model and a CSI compression parameter to obtain a compressed target channel matrix; and transmitting the target channel matrix to the network device, so that the network device determines the channel quality of the downlink channel according to the target channel matrix. The first channel matrix is used to represent the channel quality of the downlink channel; the CSI compression model comprises a channel encoder, the channel encoder comprises a plurality of sub-encoders; different sub-encoders correspond to different CSI compression parameters; and the target channel matrix is used to instruct the network device to determine the channel quality of the downlink channel.

Description

Technical Field

[0001] This disclosure relates to the field of communication technology, and more specifically, to a method, apparatus, storage medium, and chip for acquiring channel quality. Background Technology

[0002] As a key technology of 5G (the 5th Generation Mobile Communication Technology), mMIMO (massive multiple-input multiple-output) technology has become widely researched and used in the communications field in recent years. By deploying a large number of antennas at the transmitting end using centralized or distributed methods, mMIMO systems have shown good performance in terms of system stability, energy efficiency, and anti-interference capabilities. To fully utilize the advantages of mMIMO systems, accurate CSI (Channel State Information) is needed at the transmitting end. For example, terminal devices can report CSI information to network devices so that network devices can obtain the channel quality of the downlink channel and select appropriate modulation and coding schemes for downlink transmission based on the channel quality, thereby improving data transmission performance. However, in mMIMO, as the number of antennas continues to increase, the information contained in the CSI becomes increasingly rich, resulting in a growing resource overhead for CSI reporting.

[0003] In related technologies, to reduce CSI reporting overhead, CSI compression technology based on Discrete Fourier Transform (DFT) can be used, where terminal devices compress the CSI before reporting it to network devices. However, CSI compression reporting reduces the accuracy of channel quality data obtained by network devices, especially since the accuracy of CSI compression varies significantly in different scenarios, which can affect data transmission efficiency. Summary of the Invention

[0004] To overcome the aforementioned problems in related technologies, this disclosure provides a method, apparatus, storage medium, and chip for obtaining channel quality.

[0005] According to a first aspect of the present disclosure, a method for obtaining channel quality is provided, applied to a terminal device, the method comprising:

[0006] Receive pilot signals transmitted by network devices through the downlink channel;

[0007] A first channel matrix is ​​obtained based on the pilot signal; the first channel matrix is ​​used to characterize the channel quality of the downlink channel;

[0008] The first channel matrix is ​​compressed according to the Channel State Information (CSI) compression model and the CSI compression parameters to obtain the compressed target channel matrix; wherein, the CSI compression model includes a channel encoder, and the channel encoder includes multiple sub-encoders; different sub-encoders correspond to different CSI compression parameters;

[0009] The target channel matrix is ​​sent to the network device so that the network device can obtain the channel quality of the downlink channel based on the target channel matrix.

[0010] According to a second aspect of the present disclosure, a method for obtaining channel quality is provided, applied to a network device, the method comprising:

[0011] The terminal device receives a target channel matrix sent by the terminal device; the target channel matrix is ​​obtained by the terminal device after compressing the first channel matrix according to the Channel State Information (CSI) compression model and the CSI compression parameters, and the first channel matrix is ​​a matrix obtained by the terminal device based on the pilot signal to characterize the channel quality of the downlink channel.

[0012] The target channel matrix is ​​decompressed according to the Channel State Information (CSI) decompression model and the CSI compression parameters to obtain a third channel matrix; wherein, the CSI decompression model includes a channel decoder, and the channel decoder includes multiple sub-decoders, with different sub-decoders corresponding to different CSI compression parameters;

[0013] The channel quality of the downlink channel is determined based on the third channel matrix.

[0014] According to a third aspect of the present disclosure, an apparatus for acquiring channel quality is provided, applied to a terminal device, the apparatus comprising:

[0015] The first receiving module is configured to receive pilot signals transmitted by network devices through the downlink channel;

[0016] The first matrix acquisition module is configured to acquire a first channel matrix based on the pilot signal; the first channel matrix is ​​used to characterize the channel quality of the downlink channel.

[0017] The target matrix acquisition module is configured to compress the first channel matrix according to the Channel State Information (CSI) compression model and the CSI compression parameters to obtain the compressed target channel matrix; wherein, the CSI compression model includes a channel encoder, and the channel encoder includes multiple sub-encoders; different sub-encoders correspond to different CSI compression parameters;

[0018] The first transmitting module is configured to transmit the target channel matrix to the network device so that the network device can obtain the channel quality of the downlink channel based on the target channel matrix.

[0019] According to a fourth aspect of the present disclosure, an apparatus for acquiring channel quality is provided, applied to a network device, the apparatus comprising:

[0020] The second receiving module is configured to receive a target channel matrix sent by the terminal device; the target channel matrix is ​​obtained by the terminal device after compressing the first channel matrix according to the Channel State Information (CSI) compression model and the CSI compression parameters, and the first channel matrix is ​​a matrix obtained by the terminal device based on the pilot signal to characterize the channel quality of the downlink channel.

[0021] The third matrix acquisition module is configured to decompress the target channel matrix according to the Channel State Information (CSI) decompression model and the CSI compression parameters to obtain a third channel matrix; wherein, the CSI decompression model includes a channel decoder, and the channel decoder includes multiple sub-decoders, with different sub-decoders corresponding to different CSI compression parameters;

[0022] The channel quality determination module is configured to determine the channel quality of the downlink channel based on the third channel matrix.

[0023] According to a fifth aspect of the present disclosure, an apparatus for acquiring channel quality is provided, comprising:

[0024] processor;

[0025] Memory used to store processor-executable instructions;

[0026] The processor is configured to perform the steps of the method for acquiring channel quality provided in the first aspect of this disclosure.

[0027] According to a sixth aspect of the present disclosure, an apparatus for obtaining channel quality is provided, comprising:

[0028] processor;

[0029] Memory used to store processor-executable instructions;

[0030] The processor is configured to perform the steps of the method for acquiring channel quality provided in the second aspect of this disclosure.

[0031] According to a seventh aspect of the present disclosure, a computer-readable storage medium is provided that stores computer program instructions thereon, which, when executed by a processor, implement the steps of the method for obtaining channel quality provided in the first aspect of the present disclosure.

[0032] According to an eighth aspect of the present disclosure, a computer-readable storage medium is provided that stores computer program instructions thereon, which, when executed by a processor, implement the steps of the method for obtaining channel quality provided in the second aspect of the present disclosure.

[0033] According to a ninth aspect of the present disclosure, a chip is provided, comprising: a processor and an interface; the processor is configured to read instructions to perform the steps of the method for acquiring channel quality provided in the first aspect of the present disclosure.

[0034] According to a tenth aspect of the present disclosure, a chip is provided, comprising: a processor and an interface; the processor is configured to read instructions to perform the steps of the method for acquiring channel quality provided in the second aspect of the present disclosure.

[0035] The technical solutions provided by the embodiments of this disclosure can include the following beneficial effects: a terminal device receives a pilot signal transmitted by a network device through a downlink channel; obtains a first channel matrix based on the pilot signal; compresses the first channel matrix according to a Channel State Information (CSI) compression model and CSI compression parameters to obtain a compressed target channel matrix; and sends the target channel matrix to the network device so that the network device can determine the channel quality of the downlink channel based on the target channel matrix. The first channel matrix is ​​used to characterize the channel quality of the downlink channel; the CSI compression model can include a channel encoder, which includes multiple sub-encoders; different sub-encoders correspond to different CSI compression parameters; and the target channel matrix is ​​used to instruct the network device to determine the channel quality of the downlink channel. In this way, multiple sub-encoders can adapt to different CSI compression parameters, thereby adaptively obtaining a more accurate target channel matrix and sending it to the network device in scenarios where the CSI compression rate changes, so that the network device obtains a more accurate channel quality, thereby improving data transmission efficiency.

[0036] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0037] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure.

[0038] Figure 1 This is a block diagram illustrating a communication system according to an exemplary embodiment.

[0039] Figure 2 This is a flowchart illustrating a method for obtaining channel quality according to an exemplary embodiment.

[0040] Figure 3 This is a flowchart illustrating a method for obtaining channel quality according to an exemplary embodiment.

[0041] Figure 4 This is a schematic diagram of the structure of a network model for obtaining channel quality according to an exemplary embodiment.

[0042] Figure 5 This is a schematic diagram of a feature converter in a CSI compression model according to an exemplary embodiment.

[0043] Figure 6 This is a schematic diagram of a channel encoder in a CSI compression model according to an exemplary embodiment.

[0044] Figure 7 This is a schematic diagram illustrating a CSI decompression model according to an exemplary embodiment.

[0045] Figure 8 This is a flowchart illustrating a training method for a CSI compression model according to an exemplary embodiment.

[0046] Figure 9 This is a flowchart illustrating a training method for a CSI decompression model according to an exemplary embodiment.

[0047] Figure 10 This is a flowchart illustrating a method for obtaining channel quality according to an exemplary embodiment.

[0048] Figure 11 This is a block diagram illustrating an apparatus for acquiring channel quality according to an exemplary embodiment.

[0049] Figure 12 This is a block diagram illustrating an apparatus for acquiring channel quality according to an exemplary embodiment.

[0050] Figure 13 This is a block diagram illustrating an apparatus for acquiring channel quality according to an exemplary embodiment.

[0051] Figure 14 This is a block diagram illustrating an apparatus for acquiring channel quality according to an exemplary embodiment.

[0052] Figure 15 This is a block diagram illustrating an apparatus for acquiring channel quality according to an exemplary embodiment.

[0053] Figure 16 This is a block diagram illustrating an apparatus for acquiring channel quality according to an exemplary embodiment. Detailed Implementation

[0054] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure as detailed in the appended claims.

[0055] It should be noted that all actions involving the acquisition of signals, information, or data in this disclosure are carried out in compliance with the relevant data protection laws and policies of the country where the location is situated, and with authorization from the owner of the relevant device.

[0056] In this disclosure, terms such as "first" and "second" are used to distinguish similar objects and should not be construed as referring to a specific order or sequence. Furthermore, unless otherwise stated, in the description with reference to the accompanying drawings, the same reference numerals in different drawings denote the same elements.

[0057] In the description of this disclosure, unless otherwise stated, "multiple" means two or more, and other quantifiers are similar; "at least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple; "and / or" is a description of the relationship between related objects, indicating that there can be three relationships, for example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone, where A and B can be singular or plural.

[0058] Although operations are described in a specific order in the accompanying drawings in this disclosure, it should not be construed as requiring these operations to be performed in the specific order or serial order shown, or requiring all of the shown operations to obtain the desired result. In certain environments, multitasking and parallel processing may be advantageous.

[0059] With the application of mMIMO technology, in order to reduce CSI reporting overhead, CSI compression technology based on Discrete Fourier Transform (DFT) can be used, where terminal devices compress CSI before reporting to network devices. However, CSI compression reporting reduces the accuracy of channel quality data obtained by network devices, affecting data transmission efficiency.

[0060] For example, a terminal device can use a pre-trained encoding neural network to compress the CSI and report the compressed CSI to the network device. The network device, in turn, uses a decoding neural network to decompress the compressed CSI to determine the channel quality. However, different CSI compression ratios place different requirements and parameters on the encoding and decoding neural networks. Therefore, in scenarios where the CSI compression ratio changes, using the same encoding and decoding neural networks can lead to significant differences in the accuracy of CSI reporting. Furthermore, retraining the encoding and decoding neural networks when the CSI compression ratio changes is inefficient and cannot adapt to scenarios where the CSI compression ratio frequently changes.

[0061] To address the aforementioned issues, this disclosure provides a method, apparatus, storage medium, and chip for acquiring channel quality.

[0062] The implementation environment of the embodiments of this disclosure is described below.

[0063] Figure 1 This is a schematic diagram illustrating a communication system according to an exemplary embodiment, such as... Figure 1 As shown, the communication system may include terminal device 101 and network device 102. This communication system can be used to support 4G (the 4th Generation) network access technologies, such as Long Term Evolution (LTE) access technology, or 5G (the 5th Generation) network access technologies, such as New Radio Access Technology (New RAT), or other future wireless communication technologies. It should be noted that this communication system can be a communication system using FDD (Frequency Division Duplexing) technology or a communication system using TDD (Time Division Duplexing) technology. Furthermore, in this communication system, the number of network devices and terminal devices can both be one or more. Figure 1 The number of network devices and terminal devices in the communication system shown is merely an adaptive example, and this disclosure does not limit this number.

[0064] Figure 1The network equipment in this disclosure can be used to support terminal access. For example, it can be an evolved Node B (eNB or eNodeB) in LTE; or a base station in a 5G network or a future evolved public land mobile network (PLMN), a Broadband Network Gateway (BNG), an aggregation switch, or a non-3GPP (3rd Generation Partnership Project) access device. Optionally, the network equipment in this disclosure can include various forms of base stations, such as: macro base stations, micro base stations (also known as small stations), relay stations, access points, 5G base stations or future base stations, satellites, transmitting and receiving points (TRPs), transmitting points (TPs), mobile switching centers, and equipment that performs base station functions in device-to-device (D2D), vehicle-to-everything (V2X), and machine-to-machine (M2M) communications. This disclosure does not specifically limit this type of equipment. For ease of description, in all embodiments of this disclosure, the apparatus that provides wireless communication functions for terminal devices is collectively referred to as a network device or a base station.

[0065] Figure 1The terminal device in this disclosure can be an electronic device that provides voice or data connectivity, such as a user equipment (UE), subscriber unit, mobile station, station, or terminal. For example, the terminal device may include smartphones, smart wearable devices, smart speakers, smart tablets, wireless modems, wireless local loop (WLL) stations, PDAs (Personal Digital Assistants), and CPEs (Customer Premise Equipment). With the development of wireless communication technology, any device that can access a communication system, communicate with network devices within the communication system, or communicate with other objects through the communication system can be a terminal device in this disclosure. Examples include terminals and vehicles in intelligent transportation, home appliances in smart homes, electricity meter reading instruments, voltage monitoring instruments, environmental monitoring instruments in smart grids, video surveillance instruments in smart security networks, and cash registers. In this disclosure, the terminal device can communicate with network devices, such as… Figure 1 The network devices communicate with each other. Multiple terminals can also communicate with each other. The terminals can be static or mobile, and this disclosure does not limit them.

[0066] Figure 2 This is a flowchart illustrating a method for obtaining channel quality according to an exemplary embodiment. This method can be applied to terminal devices, such as… Figure 2 As shown, the method may include:

[0067] S201. The terminal device receives the pilot signal sent by the network device through the downlink channel.

[0068] For example, in the communication system described above, the network device can send a pilot signal to the terminal device via a downlink channel. Correspondingly, the terminal device can receive the pilot signal.

[0069] In some embodiments, the pilot signal may include a Channel State Information Reference Signal (CSI-RS).

[0070] S202. The terminal equipment obtains the first channel matrix based on the pilot signal.

[0071] This first channel matrix can be used to characterize the channel quality of the downlink channel.

[0072] For example, the terminal device can perform channel state information (CSI) estimation based on the received pilot signal (e.g., CSI-RS) to obtain a CSI estimation matrix; then, it can obtain a first channel matrix characterizing the downlink channel quality based on the CSI estimation matrix.

[0073] S203. The terminal device compresses the first channel matrix according to the Channel State Information (CSI) compression model and the CSI compression parameters to obtain the compressed target channel matrix.

[0074] The CSI compression model includes a channel encoder, which in turn includes multiple sub-encoders; different sub-encoders correspond to different CSI compression parameters.

[0075] In some embodiments, the terminal device may determine one or more target sub-encoders from a plurality of sub-encoders based on CSI compression parameters, and compress the first channel matrix through the target sub-encoder to obtain the target channel matrix.

[0076] For example, the CSI compression parameter can characterize the CSI compression ratio, and the value of the CSI compression ratio can be any preset value, such as 1 / 2, 1 / 4, 1 / 8, 1 / 16, 1 / 32 or 1 / 64, etc., which are not limited in this disclosure.

[0077] S204. The terminal device sends the target channel matrix to the network device.

[0078] By sending the target channel matrix, network devices can determine the channel quality of the downlink channel based on the target channel matrix.

[0079] Using the above method, the terminal device receives pilot signals transmitted by the network device through the downlink channel; obtains a first channel matrix based on the pilot signals; compresses the first channel matrix according to the Channel State Information (CSI) compression model and CSI compression parameters to obtain a compressed target channel matrix; and sends the target channel matrix to the network device so that the network device can determine the channel quality of the downlink channel based on the target channel matrix. Here, the first channel matrix characterizes the channel quality of the downlink channel; the CSI compression model includes a channel encoder, which includes multiple sub-encoders; different sub-encoders correspond to different CSI compression parameters; and the target channel matrix instructs the network device to determine the channel quality of the downlink channel. In this way, by using multiple sub-encoders to adapt to different CSI compression parameters, a more accurate target channel matrix can be adaptively obtained and sent to the network device even when the CSI compression rate changes, so that the network device can obtain a more accurate channel quality, thereby improving data transmission efficiency.

[0080] In some embodiments, the terminal device can obtain the first channel matrix described above in the following manner:

[0081] First, the spatial channel matrix is ​​measured based on the pilot signal.

[0082] Secondly, the spatial channel matrix is ​​transformed into the angular delay domain channel matrix through discrete Fourier transform.

[0083] Next, the first channel matrix is ​​determined based on the channel matrix in the angle delay domain.

[0084] For example, the above communication system can employ mMIMO technology based on OFDM (Orthogonal Frequency Division Multiplexing) with N subcarriers. s The number of antennas used for mMIMO in network devices can be N. t The aforementioned pilot signal may include a Channel State Information Reference Signal (CSI-RS). Taking this communication system as an example, the above steps are illustrated below:

[0085] First, the terminal device can measure (or estimate) the spatial channel matrix H based on the received CSI-RS. The size of the spatial channel matrix H can be N. s ×N t The spatial channel matrix H can characterize the channel quality of each subcarrier for each antenna.

[0086] Secondly, the terminal device can transform the spatial domain channel matrix into the angle delay domain channel matrix through discrete Fourier transform. For example, the angle delay domain channel matrix can be obtained through the following formula (1):

[0087] H a =F d HF a (1)

[0088] Where H represents the aforementioned spatial channel matrix, and the size of H can be N. s ×N t H a F represents the transformed angular delay domain channel matrix. d and F a Indicates a size of N s ×N s N t ×N t The discrete Fourier transform matrix, N s N represents the number of subcarriers corresponding to mMIMO. t This indicates the number of antennas corresponding to mMIMO.

[0089] Finally, the first channel matrix is ​​determined based on the principal value of the channel matrix in the angle delay domain.

[0090] It should be noted that, due to the influence of multipath delay, the aforementioned angular delay domain channel matrix H... a Only in the first N c After extracting the principal values ​​from the rows, the principal channel matrix in the angle delay domain is obtained. The size of this principal channel matrix can be N. c ×N t .

[0091] Furthermore, by separating the real and imaginary parts of the principal value channel matrix, the first channel matrix can be obtained. Where c is the real and imaginary part dimension of the channel, for example, c can be 2. This first channel matrix can be used as the input of the CSI compression model to obtain the target channel matrix.

[0092] In this way, the first channel matrix representing the downlink channel quality can be obtained from the pilot signal using the above method.

[0093] The CSI compression parameters mentioned above can be parameters received by the terminal device from the network device, or they can be parameters preset by the terminal device.

[0094] In some embodiments, the CSI compression parameters may be determined by the terminal device based on parameters received from the network device. For example, the terminal device may receive first compression parameters sent by the network device and determine the CSI compression parameters based on these first compression parameters. For instance, the terminal device may receive the first compression parameters via RRC (Radio Resource Control) signaling (e.g., broadcast signaling or proprietary signaling specific to the terminal device).

[0095] In some embodiments, the network device may pre-set the value of the CSI compression parameter, determine the first compression parameter based on the CSI compression parameter, and send the first compression parameter to the terminal device.

[0096] For example, the CSI compression parameter and the first compression parameter can be determined by a pre-set correspondence between the first compression parameters. For instance, when the CSI compression parameter represents the CSI compression ratio, the first compression parameter corresponding to 1 / 2 of the CSI compression parameter can be 1, the first compression parameter corresponding to 1 / 4 of the CSI compression parameter can be 2, the first compression parameter corresponding to 1 / 8 of the CSI compression parameter can be 3, and so on for other values.

[0097] In this way, after receiving the first compression parameter sent by the network device, the terminal device can determine the CSI compression parameter based on the first compression parameter.

[0098] In some embodiments, the network device may update the value of the CSI compression parameter and determine a new first compression parameter based on the updated CSI compression parameter; similarly, the terminal device may also determine a new CSI compression parameter based on the new first compression parameter upon receiving it.

[0099] In other embodiments, the value of the CSI compression parameter can be preset by the terminal device. For example, it can be a preset parameter value of the terminal device, or it can be a parameter value set by the terminal device according to user input.

[0100] In some embodiments, the terminal device can determine a second compression parameter based on a pre-set value of the CSI compression parameter, and send the second compression parameter to the network device to negotiate and use the same CSI compression parameter. For example, the terminal device can send the second compression parameter to the network device via RRC signaling.

[0101] Similarly, the CSI compression parameter and the second compression parameter can be determined by a pre-set correspondence between the second compression parameters.

[0102] In some embodiments, the terminal device can determine the value of the CSI compression parameter based on its own device parameters, which may include one or more of the following: the protocol version of the terminal device, the signal quality of the terminal device, the distance between the terminal device and the network device, the uplink data volume of the terminal device, and the downlink data volume of the terminal device. The signal quality of the terminal device may include RSRP (Reference Signal Receiving Power) or SINR (Signal to Interference plus Noise Ratio).

[0103] In some embodiments, the terminal device may update the value of the CSI compression parameter and determine a new second compression parameter based on the updated CSI compression parameter. For example, if the terminal device determines that the aforementioned device parameters have changed, it may update the value of the CSI compression parameter based on the changed device parameters. Similarly, the network device may also determine a new CSI compression parameter based on the new second compression parameter upon receiving it.

[0104] In some other embodiments, the CSI compression parameter may include a third compression parameter pre-set in the terminal device and the network device, respectively, and the values ​​of the third compression parameters pre-set in the terminal device and the network device may be the same.

[0105] Figure 3This is a flowchart illustrating a method for obtaining channel quality according to an exemplary embodiment. This method can be applied to network devices, such as… Figure 3 As shown, the method may include:

[0106] S301, The network device receives the target channel matrix sent by the terminal device.

[0107] The target channel matrix is ​​obtained by the terminal device after compressing the first channel matrix according to the Channel State Information (CSI) compression model and the CSI compression parameters. The first channel matrix is ​​a matrix obtained by the terminal device based on the pilot signal to characterize the channel quality of the downlink channel.

[0108] In some embodiments, the network device may transmit a pilot signal via a downlink channel so that a terminal device can receive the pilot signal and obtain a first channel matrix. For example, the pilot signal may include a Channel State Information Reference Signal (CSI-RS).

[0109] S302. The network device decompresses the target channel matrix according to the Channel State Information (CSI) decompression model and CSI compression parameters to obtain the third channel matrix.

[0110] The CSI decompression model can include a channel decoder, which includes multiple sub-decoders, each corresponding to different CSI compression parameters.

[0111] In some embodiments, the network device may determine one or more target sub-decoders from a plurality of sub-decoders based on CSI compression parameters, and compress the first channel matrix through the target sub-decoder to obtain the target channel matrix.

[0112] For example, the CSI compression parameter can characterize the CSI compression ratio, and the value of the CSI compression ratio can be any preset value, such as 1 / 2, 1 / 4, 1 / 8, 1 / 16, 1 / 32 or 1 / 64, etc., which are not limited in this disclosure.

[0113] S303. The network device determines the channel quality of the downlink channel based on the third channel matrix.

[0114] Using the above method, the target channel matrix sent by the receiving terminal device is received; the target channel matrix is ​​decompressed according to the Channel State Information (CSI) decompression model and CSI compression parameters to obtain a third channel matrix; and the channel quality of the downlink channel is determined based on the third channel matrix. The target channel matrix is ​​obtained by the terminal device after compressing the first channel matrix according to the CSI compression model and CSI compression parameters. The first channel matrix is ​​a matrix obtained by the terminal device based on pilot signals to characterize the channel quality of the downlink channel. The CSI decompression model may include a channel decoder, which includes multiple sub-decoders, each corresponding to different CSI compression parameters. In this way, the network device can adapt to different CSI compression parameters through multiple sub-decoders, thus adaptively obtaining a more accurate third channel matrix even when the CSI compression rate changes. Based on this third channel matrix, a more accurate channel quality can be obtained, and the network device can determine the modulation and coding strategy corresponding to the downlink channel based on this channel quality, thereby improving data transmission efficiency.

[0115] In some embodiments, the network device may use the sub-decoder corresponding to the CSI compression parameters as the target sub-decoder; decompress the target channel matrix using the target sub-decoder to obtain the fourth channel matrix; and determine the third channel matrix based on the fourth channel matrix.

[0116] For example, the sub-decoder may include a decompression layer that may correspond to the compression layer of the sub-encoder in the channel encoder of the terminal device.

[0117] In some embodiments, the CSI decompression model may further include a CSI reconstruction module, and the network device may input the aforementioned fourth channel matrix into the CSI reconstruction module to obtain a third channel matrix.

[0118] For example, the CSI reconstruction module may include a convolutional neural network (CNN).

[0119] In some embodiments, the value of the CSI compression parameter can be preset by the network device. For example, it can be a preset parameter value of the network device, or it can be a parameter value set by the network device according to user input.

[0120] In some embodiments, the network device may determine a first compression parameter based on the CSI compression parameter and send the first compression parameter to the terminal device to instruct the terminal device to determine the CSI compression parameter based on the first compression parameter. For example, the network device may send the first compression parameter via RRC (Radio Resource Control) signaling (e.g., broadcast signaling or proprietary signaling for the terminal device).

[0121] For example, the CSI compression parameter and the first compression parameter can be determined through a pre-set correspondence between the first compression parameters. For instance, when the CSI compression parameter represents the CSI compression ratio, the first compression parameter corresponding to a value of 1 / 2 of the CSI compression parameter can be 1, the first compression parameter corresponding to a value of 1 / 4 of the CSI compression parameter can be 2, the first compression parameter corresponding to a value of 1 / 8 of the CSI compression parameter can be 3, and so on. Thus, after receiving the first compression parameter sent by the network device, the terminal device can determine the CSI compression parameter based on the first compression parameter.

[0122] In some embodiments, the network device can set different CSI compression parameter values ​​for different terminal devices. For example, the CSI compression parameter value can be determined based on the device parameters corresponding to the terminal device. These device parameters may include one or more of the following: the protocol version of the terminal device, the signal quality of the terminal device, the distance between the terminal device and the network device, the uplink data volume of the terminal device, and the downlink data volume of the terminal device. The signal quality of the terminal device may include RSRP (Reference Signal Receiving Power) or SINR (Signal to Interference plus Noise Ratio).

[0123] In other embodiments, the network device can update the value of the CSI compression parameter and determine a new first compression parameter based on the updated CSI compression parameter. For example, the network device can obtain the device parameters corresponding to the terminal device, and if it is determined that the device parameters have changed, it can update the value of the CSI compression parameter.

[0124] Similarly, upon receiving new first compression parameters, the terminal device can also determine new CSI compression parameters based on the new first compression parameters.

[0125] In other embodiments, the CSI compression parameters may be determined by the network device based on parameters received from the terminal device. For example, the network device may receive a second compression parameter sent by the terminal device and determine the CSI compression parameters based on this second compression parameter. For instance, the network device may receive the second compression parameter via RRC signaling.

[0126] For example, the terminal device can pre-set the value of the CSI compression parameter, determine the second compression parameter based on the CSI compression parameter, and send the second compression parameter to the network device. Similarly, the CSI compression parameter and the second compression parameter can be determined through a pre-set correspondence between the second compression parameters. In this way, after receiving the second compression parameter sent by the terminal device, the network device can determine the CSI compression parameter based on the second compression parameter.

[0127] In some embodiments, the terminal device may update the value of the CSI compression parameter and determine a new second compression parameter based on the updated CSI compression parameter; similarly, the network device may also determine a new CSI compression parameter based on the new second compression parameter upon receiving it.

[0128] In other embodiments, the CSI compression parameter can also be a third compression parameter preset in the terminal device and the network device, and the values ​​of the third compression parameters preset in the terminal device and the network device can be the same.

[0129] In some embodiments, the CSI compression model and the CSI decompression model described above can together form a network model for acquiring channel quality. The following detailed description of this network model for acquiring channel quality is provided in conjunction with the accompanying drawings.

[0130] Figure 4 This is a schematic diagram of the structure of a network model for obtaining channel quality according to an exemplary embodiment.

[0131] according to Figure 4 As shown, in some embodiments, the network model for acquiring channel quality may include a Channel State Information (CSI) compression model 41 and a Channel State Information (CSI) decompression model 42. The CSI compression model 41 may be deployed on... Figure 1 In the communication system shown, the terminal device can run the CSI compression model 41 through software, hardware, or a combination of both; the CSI decompression model 42 can be deployed on... Figure 1 On network devices (e.g., base stations) in the communication system shown, the network devices can run the CSI decompression model 42 through software, hardware, or a combination of software and hardware.

[0132] The CSI compression model 41 can compress the input first channel matrix according to the CSI compression parameters and output the target channel matrix (which can also be called a codeword). The terminal device can send the target channel matrix to the network device. Correspondingly, the network device can input the received target channel matrix into the CSI decompression model 42. The CSI decompression model 42 can decode (or decompress) the target channel matrix according to the CSI compression parameters and output the third channel matrix. The network device can determine the channel quality of the downlink channel according to the third channel matrix.

[0133] according to Figure 4 As shown, in some embodiments, the CSI compression model 41 may include a channel encoder 411, which can encode the input channel matrix according to the CSI compression parameters to obtain the target channel matrix.

[0134] Furthermore, the channel encoder 411 may include multiple sub-encoders, each corresponding to a different CSI compression parameter. Thus, the terminal device can use the sub-encoder corresponding to the CSI compression parameter as the first target sub-encoder; the first target sub-encoder is used to compress the first channel matrix to obtain the target channel matrix. This target channel matrix can also be referred to as a codeword.

[0135] For example, when the CSI compression parameter includes the CSI compression rate, the channel encoder 411 may include sub-encoder 1 (corresponding to CSI compression parameter 1 / 2), sub-encoder 2 (corresponding to CSI compression parameter 1 / 4), sub-encoder 3 (corresponding to CSI compression parameter 1 / 8), sub-encoder 4 (corresponding to CSI compression parameter 1 / 16), sub-encoder 5 (corresponding to CSI compression parameter 1 / 32), sub-encoder 5 (corresponding to CSI compression parameter 1 / 64), etc.

[0136] according to Figure 4 As shown, in some embodiments, the CSI compression model may include a channel encoder 411 and a feature converter (also called a feature optimizer or joint feature optimizer) 412. The feature converter 412 can take the first channel matrix as input and extract key features from it to obtain a second channel matrix representing key CSI features. This second channel matrix can be used as input to the channel encoder 411, which can compress the second channel matrix according to CSI compression parameters to obtain a target channel matrix. For example, the terminal device can use the sub-encoder corresponding to the CSI compression parameters as a second target sub-encoder, and then compress the second channel matrix using this second target sub-encoder to obtain the target channel matrix. This target channel matrix can also be called a codeword.

[0137] Figure 5 This is a schematic diagram illustrating a feature converter in a CSI compression model according to an exemplary embodiment. Figure 5 As shown, the feature converter 412 includes a feature extraction network 4121, an attention mechanism network 4122, and a feature restoration network 4123.

[0138] In some embodiments, the terminal device can input the first channel matrix into a feature converter to extract key features from the first channel matrix and obtain a second channel matrix that characterizes the key features of CSI.

[0139] For example, the second channel matrix can be obtained through the following steps:

[0140] S51. Input the first channel matrix into the feature extraction network to obtain multiple first feature maps.

[0141] For example, the feature extraction network will extract the first channel matrix H c Converted into the first feature map Where f represents the number of first feature maps extracted. f can be any positive integer greater than 1.

[0142] In some embodiments, the feature extraction network may include two-dimensional convolutional layers with kernel sizes of f×m×m, where f represents the number of first feature maps and m×m represents the length and width of the convolutional window used by the kernel. The feature extraction network may employ a two-dimensional normalization layer to normalize the output of the convolutional layers. The activation function of the feature extraction network may include Sigmoid, ReLU, LeakyReLU, PReLU, or ELU.

[0143] For example, the activation function can be a LeakyReLU (leaky linear rectified function) activation function, which can include the following formula (2):

[0144]

[0145] Where x represents the input vector, such as the first feature map; γ represents the preset coefficient, which can be any value less than 1, for example, 0.3; LeakyReLU(x) represents the output value of the LeakyReLU activation function.

[0146] S52. Input multiple first feature maps into the attention mechanism network to obtain second feature maps.

[0147] The second feature map may include key feature information from the first feature map.

[0148] In some embodiments, a max pooling operation can be performed on multiple first feature maps using an attention mechanism network to obtain a max pooled feature map; an average pooling operation can be performed on multiple first feature maps using an attention mechanism network to obtain an average pooled feature map; and then, a second feature map can be determined based on the max pooled feature map and the average pooled feature map.

[0149] It should be noted that among the multiple first feature maps extracted by the feature extraction network, some first feature maps contain feature information that greatly helps CSI reconstruction and can be called "key feature maps." Others contain feature information that has almost no impact on CSI reconstruction and can be called "non-essential feature maps." This attention mechanism network can extract key feature maps from multiple feature maps, thus enabling the codewords generated by the subsequent encoder to contain more key features. Example:

[0150] This attention mechanism network can perform max pooling on the first feature map F to obtain a max pooled feature map M∈ For example, max pooling can be performed using the following formula (3):

[0151] m i =max{F 1,i ,F 2,i ,F 3,i ,…,F f,i}, i = {1, 2, ..., N c N t} (3)

[0152] Where, m i Let F represent the i-th element in the max-pooled feature map M, f represent the number of first feature maps, and F represent the number of first feature maps. 1,i F represents the i-th element in the first feature map. 2,i F represents the i-th element in the second first feature map. 2,i F represents the i-th element in the second first feature map. f,i N represents the i-th element in the f-th first feature map. c ×N t This indicates the size of the first channel matrix.

[0153] In this way, each element in the max-pooling feature map M is composed of the maximum element at the corresponding position in multiple first feature maps.

[0154] This attention mechanism network can also perform average pooling on the first feature map F to obtain an average pooled feature map. For example, average pooling can be performed using the following formula (4):

[0155]

[0156] Among them, v i F represents the i-th element in the average pooling feature map V, f represents the number of first feature maps, and F 1,i F represents the i-th element in the first feature map. 2,i F represents the i-th element in the second first feature map. 2,i F represents the i-th element in the second first feature map. f-1,i F represents the i-th element in the (f-1)-th first feature map. f,i N represents the i-th element in the f-th first feature map. c ×N t This indicates the size of the first channel matrix.

[0157] In this way, each element in the average pooling feature map V is composed of the mean of all elements at the corresponding positions in the multiple first feature maps.

[0158] Furthermore, the second feature map can be determined based on the max pooling feature map and the average pooling feature map.

[0159] In some embodiments, the max pooling feature map and the average pooling feature map can be input into the fusion subnetwork to obtain the fused feature map; and the second feature map can be calculated based on the fused feature map and the first feature map.

[0160] In some embodiments, the max pooling feature map M and the average pooling feature map V can be concatenated to obtain a joint feature map. The joint feature map C is processed by a fusion network to obtain the fused feature map.

[0161] In some embodiments, the fusion network may employ two-dimensional convolutional layers with kernel sizes of m×n×n. The fusion network may also include two-dimensional normalization layers and activation functions, which may include sigmoid activation functions.

[0162] Then, multiplying the fused feature map D with the first feature matrix F yields the second feature map. This second feature map F′ can also be called the optimized feature map.

[0163] In this way, the feature information in the key feature map is highlighted in the second feature map F′, while the feature information in the non-essential feature map is weakened, thus enabling the representation of the key features of CSI.

[0164] S53. Input the second feature map into the feature restoration network to obtain the second channel matrix.

[0165] For example, the second feature map F′ described above can be restored to the second channel matrix using this feature restoration network.

[0166] In some embodiments, the feature restoration network may include a two-dimensional convolutional layer, a two-dimensional normalization layer, and an activation function. The activation function may be the LeakyReLU activation function or other activation functions in related technologies, and this disclosure does not limit it.

[0167] In some embodiments, the feature restoration network and the feature extraction network may employ the same activation function to avoid feature distortion.

[0168] Thus, the second channel matrix H obtained in this way e It highlights key feature information and weakens unnecessary feature information, thus enabling the characterization of key CSI features.

[0169] Figure 6 This is a schematic diagram of a channel encoder in a CSI compression model according to an exemplary embodiment. Figure 6 As shown, the channel encoder 411 may include multiple sub-encoders, such as sub-encoder 1, sub-encoder 2, ..., sub-encoder T-1 and sub-encoder T in the figure. Where T is the number of sub-encoders.

[0170] In some embodiments, the channel encoder may first input the second channel matrix H e Preprocessing can be performed, such as dimensional transformation, so that the dimension of the transformed second channel matrix can be [missing information]. The second channel matrix after dimensional transformation is input into the above sub-encoder for processing.

[0171] In some embodiments, the CSI compression parameters may include a set of preset CSI compression rates, the set of which may include σ = {σ1, σ2, ..., σ...} T The preset CSI compression ratios in the set are sorted in descending order from largest to smallest. For each compression ratio σ in the set... j j = {1,2,…,T}, each of the channel encoders has a sub-encoder corresponding to the preset CSI compression rate, that is, the number of sub-encoders is equal to the number of preset CSI compression rates.

[0172] For example, the preset CSI compression ratio σ1 corresponding to sub-encoder 1 can be 1 / 2, the preset CSI compression ratio σ2 corresponding to sub-encoder 2 can be 1 / 4, the preset CSI compression ratio σ3 corresponding to sub-encoder 3 can be 1 / 8, the preset CSI compression ratio σ4 corresponding to sub-encoder 4 can be 1 / 16, the preset CSI compression ratio σ5 corresponding to sub-encoder 5 can be 1 / 32, the preset CSI compression ratio σ6 corresponding to sub-encoder 6 can be 1 / 64, and so on. It should be noted that the preset CSI compression ratio values ​​here are just examples, and this disclosure does not limit the specific values.

[0173] In some embodiments, the set of preset CSI compression rates may include σ = {1 / 4, 1 / 16, 1 / 32, 1 / 64}, i.e., the number of sub-encoders T = 4, and the preset CSI compression rates in the set are arranged in descending order. For each preset CSI compression rate in the set, there is a corresponding sub-encoder in the adaptive encoder that matches it, i.e., the number of sub-encoders is equal to the number of dynamic compression rates. For example, the preset CSI compression rate σ1 corresponding to sub-encoder 1 can be 1 / 4, the preset CSI compression rate σ2 corresponding to sub-encoder 2 can be 1 / 16, the preset CSI compression rate σ3 corresponding to sub-encoder 3 can be 1 / 32, and the preset CSI compression rate σ4 corresponding to sub-encoder 4 can be 1 / 64.

[0174] In some embodiments, each sub-encoder may include a compression layer, which may include a fully connected layer. The size of the compression layer of each sub-encoder can be determined based on CSI compression parameters (e.g., a preset CSI compression rate) and the size of the first channel matrix; for example, the size of the first channel matrix is ​​c×N. c ×N t If the preset CSI compression rate is σ1, then the size of the compression layer of the sub-encoder corresponding to σ1 is (c×N). c ×N t )×d1, where d1 can be c×N c ×N t ×σ1.

[0175] For example, if the size of the first channel matrix is ​​2×32×32 and the preset CSI compression rate is 1 / 4, then the size of the compression layer of the sub-encoder corresponding to the preset CSI compression rate can be 2048×512.

[0176] In other embodiments, each sub-encoder may include a compression layer and an encoding switch. In practical use, the target sub-encoder (e.g., the first target sub-encoder or the second target sub-encoder in the above embodiments) corresponding to the CSI compression parameters (e.g., CSI compression ratio) can be determined, and the encoding switch of the target sub-encoder can be closed while the switches of other sub-encoders are opened. Thus, the compression layer of the target sub-encoder can be used to compress the input second channel matrix and output the undetermined channel matrix of the target sub-encoder. The undetermined channel matrix can be used as the target channel matrix M.

[0177] In some embodiments, the preset CSI compression rate with the largest value is taken as the maximum CSI compression rate, and the sub-encoder corresponding to the maximum CSI compression rate (e.g.) Figure 5 The compression layer of sub-encoder 1) is used as the maximum compression layer. The output of the maximum compression layer can be used as the input of the compression layer of other sub-encoders, which can improve the compression efficiency.

[0178] For example, if the preset CSI compression rate σ1 is the maximum compression rate, the size of the compression layer 1 of the sub-encoder 1 corresponding to σ1 can be (c×N). c ×N t )×d1, where d1 can be c×N c ×N t ×σ1. The output of compression layer 1 can be used as the input of the compression layers corresponding to other sub-encoders, and the size of the compression layers corresponding to other sub-encoders can be d1×d. k d k It can be c×N c ×N t ×σ k Where k represents the sub-encoder number, σ k c×N represents the preset CSI compression ratio corresponding to the k-th sub-encoder. c ×N t This indicates the size of the first channel matrix.

[0179] For example, if the size of the first channel matrix is ​​2×32×32 and the maximum compression ratio σ1 is 1 / 2, then the size of the compression layer 1 of the sub-encoder 1 corresponding to σ1 can be 2048×1024; if the preset CSI compression ratio σ2 corresponding to the sub-encoder 2 is 1 / 4, then the size of the compression layer 2 of the sub-encoder 2 can be 1024×512; if the preset CSI compression ratio σ3 corresponding to the sub-encoder 3 is 1 / 16, then the size of the compression layer 3 of the sub-encoder 3 can be 1024×256.

[0180] In this way, Figure 6 The compression layers in sub-encoders 2 to T can further reduce dimensionality.

[0181] Figure 7 This is a schematic diagram illustrating a CSI decompression model according to an exemplary embodiment. Figure 7 As shown, the CSI decompression model 42 may include a channel decoder 421.

[0182] In some embodiments, the channel decoder 421 may consist of multiple sub-decoders, each corresponding to different CSI compression parameters. Each sub-decoder can decompress the received target channel matrix through a decompression layer to obtain a third channel matrix. This decompression layer may include a fully connected layer.

[0183] like Figure 7 As shown, the channel decoder 421 may include multiple sub-decoders, such as sub-decoder 1, sub-decoder 2, ..., sub-decoder T-1, and sub-decoder T. Here, T represents the number of sub-decoders.

[0184] In some embodiments, the CSI compression parameters may include a set of preset CSI compression rates, the set of which may include σ = {σ1, σ2, ..., σ...} T The preset CSI compression ratios in the set are sorted in descending order from largest to smallest. For each compression ratio σ in the set... j , j={1,2,…,T}, each of the channel decoders has a sub-decoder corresponding to the preset CSI compression rate, that is, the number of sub-decoders is equal to the number of preset CSI compression rates.

[0185] For example, the preset CSI compression ratio σ1 corresponding to sub-decoder 1 can be 1 / 2, the preset CSI compression ratio σ2 corresponding to sub-decoder 2 can be 1 / 4, the preset CSI compression ratio σ3 corresponding to sub-decoder 3 can be 1 / 8, the preset CSI compression ratio σ4 corresponding to sub-decoder 4 can be 1 / 16, the preset CSI compression ratio σ5 corresponding to sub-decoder 5 can be 1 / 32, the preset CSI compression ratio σ6 corresponding to sub-decoder 6 can be 1 / 64, and so on. It should be noted that the preset CSI compression ratio values ​​here are just examples, and this disclosure does not limit the specific values.

[0186] In some embodiments, each sub-decoder may include a decompression layer and a decoding switch. In practical use, a target sub-decoder corresponding to a CSI compression parameter (e.g., CSI compression rate) can be determined, and the decoding switch of the target sub-decoder can be closed while the switches of other sub-decoders are opened. Thus, the decompression layer of the target sub-decoder can be used to decompress the input target channel matrix and output the undetermined channel matrix corresponding to the target sub-decoder. The undetermined channel matrix can be used as the fourth channel matrix, and the third channel matrix can be further determined based on the fourth channel matrix.

[0187] In some embodiments, the CSI decompression model may further include a CSI reconstruction module 422. This CSI reconstruction module may include CNNs (Convolutional Neural Networks). For example, the CSI reconstruction module includes two CNN networks, each CNN network comprising five convolutional layers. The kernel sizes of each convolutional layer are c×k×k, f1×l×l, f2×l×l, f2×n×n, and c×m×m, respectively (f1, f2, k, l, m, n are all preset values, which can be preset differently depending on the convolutional layer). The stride of each convolutional layer is t, and both can employ a normalization layer and the LeakyReLU activation function. Then, the output element values ​​of the second CNN module are mapped to the [0,1] interval through a Sigmoid activation function layer. In this way, the CSI reconstruction module can output the third channel matrix corresponding to the target channel matrix.

[0188] It should be noted that this CSI reconstruction module can output different third channel matrices for different CSI compression parameters (e.g., CSI compression ratio). For example, the set of preset CSI compression ratios in the CSI compression parameters may include σ = {σ1, σ2, ..., σ...} T}, then the corresponding output third channel matrix can be respectively

[0189] In this way, through this CSI decompression model, network devices can decompress the received target channel matrix to obtain a third channel matrix, so as to determine the channel quality of the downlink channel based on the third channel matrix.

[0190] In some embodiments, the CSI compression model and CSI decompression model can be obtained through offline training. For example, the CSI compression model and CSI decompression model can be jointly trained to obtain the parameters of the CSI compression model and CSI decompression model, so that the CSI compression model and CSI decompression model can be matched.

[0191] The training of the above model can be performed on terminal devices or network devices, as explained below with reference to the accompanying drawings.

[0192] Figure 8 This is a flowchart illustrating a training method for a CSI compression model according to an exemplary embodiment. This training method can be applied to terminal devices. Figure 8 As shown, the training method may include:

[0193] S801, The terminal device acquires the first sample channel matrix for training.

[0194] The first sample channel matrix is ​​a matrix obtained by the terminal device based on the received pilot signal, used to characterize the downlink channel quality.

[0195] In some embodiments, in an FDD downlink mMIMO system, N can be configured at half-wavelength intervals using a ULA (Uniform Linear Array) configuration on the network device (e.g., base station) side. t =32 antennas, configured as a single antenna on the terminal device. Using the COST2100 channel model, 150,000 spatial CSI matrix samples were generated in a 5.3GHz indoor microcell scenario, and a training set of 100,000 samples, a validation set of 30,000 samples, and a test set of 20,000 samples were obtained. The training set can be used as the first sample channel matrix mentioned above.

[0196] S802. The terminal device trains the first target network model based on the first sample channel matrix to obtain the CSI compression model.

[0197] The first target network model includes a first target compression model and a first target decompression model. The network structure of the first target compression model is the same as that of the CSI compression model. For example, both can include a channel encoder, which includes multiple sub-encoders, and different sub-encoders correspond to different CSI compression parameters. The first target decompression model includes a channel decoder, which includes multiple sub-decoders, and different sub-decoders correspond to different CSI compression parameters.

[0198] In some embodiments, the model structure of the first target compression model described above can be the same as... Figure 4 The CSI compression model shown is the same; for example, this first target compression model may include a channel encoder and a feature converter, the structure of which may be as follows: Figure 5 As shown, the structure of the channel encoder can be as follows: Figure 6 As shown; the model structure of the first target decompression model described above can be compared with... Figure 7 The CSI decompression model shown is the same, and the structure of the above model will not be described again here.

[0199] In some embodiments, the first model training step can be executed cyclically until the trained first target network model is determined to meet the first preset stopping iteration condition based on the first sample channel matrix and the first predicted channel matrix, and the first target compression model in the trained first target network model is used as the CSI compression model.

[0200] The first predicted channel matrix is ​​the matrix output after the first sample channel matrix is ​​input into the first target network model.

[0201] For example, the parameters of the first compression model corresponding to the first target compression model can be used as the model parameters of the CSI compression model.

[0202] The first model training step may include:

[0203] S81. Input the first sample channel matrix into the first target compression model, and compress the first sample channel matrix through multiple sub-encoders to obtain the first target sample channel matrix.

[0204] S82. Input the first target sample channel matrix into the first target decompression model, and after compressing the first sample channel matrix through multiple sub-decoders, obtain the first prediction channel matrix.

[0205] S83. If the first target network model does not meet the first preset stopping iteration condition based on the first sample channel matrix and the first predicted channel matrix, the first loss value is determined based on the first sample channel matrix and the first predicted channel matrix, the parameters of the first target network model are updated based on the first loss value, the trained first target network model is obtained, and the trained first target network model is used as the new first target network model.

[0206] It should be noted that the first preset stopping iteration condition mentioned above can be determined based on the loss function used during training.

[0207] In some embodiments, the encoding switches corresponding to all sub-encoders can be closed, the decoding switches corresponding to all sub-decoders can also be closed, and joint optimization can be performed on the training error under multiple CSI compression parameters. The loss function used during training can include the following formula (5):

[0208]

[0209] Where σ1 represents the first preset CSI compression ratio, σ2 represents the first preset CSI compression ratio, and σ T This represents the T-th preset CSI compression rate, where T represents the number of preset CSI compression rates in the CSI compression parameters (it can also represent the number of sub-encoders or sub-decoders). σ1 represents the loss value corresponding to the preset CSI compression rate σ1, and Loss represents the first loss value of the first sample channel matrix and the first predicted channel matrix.

[0210] In this way, by jointly optimizing the training errors corresponding to multiple CSI compression parameters (such as preset CSI compression ratios), the network can optimize the training errors corresponding to all preset CSI compression ratios in a single training cycle, thereby improving the network's adaptability to dynamic changes in CSI compression ratios. After training, the parameters of the CSI compression model can be obtained.

[0211] In some embodiments, the terminal device can also obtain the parameters of the first target decompression model after training through the above training process. The parameters of the first target decompression model after training can be used as the parameters of the CSI decompression model on the network device side.

[0212] In some embodiments, after determining the model parameters through the training method, the terminal device can obtain the first decompression model parameters corresponding to the first target decompression model in the trained first target network model; and send the first decompression model parameters to the network device to instruct the network device to determine the CSI decompression model based on the first decompression model parameters. The CSI decompression model is used by the network device to determine the channel quality of the downlink channel based on the target channel matrix.

[0213] For example, the terminal device can send the first decompression model parameters to the network device via signaling or data packets.

[0214] In other embodiments of this disclosure, the above training method can be executed on a network device, and the terminal device can receive the second compression model parameters sent by the network device; and determine the CSI compression model based on the second compression model parameters.

[0215] For example, the second compression model parameters can be used as model parameters for the CSI compression model.

[0216] For example, the terminal device can receive the second compression model parameters sent by the network device through signaling or data packets, and determine the parameters corresponding to the CSI compression model based on the second compression model parameters.

[0217] Figure 9 This is a flowchart illustrating a training method for a CSI decompression model according to an exemplary embodiment, which can be applied to network devices. Figure 9 As shown, the training method may include:

[0218] S901, The network device acquires the second sample channel matrix for training.

[0219] The second sample channel matrix can be a matrix obtained by the terminal device based on the received pilot signal to characterize the downlink channel quality.

[0220] In some embodiments, in an FDD downlink mMIMO system, N can be configured in a ULA manner with half-wavelength spacing on the network device (e.g., base station) side. t =32 antennas, with a single antenna configured on the terminal device. Using the COST2100 channel model, 150,000 spatial CSI matrix samples are generated in a 5.3GHz indoor microcell scenario, and obtained as a training set of 100,000 samples, a validation set of 30,000 samples, and a test set of 20,000 samples. The training set can be used as the first sample channel matrix.

[0221] S902. The network device trains the second target network model based on the second sample channel matrix to obtain the CSI decompression model.

[0222] The second target network model includes a second target compression model and a second target decompression model. The network structure of the second target decompression model is the same as that of the CSI decompression model. For example, both include a channel decoder, which may include multiple sub-decoders, and different sub-decoders correspond to different CSI compression parameters. The second target compression model includes a channel encoder, which includes multiple sub-encoders, and different sub-encoders correspond to different CSI compression parameters.

[0223] In some embodiments, the model structure of the second target compression model described above can be the same as... Figure 4 The CSI compression model shown is the same; for example, this second target compression model may include a channel encoder and a feature converter, the structure of which may be as follows: Figure 5 As shown, the structure of the channel encoder can be as follows: Figure 6 As shown; the model structure of the above-mentioned second objective decompression model can be compared with... Figure 7 The CSI decompression model shown is the same, and the structure of the above model will not be described again here.

[0224] In some embodiments, the second model training step can be executed cyclically until the trained second target network model satisfies the second preset stopping iteration condition based on the second sample channel matrix and the second predicted channel matrix. The second target decompression model in the trained second target network model is then used as the CSI decompression model. The second predicted channel matrix is ​​the matrix output after the second sample channel matrix is ​​input into the second target network model.

[0225] The second model training step may include:

[0226] S91. Input the second sample channel matrix into the second target compression model, and compress the second sample channel matrix through multiple sub-encoders to obtain the second target sample channel matrix.

[0227] S92. Input the second target sample channel matrix into the second target decompression model. After decompressing the second target sample channel matrix through multiple sub-decoders, the second prediction channel matrix is ​​obtained.

[0228] S93. If the second target network model does not meet the second preset stopping iteration condition based on the second sample channel matrix and the second predicted channel matrix, the second loss value is determined based on the second sample channel matrix and the second predicted channel matrix. The parameters of the second target network model are updated based on the second loss value to obtain the trained second target network model. The trained second target network model is then used as the new second target network model.

[0229] It should be noted that the second preset stopping iteration condition mentioned above can also be determined based on the loss function used during training.

[0230] In some embodiments, the encoding switches corresponding to all sub-encoders can be closed, and the decoding switches corresponding to all sub-decoders can also be closed. Joint optimization can be performed on the training error under multiple CSI compression parameters. The loss function used during training can also include the above formula (5), which will not be repeated here.

[0231] In this way, by jointly optimizing the training errors corresponding to multiple CSI compression parameters (such as preset CSI compression ratios), the network can optimize the training errors corresponding to all preset CSI compression ratios in a single training cycle, thereby improving the network's adaptability to dynamic changes in CSI compression ratios. After training, the parameters of the CSI decompression model can be obtained.

[0232] In some embodiments, the network device can also obtain the parameters of the second target compression model after training through the above training process, and the parameters of the second target compression model after training can be used as the parameters of the CSI compression model on the terminal device side.

[0233] In some embodiments, the network device may obtain the second compression model parameters corresponding to the second target compression model in the trained second target network model; and send the second compression model parameters to the terminal device to instruct the terminal device to determine the CSI compression model based on the second compression model parameters. The CSI compression model is used by the terminal device to obtain the target channel matrix based on the first channel matrix.

[0234] For example, a network device can send the second compression model parameters to a terminal device via signaling or data packets.

[0235] In other embodiments of this disclosure, the above training method can be executed on a terminal device, and the network device can receive the first decompression model parameters sent by the terminal device; and determine the CSI decompression model based on the first decompression model parameters.

[0236] For example, a network device can receive the first decompression model parameters sent by a terminal device via signaling or data packets, and determine the parameters corresponding to the CSI decompression model based on the first decompression model parameters.

[0237] Figure 10 This is an exemplary embodiment illustrating a method for obtaining channel quality, such as... Figure 10 As shown, the method may include:

[0238] S1001. Network devices transmit pilot signals through the downlink channel.

[0239] S1002. The terminal device receives the pilot signal through the downlink channel and obtains the first channel matrix based on the pilot signal.

[0240] The first channel matrix is ​​used to characterize the channel quality of the downlink channel;

[0241] S1003. The terminal device compresses the first channel matrix according to the Channel State Information (CSI) compression model and the CSI compression parameters to obtain the compressed target channel matrix.

[0242] In some embodiments, the CSI compression model includes a channel encoder, which includes multiple sub-encoders; different sub-encoders correspond to different CSI compression parameters.

[0243] In some embodiments, the CSI compression model may include a channel encoder and a feature converter.

[0244] S1004. The terminal device sends the target channel matrix to the network device.

[0245] S1005. The network device receives the target channel matrix sent by the terminal device, and decompresses the target channel matrix according to the Channel State Information (CSI) decompression model and CSI compression parameters to obtain the third channel matrix.

[0246] The CSI decompression model includes a channel decoder, which in turn includes multiple sub-decoders, with different sub-decoders corresponding to different CSI compression parameters.

[0247] S1006. The network device determines the channel quality of the downlink channel based on the third channel matrix.

[0248] It should be noted that the CSI compression model and the CSI decompression model described above can be any of the model structures provided in the above embodiments, and will not be elaborated further here.

[0249] In this way, by using multiple sub-encoders and multiple sub-decoders, different CSI compression parameters can be adapted. Thus, in scenarios where the CSI compression rate changes, the terminal device can adaptively obtain a more accurate target channel matrix and send it to the network device, so that the network device can obtain a more accurate channel quality, thereby improving data transmission efficiency.

[0250] Figure 11 This is a block diagram illustrating an apparatus 1100 for acquiring channel quality according to an exemplary embodiment, which can be applied to a terminal device. Figure 11 As shown, the device 1100 may include:

[0251] The first receiving module 1101 is configured to receive pilot signals transmitted by the network device through the downlink channel;

[0252] The first matrix acquisition module 1102 is configured to acquire a first channel matrix based on the pilot signal; the first channel matrix is ​​used to characterize the channel quality of the downlink channel.

[0253] The target matrix acquisition module 1103 is configured to compress the first channel matrix according to the CSI compression model and the CSI compression parameters of the channel state information to obtain the compressed target channel matrix; wherein, the CSI compression model includes a channel encoder, and the channel encoder includes multiple sub-encoders; different sub-encoders correspond to different CSI compression parameters;

[0254] The first transmitting module 1104 is configured to transmit the target channel matrix to the network device so that the network device can determine the channel quality of the downlink channel based on the target channel matrix.

[0255] In some embodiments, the target matrix acquisition module 1103 is configured to use the sub-encoder corresponding to the CSI compression parameters as the first target sub-encoder; and to compress the first channel matrix through the first target sub-encoder to obtain the target channel matrix.

[0256] In some embodiments, the CSI compression model further includes a feature converter; the target matrix acquisition module 1103 is configured to input the first channel matrix into the feature converter, extract key features from the first channel matrix to obtain a second channel matrix characterizing key CSI features; and compress the second channel matrix according to the CSI compression parameters and the channel encoder to obtain the target channel matrix.

[0257] In some embodiments, the target matrix acquisition module 1103 is configured to use the sub-encoder corresponding to the CSI compression parameters as the second target sub-encoder; and to compress the second channel matrix through the second target sub-encoder to obtain the target channel matrix.

[0258] In some embodiments, the feature converter includes a feature extraction network, an attention mechanism network, and a feature restoration network; the target matrix acquisition module 1103 is configured to input the first channel matrix into the feature extraction network to obtain a plurality of first feature maps; input the plurality of first feature maps into the attention mechanism network to obtain a second feature map; the second feature map includes key feature information from the first feature map; and input the second feature map into the feature restoration network to obtain a second channel matrix.

[0259] In some embodiments, the target matrix acquisition module 1103 is configured to perform max pooling on a plurality of first feature maps through the attention mechanism network to obtain max pooled feature maps; perform average pooling on a plurality of first feature maps through the attention mechanism network to obtain average pooled feature maps; and determine a second feature map based on the max pooled feature map and the average pooled feature map.

[0260] In some embodiments, the attention mechanism network includes a fusion sub-network; the target matrix acquisition module 1103 is configured to input the max pooling feature map and the average pooling feature map into the fusion sub-network to obtain a fused feature map; and to calculate a second feature map based on the fused feature map and the first feature map.

[0261] In some embodiments, the first matrix acquisition module 1102 is configured to measure a spatial channel matrix based on the pilot signal; transform the spatial channel matrix into an angle delay domain channel matrix through a discrete Fourier transform; and determine the first channel matrix based on the angle delay domain channel matrix.

[0262] Figure 12 This is a block diagram illustrating an apparatus 1100 for acquiring channel quality according to an exemplary embodiment, such as... Figure 12 As shown, the device 1100 may further include a first training module 1105, which is configured to train the CSI compression model obtained by:

[0263] Obtain a first sample channel matrix for training; the first sample channel matrix is ​​a matrix obtained by the terminal device based on the received pilot signal to characterize the downlink channel quality;

[0264] The first target network model is trained based on the first sample channel matrix to obtain the CSI compression model;

[0265] The first target network model includes a first target compression model and a first target decompression model. The network structure of the first target compression model is the same as that of the CSI compression model. The first target decompression model includes a channel decoder, which includes multiple sub-decoders. Different sub-decoders correspond to different CSI compression parameters.

[0266] In some embodiments, the first training module 1105 is configured to repeatedly execute the first model training step until the trained first target network model satisfies the first preset stopping iteration condition based on the first sample channel matrix and the first predicted channel matrix, and the first target compression model in the trained first target network model is used as the CSI compression model; the first predicted channel matrix is ​​the matrix output after the first sample channel matrix is ​​input into the first target network model;

[0267] The first model training steps include:

[0268] The first sample channel matrix is ​​input into the first target compression model, and the first sample channel matrix is ​​compressed by multiple sub-encoders to obtain the first target sample channel matrix.

[0269] The first target sample channel matrix is ​​input into the first target decompression model. After the first target sample channel matrix is ​​decompressed by multiple sub-decoders, the first prediction channel matrix is ​​obtained.

[0270] If, based on the first sample channel matrix and the first predicted channel matrix, it is determined that the first target network model does not meet the first preset stopping iteration condition, a first loss value is determined based on the first sample channel matrix and the first predicted channel matrix, and the parameters of the first target network model are updated based on the first loss value to obtain the trained first target network model, and the trained first target network model is used as the new first target network model.

[0271] In some embodiments, the first sending module 1104 is further configured to acquire first decompression model parameters corresponding to the first target decompression model in the trained first target network model; and send the first decompression model parameters to the network device to instruct the network device to determine the CSI decompression model based on the first decompression model parameters, wherein the CSI decompression model is used by the network device to determine the channel quality of the downlink channel based on the target channel matrix.

[0272] In some embodiments, the first receiving module 1101 is further configured to receive second compression model parameters sent by the network device; and determine a CSI compression model based on the second compression model parameters.

[0273] In some embodiments, the first receiving module 1101 is further configured to receive a first compression parameter sent by the network device; and determine the CSI compression parameter based on the first compression parameter.

[0274] Figure 13 This is a block diagram illustrating an apparatus 1300 for acquiring channel quality according to an exemplary embodiment, which can be applied to network devices. Figure 13 As shown, the device 1300 may include:

[0275] The second receiving module 1301 is configured to receive a target channel matrix sent by a terminal device; the target channel matrix is ​​obtained by the terminal device after compressing a first channel matrix according to a CSI compression model and CSI compression parameters of channel state information, and the first channel matrix is ​​a matrix obtained by the terminal device based on pilot signals to characterize the channel quality of the downlink channel.

[0276] The third matrix acquisition module 1302 is configured to decompress the target channel matrix according to the CSI decompression model and the CSI compression parameters to obtain a third channel matrix; wherein, the CSI decompression model includes a channel decoder, the channel decoder includes multiple sub-decoders, and different sub-decoders correspond to different CSI compression parameters.

[0277] The channel quality determination module 1303 is configured to determine the channel quality of the downlink channel based on the third channel matrix.

[0278] In some embodiments, the third matrix acquisition module 1302 is configured to use the sub-decoder corresponding to the CSI compression parameters as the target sub-decoder; decompress the target channel matrix through the target sub-decoder to obtain a fourth channel matrix; and determine the third channel matrix based on the fourth channel matrix.

[0279] In some embodiments, the CSI decompression model further includes a CSI reconstruction module; the third matrix acquisition module 1302 is configured to input the fourth channel matrix into the CSI reconstruction module to obtain the third channel matrix.

[0280] Figure 14 This is a block diagram illustrating an apparatus 1300 for acquiring channel quality according to an exemplary embodiment, such as... Figure 14As shown, the device 1300 may further include a second training module 1304, which is configured to train the CSI decompression model obtained in the following manner:

[0281] Obtain a second sample channel matrix for training; the second sample channel matrix is ​​a matrix obtained by the terminal device based on the received pilot signal to characterize the downlink channel quality;

[0282] The second target network model is trained based on the second sample channel matrix to obtain the CSI compression model;

[0283] The second target network model includes a second target compression model and a second target decompression model. The network structure of the second target decompression model is the same as that of the CSI decompression model. The second target compression model includes a channel encoder, which includes multiple sub-encoders. Different sub-encoders correspond to different CSI compression parameters.

[0284] In some embodiments, the second training module 1304 is configured to repeatedly execute the second model training step until the trained second target network model satisfies the second preset stopping iteration condition based on the second sample channel matrix and the second predicted channel matrix, and the second target decompression model in the trained second target network model is used as the CSI decompression model; the second predicted channel matrix is ​​the matrix output after the second sample channel matrix is ​​input into the second target network model;

[0285] The second model training steps include:

[0286] The second sample channel matrix is ​​input into the second target compression model, and after the second sample channel matrix is ​​compressed by multiple sub-encoders, the second target sample channel matrix is ​​obtained.

[0287] The second target sample channel matrix is ​​input into the second target decompression model. After the second target sample channel matrix is ​​decompressed by multiple sub-decoders, the second prediction channel matrix is ​​obtained.

[0288] If, based on the second sample channel matrix and the second predicted channel matrix, it is determined that the second target network model does not meet the second preset stopping iteration condition, a second loss value is determined based on the second sample channel matrix and the second predicted channel matrix, and the parameters of the second target network model are updated based on the second loss value to obtain the trained second target network model, and the trained second target network model is used as the new second target network model.

[0289] Figure 15This is a block diagram illustrating an apparatus 1300 for acquiring channel quality according to an exemplary embodiment, such as... Figure 15 As shown, the device 1300 may further include:

[0290] The second sending module 1305 is configured to acquire the second compression model parameters corresponding to the second target compression model in the trained second target network model; and send the second compression model parameters to the terminal device so as to instruct the terminal device to determine the CSI compression model according to the second compression model parameters, wherein the CSI compression model is used by the terminal device to obtain the target channel matrix according to the first channel matrix.

[0291] In some embodiments, the second receiving module 1301 is further configured to receive first decompression model parameters sent by the terminal device; and determine a CSI decompression model based on the first decompression model parameters.

[0292] In some embodiments, the second sending module 1305 is configured to determine a first compression parameter based on the CSI compression parameter and send the first compression parameter to the terminal device.

[0293] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0294] Figure 16 This is a block diagram illustrating an apparatus for acquiring channel quality according to an exemplary embodiment. The apparatus 2000 for acquiring channel quality may be... Figure 1 The terminal equipment in the communication system shown can also be the network equipment in the same communication system.

[0295] Reference Figure 16 The device 2000 may include one or more of the following components: a processing component 2002, a memory 2004, and a communication component 2006.

[0296] Processing component 2002 can control the overall operation of device 2000, such as operations associated with display, telephone calls, data communication, camera operation, and recording operation. Processing component 2002 may include one or more processors 2020 to execute instructions to complete all or part of the steps of the method for acquiring channel quality described above. Furthermore, processing component 2002 may include one or more modules to facilitate interaction between processing component 2002 and other components. For example, processing component 2002 may include a multimedia module to facilitate interaction between multimedia components and processing component 2002.

[0297] Memory 2004 is configured to store various types of data to support the operation of device 2000. Examples of this data include instructions for any application or method operating on device 2000, contact data, phonebook data, messages, pictures, videos, etc. Memory 2004 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0298] Communication component 2006 is configured to facilitate wired or wireless communication between device 2000 and other devices. Device 2000 can access wireless networks based on communication standards, such as WiFi, 2G, 3G, 4G, 5G, or 6G communication technologies, or combinations thereof. In one exemplary embodiment, communication component 2006 receives broadcast signals or broadcast-related information from an external broadcast management system via a broadcast channel. In one exemplary embodiment, communication component 2006 also includes a near-field communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on radio frequency identification (RFID) technology, Infrared Data Association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.

[0299] In an exemplary embodiment, the apparatus 2000 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above-described method for acquiring channel quality.

[0300] The aforementioned device 2000 can be a standalone electronic device or part of a standalone electronic device. For example, in one embodiment, the electronic device can be an integrated circuit (IC) or a chip, wherein the integrated circuit can be a single IC or a collection of multiple ICs. The chip can include, but is not limited to, the following types: GPU (Graphics Processing Unit), CPU (Central Processing Unit), FPGA (Field Programmable Gate Array), DSP (Digital Signal Processor), ASIC (Application Specific Integrated Circuit), SOC (System on Chip), etc. The aforementioned integrated circuit or chip can be used to execute executable instructions (or code) to implement the aforementioned method for obtaining channel quality. The executable instructions can be stored in the integrated circuit or chip or obtained from other devices or equipment. For example, the integrated circuit or chip may include a processor, memory, and an interface for communicating with other devices. The executable instruction can be stored in the processor, and when the executable instruction is executed by the processor, it implements the above-described method for obtaining channel quality; or, the integrated circuit or chip can receive the executable instruction through the interface and transmit it to the processor for execution to implement the above-described method for obtaining channel quality.

[0301] In an exemplary embodiment, this disclosure also provides a computer-readable storage medium storing computer program instructions thereon, which, when executed by a processor, implement the steps of the method for acquiring channel quality provided in this disclosure. For example, the computer-readable storage medium may be a non-transitory computer-readable storage medium including instructions, such as the aforementioned memory 2004 including instructions, which can be executed by the processor 2020 of the device 2000 to complete the aforementioned method for acquiring channel quality. For example, the non-transitory computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0302] In another exemplary embodiment, a computer program product is also provided, the computer program product comprising a computer program executable by a programmable device, the computer program having a code portion for performing the above-described method of obtaining channel quality when executed by the programmable device.

[0303] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of this disclosure. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the following claims.

[0304] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims.

Claims

1. A method for obtaining channel quality, characterized in that, Applied to a terminal device, the method includes: Receive pilot signals transmitted by network devices through the downlink channel; A first channel matrix is ​​obtained based on the pilot signal; the first channel matrix is ​​used to characterize the channel quality of the downlink channel; The first channel matrix is ​​compressed according to the Channel State Information (CSI) compression model and the CSI compression parameters to obtain the compressed target channel matrix; wherein, the CSI compression model includes a channel encoder, and the channel encoder includes multiple sub-encoders; different sub-encoders correspond to different CSI compression parameters; The target channel matrix is ​​sent to the network device so that the network device can determine the channel quality of the downlink channel based on the target channel matrix; The step of obtaining the first channel matrix based on the pilot signal includes: Based on the received pilot signals, Channel State Information (CSI) is estimated to obtain the CSI estimation matrix; The first channel matrix is ​​determined based on the CSI estimation matrix; The step of compressing the first channel matrix according to the Channel State Information (CSI) compression model and CSI compression parameters to obtain the compressed target channel matrix includes: The sub-encoder corresponding to the CSI compression parameters is taken as the first target sub-encoder; The first target channel matrix is ​​obtained by compressing the first channel matrix using the first target sub-encoder.

2. The method according to claim 1, characterized in that, The CSI compression model also includes a feature converter; The step of compressing the first channel matrix according to the Channel State Information (CSI) compression model and CSI compression parameters to obtain the compressed target channel matrix includes: The first channel matrix is ​​input into the feature converter, and key features are extracted from the first channel matrix to obtain a second channel matrix representing the key features of CSI. The second channel matrix is ​​compressed according to the CSI compression parameters and the channel encoder to obtain the target channel matrix.

3. The method according to claim 2, characterized in that, The step of compressing the second channel matrix according to the CSI compression parameters and the channel encoder to obtain the target channel matrix includes: The sub-encoder corresponding to the CSI compression parameters is used as the second target sub-encoder; The second channel matrix is ​​compressed by the second target sub-encoder to obtain the target channel matrix.

4. The method according to claim 2, characterized in that, The feature converter includes a feature extraction network, an attention mechanism network, and a feature restoration network; the step of inputting the first channel matrix into the feature converter and extracting key features from the first channel matrix to obtain a second channel matrix representing the key features of CSI includes: The first channel matrix is ​​input into the feature extraction network to obtain multiple first feature maps; Multiple first feature maps are input into the attention mechanism network to obtain a second feature map; the second feature map includes key feature information from the first feature maps; The second feature map is input into the feature reconstruction network to obtain the second channel matrix.

5. The method according to claim 4, characterized in that, The step of inputting multiple first feature maps into the attention mechanism network to obtain second feature maps includes: Max pooling is performed on multiple first feature maps through the attention mechanism network to obtain max pooling feature maps; The attention mechanism network is used to perform average pooling on multiple first feature maps to obtain average pooled feature maps. A second feature map is determined based on the max pooling feature map and the average pooling feature map.

6. The method according to claim 5, characterized in that, The attention mechanism network includes a fusion sub-network; determining the second feature map based on the max pooling feature map and the average pooling feature map includes: The max pooling feature map and the average pooling feature map are input into the fusion sub-network to obtain the fused feature map; A second feature map is determined based on the fused feature map and the first feature map.

7. The method according to claim 1, characterized in that, The step of obtaining the first channel matrix based on the pilot signal includes: The spatial channel matrix is ​​measured based on the pilot signal. The spatial domain channel matrix is ​​transformed into an angular time delay domain channel matrix by using the discrete Fourier transform. The first channel matrix is ​​determined based on the angle delay domain channel matrix.

8. The method according to any one of claims 1 to 7, characterized in that, The CSI compression model was trained in the following way: Obtain a first sample channel matrix for training; the first sample channel matrix is ​​a matrix obtained by the terminal device based on the received pilot signal to characterize the downlink channel quality; The first target network model is trained based on the first sample channel matrix to obtain the CSI compression model; The first target network model includes a first target compression model and a first target decompression model. The network structure of the first target compression model is the same as that of the CSI compression model. The first target decompression model includes a channel decoder, which includes multiple sub-decoders. Different sub-decoders correspond to different CSI compression parameters.

9. The method according to claim 8, characterized in that, The step of training the first target network model based on the first sample channel matrix includes: The first model training step is executed repeatedly until the first target network model after training satisfies the first preset stopping iteration condition based on the first sample channel matrix and the first predicted channel matrix. The first target compression model in the trained first target network model is then used as the CSI compression model. The first predicted channel matrix is ​​the matrix output after the first sample channel matrix is ​​input into the first target network model. The first model training steps include: The first sample channel matrix is ​​input into the first target compression model, and the first sample channel matrix is ​​compressed by multiple sub-encoders to obtain the first target sample channel matrix. The first target sample channel matrix is ​​input into the first target decompression model. After the first target sample channel matrix is ​​decompressed by multiple sub-decoders, the first prediction channel matrix is ​​obtained. If, based on the first sample channel matrix and the first predicted channel matrix, it is determined that the first target network model does not meet the first preset stopping iteration condition, a first loss value is determined based on the first sample channel matrix and the first predicted channel matrix, and the parameters of the first target network model are updated based on the first loss value to obtain the trained first target network model, and the trained first target network model is used as the new first target network model.

10. The method according to claim 8, characterized in that, The method further includes: Obtain the parameters of the first decompression model corresponding to the first target decompression model in the trained first target network model; The first decompression model parameters are sent to the network device to instruct the network device to determine the Channel State Information (CSI) decompression model based on the first decompression model parameters. The CSI decompression model is used by the network device to determine the channel quality of the downlink channel based on the target channel matrix.

11. The method according to any one of claims 1 to 7, characterized in that, The method further includes: Receive the second compression model parameters sent by the network device; The CSI compression model is determined based on the parameters of the second compression model.

12. The method according to any one of claims 1 to 7, characterized in that, The method further includes: Receive the first compression parameter sent by the network device; The CSI compression parameters are determined based on the first compression parameters.

13. A method for obtaining channel quality, characterized in that, Applied to network devices, the method includes: The terminal device receives a target channel matrix sent by itself. The target channel matrix is ​​obtained by compressing a first channel matrix using a Channel State Information (CSI) compression model and CSI compression parameters. The first channel matrix is ​​a matrix obtained by the terminal device based on pilot signals, characterizing the channel quality of the downlink channel. The first channel matrix is ​​determined by the terminal device through CSI estimation based on received pilot signals. The target channel matrix is ​​obtained by the terminal device using the sub-encoder corresponding to the CSI compression parameters as a first target sub-encoder, and compressing the first channel matrix using the first target sub-encoder. The target channel matrix is ​​decompressed according to the Channel State Information (CSI) decompression model and the CSI compression parameters to obtain a third channel matrix; wherein, the CSI decompression model includes a channel decoder, and the channel decoder includes multiple sub-decoders, with different sub-decoders corresponding to different CSI compression parameters; The channel quality of the downlink channel is determined based on the third channel matrix.

14. The method according to claim 13, characterized in that, The step of decompressing the target channel matrix according to the Channel State Information (CSI) decompression model and the CSI compression parameters to obtain the third channel matrix includes: Use the sub-decoder corresponding to the CSI compression parameters as the target sub-decoder; The target channel matrix is ​​decompressed by the target sub-decoder to obtain the fourth channel matrix; The third channel matrix is ​​determined based on the fourth channel matrix.

15. The method according to claim 14, characterized in that, The CSI decompression model further includes a CSI reconstruction module; determining the third channel matrix based on the fourth channel matrix includes: The fourth channel matrix is ​​input into the CSI reconstruction module to obtain the third channel matrix.

16. The method according to any one of claims 13 to 15, characterized in that, The CSI decompression model was trained in the following way: Obtain a second sample channel matrix for training; the second sample channel matrix is ​​a matrix obtained by the terminal device based on the received pilot signal to characterize the downlink channel quality; The second target network model is trained based on the second sample channel matrix to obtain the CSI decompression model; The second target network model includes a second target compression model and a second target decompression model. The network structure of the second target decompression model is the same as that of the CSI decompression model. The second target compression model includes a channel encoder, which includes multiple sub-encoders. Different sub-encoders correspond to different CSI compression parameters.

17. The method according to claim 16, characterized in that, The step of training the second target network model based on the second sample channel matrix includes: The second model training step is executed repeatedly until the trained second target network model satisfies the second preset stopping iteration condition based on the second sample channel matrix and the second predicted channel matrix. The second target decompression model in the trained second target network model is then used as the CSI decompression model. The second predicted channel matrix is ​​the matrix output after the second sample channel matrix is ​​input into the second target network model. The second model training steps include: The second sample channel matrix is ​​input into the second target compression model, and after the second sample channel matrix is ​​compressed by multiple sub-encoders, the second target sample channel matrix is ​​obtained. The second target sample channel matrix is ​​input into the second target decompression model. After the second target sample channel matrix is ​​decompressed by multiple sub-decoders, the second prediction channel matrix is ​​obtained. If, based on the second sample channel matrix and the second predicted channel matrix, it is determined that the second target network model does not meet the second preset stopping iteration condition, a second loss value is determined based on the second sample channel matrix and the second predicted channel matrix, and the parameters of the second target network model are updated based on the second loss value to obtain the trained second target network model, and the trained second target network model is used as the new second target network model.

18. The method according to claim 16, characterized in that, The method further includes: Obtain the parameters of the second compression model corresponding to the second objective compression model in the trained second objective network model; The second compression model parameters are sent to the terminal device to instruct the terminal device to determine the CSI compression model based on the second compression model parameters. The CSI compression model is used by the terminal device to obtain the target channel matrix based on the first channel matrix.

19. The method according to any one of claims 13 to 15, characterized in that, The method further includes: Receive the first decompression model parameters sent by the terminal device; The CSI decompression model is determined based on the parameters of the first decompression model.

20. The method according to any one of claims 13 to 15, characterized in that, The method further includes: The first compression parameter is determined based on the CSI compression parameter; The first compression parameter is sent to the terminal device.

21. An apparatus for acquiring channel quality, characterized in that, Applied to a terminal device, the device includes: The first receiving module is configured to receive pilot signals transmitted by network devices through the downlink channel; The first matrix acquisition module is configured to acquire a first channel matrix based on the pilot signal; the first channel matrix is ​​used to characterize the channel quality of the downlink channel. The target matrix acquisition module is configured to compress the first channel matrix according to the Channel State Information (CSI) compression model and the CSI compression parameters to obtain the compressed target channel matrix; wherein, the CSI compression model includes a channel encoder, and the channel encoder includes multiple sub-encoders; different sub-encoders correspond to different CSI compression parameters; A first transmitting module is configured to transmit the target channel matrix to the network device, so that the network device can determine the channel quality of the downlink channel based on the target channel matrix; The first matrix acquisition module is configured to perform channel state information (CSI) estimation based on the received pilot signal to obtain a CSI estimation matrix; and to determine the first channel matrix based on the CSI estimation matrix. The target matrix acquisition module is configured to use the sub-encoder corresponding to the CSI compression parameters as the first target sub-encoder; and to compress the first channel matrix through the first target sub-encoder to obtain the target channel matrix.

22. An apparatus for acquiring channel quality, characterized in that, Applied to network devices, the device includes: The second receiving module is configured to receive a target channel matrix transmitted by a terminal device. The target channel matrix is ​​obtained by the terminal device after compressing a first channel matrix according to a CSI compression model and Channel State Information (CSI) compression parameters. The first channel matrix is ​​a matrix obtained by the terminal device based on pilot signals, used to characterize the channel quality of the downlink channel. The first channel matrix is ​​determined by the terminal device through CSI estimation based on received pilot signals. The target channel matrix is ​​obtained by the terminal device using the sub-encoder corresponding to the CSI compression parameters as a first target sub-encoder to compress the first channel matrix. The third matrix acquisition module is configured to decompress the target channel matrix according to the Channel State Information (CSI) decompression model and the CSI compression parameters to obtain a third channel matrix; wherein, the CSI decompression model includes a channel decoder, and the channel decoder includes multiple sub-decoders, with different sub-decoders corresponding to different CSI compression parameters; The channel quality determination module is configured to determine the channel quality of the downlink channel based on the third channel matrix.

23. An apparatus for acquiring channel quality, characterized in that, The device includes: processor; Memory used to store processor-executable instructions; The processor is configured to perform the steps of the method according to any one of claims 1 to 12, or the processor is configured to perform the steps of the method according to any one of claims 13 to 20.

24. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 12, or when the computer program instructions are executed by a processor, they implement the steps of the method according to any one of claims 13 to 20.

25. A chip, characterized in that, It includes a processor and an interface; the processor is configured to read instructions to perform the steps of the method according to any one of claims 1 to 12, or the processor is configured to read instructions to perform the steps of the method according to any one of claims 13 to 20.

Images