Channel state transmission method, channel state acquisition method, channel state transmission device, channel state acquisition device, terminal and network equipment

Abstract

The invention provides a channel state transmission and acquisition method and device, a terminal and network equipment, and the method comprises the steps: obtaining a plurality of pieces of first downlink channel state information (CSI), and enabling each piece of first downlink CSI to correspond to one piece of network equipment; compressing the plurality of pieces of first downlink CSI to obtain target information; and sending the target information to at least two network devices, wherein the at least two network devices provide coherent joint transmission for the terminal. According to the scheme, the target information is acquired by compressing the plurality of first downlink CSI, and the target information is sent to the at least two network devices, so that the acquisition time delay of the downlink CSI from each cooperative network device to the UE is reduced, and the signaling overhead is saved.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a channel state transmission and acquisition method, apparatus, terminal and network device. Background Technology

[0002] In distributed multiple-in multiple-out (MIMO) systems (e.g., distributed MIMO, cell-free massive MIMO, or cooperative communication), multiple base stations, transmit and receive points (TRPs), or access points (APs) serve a single user equipment (UE, also known as a terminal). To perform coherent joint transmission (CJT), the cooperating base stations need to obtain downlink channel state information (CSI) from each cooperating base station to the UE. Current methods for obtaining downlink CSI from each cooperating base station to the UE suffer from significant latency or high signaling overhead. Summary of the Invention

[0003] This application provides a channel state transmission and acquisition method, apparatus, terminal, and network device to solve the problems of time delay and large signaling overhead in the acquisition method of downlink CSI from each cooperating base station to the UE.

[0004] To address the aforementioned technical problems, embodiments of this application provide a channel state information transmission method, applied to a terminal, comprising:

[0005] Acquire multiple First Downlink Channel State Information (CSI) entries, with each First Downlink CSI corresponding to a network device.

[0006] The multiple first downlink CSIs are compressed to obtain target information;

[0007] The target information is sent to at least two network devices, which provide coherent joint transmission for the terminal.

[0008] Optionally, compressing the plurality of first downlink CSIs to obtain target information includes:

[0009] The plurality of first downlink CSIs are input into the first target model to obtain the target information output by the first target model;

[0010] The first target model is a two-layer compression model. The first layer compression model compresses multiple first downlink CSIs respectively, and the second layer compression model compresses multiple second downlink CSIs output by the first layer compression model. The first target model is obtained by training on first training data, which includes at least one of the following: third downlink CSIs corresponding to multiple network devices, the number of transmit antenna ports of multiple network devices, the number of subbands occupied by the transmit signals of multiple network devices, and tag data information.

[0011] The third downlink CSI is the downlink CSI used for model training.

[0012] Optionally, the first target model is a model trained using a two-sided model, or the first target model is a model trained using a one-sided model.

[0013] Optionally, the bilateral model includes at least one of the following:

[0014] Autoencoder model, convolutional neural network model, generative adversarial network model, Transformer structure model.

[0015] Optionally, when the first target model is a model trained using a two-sided model, the method further includes at least one of the following:

[0016] The second target model is sent to multiple network devices. The second target model is a model used to decompress the target information. The second target model is obtained by the terminal through training using the first training data.

[0017] The first target model sent by the network device;

[0018] Obtain the pre-configured first target model.

[0019] Optionally, when the first target model is a model trained by a bilateral model, the label data information included in the first training data is the third downlink CSI corresponding to multiple network devices respectively.

[0020] Optionally, when the first target model is a model trained by a one-sided model, the label data information included in the first training data includes first label data and second label data. The first label data is information obtained by compressing the third downlink CSI of multiple network devices using the compressed sensing principle, and the second label information is information obtained by compressing the first label data information using the compressed sensing principle.

[0021] Optionally, sending the target information to at least two network devices includes one of the following:

[0022] Send target information to at least two network devices using a directional beam;

[0023] The target information is transmitted to at least two network devices via an omnidirectional antenna.

[0024] Optionally, the method further includes:

[0025] The terminal receives indication information sent by network devices, the indication information being used to indicate whether the terminal performs joint compression and joint feedback on the first downlink CSI from multiple network devices in cooperative communication mode;

[0026] The step of compressing the plurality of first downlink CSIs to obtain target information includes:

[0027] When the instruction information instructs the terminal to perform joint compression and joint feedback of the first downlink CSI from multiple network devices in the cooperative communication mode, the multiple first downlink CSIs are compressed to obtain the target information.

[0028] Optionally, the method further includes:

[0029] Send the identification information of the network device associated with the target information to at least two network devices.

[0030] This application also provides a channel state information acquisition method, applied to a network device, including:

[0031] The target information sent by the receiving terminal is obtained by compression of multiple first downlink channel state information (CSI), and each first downlink CSI corresponds to a network device.

[0032] The target information is decompressed to obtain multiple decompressed first downlink CSIs.

[0033] Optionally, the step of decompressing the target information to obtain multiple decompressed first downlink CSIs includes:

[0034] The target information is input into the second target model to obtain multiple decompressed first downlink CSIs output by the second target model;

[0035] The second target model is a two-layer decompression model. The first-layer decompression model decompresses the target information to obtain the fourth downlink CSI, and the second-layer multiple decompression models decompress the fourth downlink CSI respectively.

[0036] The second target model is obtained by training a bilateral model and the first training data; or, the second target model is obtained by training a unilateral model and the second training data, wherein the second training data includes: training data of the first layer decompression model, label data of the first layer decompression model, training data of the second layer decompression model, and label data of the second layer decompression model, wherein the training data of the first layer decompression model is the second label data information, the label data of the first layer decompression model is the first label data information, the training data of the second layer decompression model is the first label data information, and the label data of the second layer decompression model is the third downlink CSI from multiple network devices;

[0037] The first training data includes at least one of the following: the third downlink CSI corresponding to multiple network devices, the number of transmit antenna ports of multiple network devices, the number of subbands occupied by the transmit signals of multiple network devices, and tag data information;

[0038] The first tag data is information obtained by compressing the third downlink CSI of multiple network devices using the principle of compressed sensing, and the second tag information is information obtained by compressing the first tag data using the principle of compressed sensing.

[0039] Optionally, the bilateral model includes at least one of the following:

[0040] Autoencoder model, convolutional neural network model, generative adversarial network model, Transformer structure model.

[0041] Optionally, when the second target model is obtained by training using a bilateral model and the first training data, the method further includes at least one of the following:

[0042] The first target model is sent to the terminal. The first target model is a model used for multiple first downlink CSI compressions. The first target model is obtained by the network device through training using the first training data.

[0043] The second target model sent by the receiving terminal.

[0044] Optionally, the method further includes:

[0045] The terminal is sent an indication message, which is used to indicate whether the terminal performs joint compression and joint feedback on the first downlink CSI from multiple network devices in the cooperative communication mode.

[0046] Optionally, the method further includes one of the following:

[0047] The receiving terminal sends the identification information of the network device associated with the target information;

[0048] The terminal receives identification information of the network device that provides cooperative communication for the terminal, sent by the control center.

[0049] This application also provides a terminal, including a memory, a transceiver, and a processor:

[0050] A memory for storing computer programs; a transceiver for sending and receiving data under the control of the processor; and a processor for reading the computer programs from the memory and performing the following operations:

[0051] Acquire multiple First Downlink Channel State Information (CSI) entries, with each First Downlink CSI corresponding to a network device.

[0052] The multiple first downlink CSIs are compressed to obtain target information;

[0053] The target information is transmitted to at least two network devices via a transceiver, and the at least two network devices provide coherent joint transmission for the terminal.

[0054] Optionally, the processor is configured to read the computer program in the memory and perform the following operations:

[0055] The plurality of first downlink CSIs are input into the first target model to obtain the target information output by the first target model;

[0056] The first target model is a two-layer compression model. The first layer compression model compresses multiple first downlink CSIs respectively, and the second layer compression model compresses multiple second downlink CSIs output by the first layer compression model. The first target model is obtained by training on first training data, which includes at least one of the following: third downlink CSIs corresponding to multiple network devices, the number of transmit antenna ports of multiple network devices, the number of subbands occupied by the transmit signals of multiple network devices, and tag data information.

[0057] The third downlink CSI is the downlink CSI used for model training.

[0058] Optionally, the first target model is a model trained using a two-sided model, or the first target model is a model trained using a one-sided model.

[0059] Optionally, the bilateral model includes at least one of the following:

[0060] Autoencoder model, convolutional neural network model, generative adversarial network model, Transformer structure model.

[0061] Optionally, if the first target model is a model trained using a bilateral model, the processor, when reading the computer program from the memory, further performs at least one of the following operations:

[0062] The second target model is sent to multiple network devices. The second target model is a model used to decompress the target information. The second target model is obtained by the terminal through training using the first training data.

[0063] The first target model sent by the network device;

[0064] Obtain the pre-configured first target model.

[0065] Optionally, when the first target model is a model trained by a bilateral model, the label data information included in the first training data is the third downlink CSI corresponding to multiple network devices respectively.

[0066] Optionally, when the first target model is a model trained by a one-sided model, the label data information included in the first training data includes first label data and second label data. The first label data is information obtained by compressing the third downlink CSI of multiple network devices using the compressed sensing principle, and the second label information is information obtained by compressing the first label data information using the compressed sensing principle.

[0067] Optionally, the processor is configured to read a computer program from the memory and perform one of the following operations:

[0068] Send target information to at least two network devices using a directional beam;

[0069] The target information is transmitted to at least two network devices via an omnidirectional antenna.

[0070] Optionally, the processor, for reading the computer program in the memory, further performs the following operations:

[0071] The terminal receives indication information sent by network devices through a transceiver. The indication information is used to indicate whether the terminal should perform joint compression and joint feedback on the first downlink CSI from multiple network devices in cooperative communication mode.

[0072] When the instruction information instructs the terminal to perform joint compression and joint feedback of the first downlink CSI from multiple network devices in the cooperative communication mode, the multiple first downlink CSIs are compressed to obtain the target information.

[0073] Optionally, the processor, for reading the computer program in the memory, further performs the following operations:

[0074] The target information is transmitted via a transceiver to at least two network devices, along with the identification information of the network device associated with it.

[0075] This application also provides a network device, including a memory, a transceiver, and a processor:

[0076] A memory for storing computer programs; a transceiver for sending and receiving data under the control of the processor; and a processor for reading the computer programs from the memory and performing the following operations:

[0077] The target information is received by the transceiver terminal. The target information is obtained by compression of multiple first downlink channel state information (CSI) and each first downlink CSI corresponds to a network device.

[0078] The target information is decompressed to obtain multiple decompressed first downlink CSIs.

[0079] Optionally, the processor is configured to read the computer program in the memory and perform the following operations:

[0080] The target information is input into the second target model to obtain multiple decompressed first downlink CSIs output by the second target model;

[0081] The second target model is a two-layer decompression model. The first-layer decompression model decompresses the target information to obtain the fourth downlink CSI, and the second-layer multiple decompression models decompress the fourth downlink CSI respectively.

[0082] The second target model is obtained by training a bilateral model and the first training data; or, the second target model is obtained by training a unilateral model and the second training data, wherein the second training data includes: training data of the first layer decompression model, label data of the first layer decompression model, training data of the second layer decompression model, and label data of the second layer decompression model, wherein the training data of the first layer decompression model is the second label data information, the label data of the first layer decompression model is the first label data information, the training data of the second layer decompression model is the first label data information, and the label data of the second layer decompression model is the third downlink CSI from multiple network devices;

[0083] The first training data includes at least one of the following: the third downlink CSI corresponding to multiple network devices, the number of transmit antenna ports of multiple network devices, the number of subbands occupied by the transmit signals of multiple network devices, and tag data information;

[0084] The first tag data is information obtained by compressing the third downlink CSI of multiple network devices using the principle of compressed sensing, and the second tag information is information obtained by compressing the first tag data using the principle of compressed sensing.

[0085] Optionally, the bilateral model includes at least one of the following:

[0086] Autoencoder model, convolutional neural network model, generative adversarial network model, Transformer structure model.

[0087] Optionally, when the second target model is acquired through training using a bilateral model and the first training data, the processor, for reading the computer program in the memory, further performs at least one of the following operations:

[0088] The first target model is sent to the terminal. The first target model is a model used for multiple first downlink CSI compressions. The first target model is obtained by the network device through training using the first training data.

[0089] The second target model sent by the receiving terminal.

[0090] Optionally, the processor, for reading the computer program in the memory, further performs the following operations:

[0091] The transceiver sends an indication message to the terminal, which is used to indicate whether the terminal should perform joint compression and joint feedback on the first downlink CSI from multiple network devices in cooperative communication mode.

[0092] Optionally, the processor, for reading the computer program in the memory, further performs one of the following operations:

[0093] The transceiver receives the identification information of the network device associated with the target information sent by the terminal;

[0094] The transceiver receives identification information of the network devices that provide cooperative communication for the terminal, sent by the control center.

[0095] This application also provides a channel state information transmission device, applied to a terminal, including:

[0096] The first acquisition unit is used to acquire multiple first downlink channel state information (CSI), each first downlink CSI corresponding to a network device.

[0097] The second acquisition unit is used to compress the plurality of first downlink CSIs to acquire target information;

[0098] The first sending unit is used to send the target information to at least two network devices, wherein the at least two network devices provide coherent joint transmission for the terminal.

[0099] This application embodiment also provides a channel state information acquisition device, applied to a network device, including:

[0100] The first receiving unit is used to receive target information sent by the terminal. The target information is obtained by compression of multiple first downlink channel state information (CSI) and each first downlink CSI corresponds to a network device.

[0101] The third acquisition unit is used to decompress the target information and acquire multiple decompressed first downlink CSIs.

[0102] This application also provides a processor-readable storage medium storing a computer program for causing the processor to perform the above-described method.

[0103] This application also provides a computer program product, including computer instructions, which, when executed by a processor, implement the steps of the above-described method.

[0104] The beneficial effects of this application are:

[0105] The above scheme reduces the downlink CSI acquisition latency from each cooperating network device to the UE and saves signaling overhead by compressing multiple first downlink CSIs and sending the target information to at least two network devices. Attached Figure Description

[0106] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0107] Figure 1 This diagram illustrates the structure of a network system applicable to embodiments of this application.

[0108] Figure 2 A flowchart illustrating a channel state information transmission method according to an embodiment of this application;

[0109] Figure 3 A schematic diagram illustrating the structure of a two-layer compressed AI model deployed on a terminal;

[0110] Figure 4A schematic diagram illustrating the structure of a two-layer decompression AI model deployed across various collaborative network devices;

[0111] Figure 5 One of the schematic diagrams illustrating CSI joint compression and feedback;

[0112] Figure 6 This is the second schematic diagram illustrating CSI joint compression and feedback.

[0113] Figure 7 A flowchart illustrating the channel state information acquisition method according to an embodiment of this application;

[0114] Figure 8 A schematic diagram of a channel state information transmission apparatus according to an embodiment of this application;

[0115] Figure 9 A structural diagram of the terminal according to an embodiment of this application;

[0116] Figure 10 A schematic diagram of a channel state information acquisition device according to an embodiment of this application;

[0117] Figure 11 This is a structural diagram of a network device according to an embodiment of this application. Detailed Implementation

[0118] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0119] The terms “first,” “second,” etc., used in the specification and claims of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the application described herein may be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0120] In this application's embodiments, the term "and / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship. In this application's embodiments, the term "multiple" refers to two or more, and other quantifiers are similar.

[0121] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0122] The relevant concepts mentioned in this application will be briefly explained below.

[0123] The current method for obtaining downlink CSI from each cooperating base station has the following drawbacks:

[0124] Scenario 1: When the terminal uses a directional antenna or a narrow transmit beam to feed back the estimated downlink CSI to the corresponding base station, and then each cooperating base station obtains the downlink CSI from other cooperating base stations to the terminal through the backhaul link, each downlink CSI is sent to each cooperating base station via spatial division. The terminal can use the same time and frequency resources to discover the downlink CSI. In this case, the time and frequency resources occupied by feeding back the downlink CSI are minimized. However, since each cooperating base station needs to obtain the downlink CSI between other cooperating base stations and the terminal through the backhaul link, it introduces a large transmission delay.

[0125] Scenario 2: The terminal uses an omnidirectional antenna to feed back the estimated downlink CSI to the corresponding base station in time-division or frequency-division mode. Then, each cooperating base station obtains the downlink CSI from other cooperating base stations to the terminal through the backhaul link. When the terminal uses an omnidirectional antenna, each cooperating base station needs to send downlink CSI to the base station separately in time-division or frequency-division mode, and the time and frequency resources used are N times that of Scenario 1. However, in Scenario 2, if each base station coordinates the time and frequency resources for receiving downlink CSI, each cooperating base station can also receive downlink CSI from other cooperating base stations. At this time, Scenario 2 evolves into Scenario 4 as follows: The time and frequency resources used for downlink CSI feedback in Scenario 4 are N times that of Scenario 1, but no additional delay is introduced.

[0126] Case 3: When the terminal uses a directional antenna or a narrower transmit beam to send downlink CSI from all cooperating base stations to all cooperating base stations, the time and frequency resources it occupies are N times that of Case 1 because each transmit beam contains downlink CSI from all cooperating base stations, but it does not introduce additional latency.

[0127] Case 4: The downlink CSI feedback method in which the terminal uses an omnidirectional antenna to send downlink CSI from all cooperating base stations to all cooperating base stations uses N times the time and frequency resources of Case 1, but does not introduce additional latency.

[0128] The embodiments of this application are described below with reference to the accompanying drawings. The channel state transmission and acquisition method, apparatus, terminal, and network equipment provided in the embodiments of this application can be applied to wireless communication systems. This wireless communication system can be a system employing fifth-generation (5G) mobile communication technology (hereinafter referred to as a 5G system). Those skilled in the art will understand that the 5G NR system is merely an example and not a limitation.

[0129] See Figure 1 , Figure 1 This is a structural diagram of a network system that can be applied to the embodiments of this application, such as... Figure 1 As shown, the system includes a user terminal 11 and a base station 12. The user terminal 11 can be a user equipment (UE), such as a mobile phone, tablet personal computer, laptop computer, personal digital assistant (PDA), mobile internet device (MID), or wearable device. It should be noted that the specific type of user terminal 11 is not limited in this embodiment. The base station 12 can be a 5G or later version base station (e.g., gNB, 5G NR NB), or a base station in other communication systems, also referred to as a node B. It should be noted that this embodiment only uses a 5G base station as an example, but the specific type of base station 12 is not limited.

[0130] This application provides a channel state transmission and acquisition method, apparatus, terminal, and network device to solve the problems of time delay and large signaling overhead in the acquisition method of downlink CSI from each cooperating base station to the UE.

[0131] The method and apparatus are based on the same concept of the application. Since the methods and apparatus solve problems in similar ways, the implementation of the apparatus and methods can refer to each other, and the repeated parts will not be described again.

[0132] like Figure 2 As shown, this application embodiment provides a channel state information transmission method, executed by a terminal, including:

[0133] Step S201: Obtain multiple first downlink channel state information (CSIs), each first downlink CSI corresponding to a network device;

[0134] Step 202: Compress the plurality of first downlink CSIs to obtain target information;

[0135] Step 203: Send the target information to at least two network devices, which provide coherent joint transmission for the terminal.

[0136] It should be noted that, in this embodiment of the application, target information is obtained by compressing multiple first downlink CSIs and sending the target information to at least two network devices, thereby reducing the acquisition latency of downlink CSIs from each cooperating network device to the UE and saving signaling overhead.

[0137] It should be noted that the embodiments of this application are applied to distributed MIMO, cell-free massive MIMO, or cooperative communication systems, in scenarios where multiple network devices serve a single terminal. Optionally, the network devices in the embodiments of this application can also be referred to as cooperative network devices. Optionally, the network device can be a base station, a TRP, or an AP.

[0138] Optionally, each cooperating network device sends a network device-specific downlink channel state information reference signal (CSI-RS), and the terminal receives the downlink CSI-RS sent by each cooperating network device. Based on the received downlink CSI-RS sent by each cooperating network device, the terminal estimates the downlink CSI from each cooperating network device to the terminal.

[0139] Optionally, in one implementation, compressing the plurality of first downlink CSIs to obtain target information includes:

[0140] The plurality of first downlink CSIs are input into the first target model to obtain the target information output by the first target model;

[0141] The first target model is a two-layer compression model. The first layer compression model compresses multiple first downlink CSIs respectively, and the second layer compression model compresses multiple second downlink CSIs output by the first layer compression model. The first target model is obtained by training on first training data, which includes at least one of the following: third downlink CSIs corresponding to multiple network devices, the number of transmit antenna ports of multiple network devices, the number of subbands occupied by the transmit signals of multiple network devices, and tag data information.

[0142] The third downlink CSI is the downlink CSI used for model training.

[0143] It should be noted that the third downlink CSI mentioned here refers to the downlink CSI used for model training, while the first downlink CSI refers to the downlink CSI obtained by the terminal in actual application. It is worth noting that both the first and third downlink CSI are obtained by measuring the CSI-RS sent by the terminal measurement network device. The difference between the first and third downlink CSI is that these two downlink CSIs are obtained at different times by different CSI-RS.

[0144] Optionally, in one implementation, the first target model is a model trained using a two-sided model, or the first target model is a model trained using a one-sided model.

[0145] It should be noted that a two-sided model can be understood as a joint training of the compression and decompression models, meaning that the compression and decompression models can be determined by training on one side, while a one-sided model means that the compression and decompression models need to be trained independently.

[0146] Optionally, the bilateral model includes at least one of the following:

[0147] Autoencoder model, convolutional neural network model, generative adversarial network (GAN) model, and Transformer structure model.

[0148] Optionally, in one implementation, when the first target model is a model trained using a two-sided model, the method further includes at least one of the following:

[0149] A11. The second target model is sent to multiple network devices. The second target model is a model used to decompress the target information. The second target model is obtained by the terminal through training using the first training data.

[0150] Alternatively, this can be understood as the model training being performed on the terminal side, and after the terminal side completes the training, it sends the model for decompression to the network device.

[0151] A12, Receive the first target model sent by the network device;

[0152] Alternatively, this can be understood as the model training being performed on the network device side, and the network device sending the compressed model to the terminal after completing the training.

[0153] A13. Obtain the pre-configured first target model;

[0154] Alternatively, this situation can be understood as the model being trained on devices other than terminals and network devices. For example, the model is trained by the terminal manufacturer. After training the first target model and the second target model, the terminal manufacturer sends the first target model to the terminal and the second target model to the network device. In this case, the model can be understood as a pre-configured model.

[0155] It should be noted here that, when the first target model is a model trained using a bilateral model, the first training data may optionally include at least: the third downlink CSI corresponding to multiple network devices, the number of transmit antenna ports of the multiple network devices, the number of subbands occupied by the transmitted signals of the multiple network devices, and tag data information. In this case, the tag data information is the third downlink CSI corresponding to the multiple network devices.

[0156] Optionally, when the first target model is a model trained using a one-sided model, the first training data includes: the third downlink CSI corresponding to multiple network devices, the number of transmit antenna ports of the multiple network devices, the number of subbands occupied by the transmitted signals of the multiple network devices, and tag data information. Optionally, the tag data information includes first tag data and second tag data, wherein the first tag data is information obtained by compressing the third downlink CSI of the multiple network devices using the compressed sensing principle, and the second tag information is information obtained by compressing the first tag data information using the compressed sensing principle.

[0157] The specific applications of the bilateral and unilateral models are explained in detail below.

[0158] I. The first target model is the model obtained through bilateral model training.

[0159] Optionally, the design principles of the two-sided model:

[0160] To fully utilize the common sparsity of CSIs among different cooperating network devices under distributed MIMO, this application proposes a two-layer compression AI model structure: the first layer compresses downlink CSIs from each cooperating network device individually, and the second layer jointly compresses the compressed downlink CSIs from each cooperating network device to remove the correlation between downlink CSIs from each cooperating network device. The structure of the two-layer compression model is as follows: Figure 3 As shown, the first layer may include multiple AI compression models, each corresponding to the downlink CSI of each cooperative network device. The second layer includes an AI compression model that jointly compresses the outputs of the multiple AI models in the first layer to obtain the compressed downlink CSI.

[0161] Optionally, the joint compression mentioned in this application embodiment refers to compressing downlink CSIs from multiple cooperating network devices together. The purpose is not only to remove the correlation within the downlink CSIs from each cooperating network device, but also to remove the correlation between the downlink CSIs from multiple cooperating network devices.

[0162] Corresponding to the two-layer compression model, the two-layer decompression AI model also contains a two-layer structure. The first layer is used to decompress the jointly compressed data, and the second layer is used to decompress the downlink CSI of each cooperating network device separately. The specific structure is as follows: Figure 4 As shown, the first layer includes an AI decompression model that decompresses the jointly compressed information from various collaborative network devices. The second layer includes multiple AI decompression models that decompress the compressed downlink CSI corresponding to each collaborative network device after decompressing the jointly compressed information from various collaborative network devices, in order to obtain the downlink CSI corresponding to each collaborative network device.

[0163] Specifically, the process of obtaining a dual-layer compressed / decompressed AI model includes the following steps:

[0164] 1. Model Input

[0165] The terminal estimates the downlink CSI from each cooperating network device, the number of transmit antenna ports of each cooperating network device, and the number of subbands occupied by the transmitted signal.

[0166] 2. Model output

[0167] The CSI is a jointly compressed CSI from downlink CSIs from various cooperating network devices.

[0168] 3. Data Collection

[0169] Optionally, the data acquisition process during model training on the network device side includes one of the following methods:

[0170] Method 1 mainly includes:

[0171] Each terminal independently reports its corresponding uncompressed downlink CSI to each cooperating network device;

[0172] Each collaborative network device receives downlink CSI reports from the terminal;

[0173] The network device responsible for training the model obtains the downlink CSI between other cooperating network devices and the terminal at the same time through Backhaul;

[0174] The network device responsible for training the model stores the downlink CSIs acquired from all collaborating network devices according to the acquisition time.

[0175] The collected data is the input data of the first layer encoder.

[0176] Method 2 mainly includes:

[0177] The terminal jointly reports the uncompressed downlink CSI to all cooperating network devices;

[0178] Each cooperating network device (including the network device responsible for training the model) receives downlink CSI reports from all cooperating network devices reported by the terminal;

[0179] The network device responsible for training the model stores the downlink CSIs acquired from all collaborating network devices according to the acquisition time.

[0180] Optionally, the data acquisition process during model training on the terminal side is as follows:

[0181] The terminal estimates the downlink CSI from each cooperating network device based on the downlink CSI-RS sent by each cooperating network device;

[0182] The terminal stores downlink CSIs from each cooperating network device according to time.

[0183] The terminal can also report the stored information to the terminal manufacturer, who can then train the model and deploy it to the terminal.

[0184] 4. Model Training

[0185] Terminals (or terminal manufacturers) or network devices can use the collected training data and label data information to train models using the following two model structures:

[0186] The first type is a model structure based on an auto-encoder.

[0187] The input of the nth (1≤n≤N) encoder model in the first layer is the label data information of the nth decoder model in the second layer.

[0188] The first encoder is used to compress the downlink CSI from each cooperating network device separately, and the second encoder is used to jointly compress the downlink CSI from all cooperating network devices. The first and second encoders can use the same convolutional neural network or self-attention mechanism network, but the input and output dimensions are different.

[0189] The first-layer decoder is used to jointly decompress the downlink CSI from each cooperating network device, and the second-layer decoder is used to decompress the downlink CSI from each cooperating network device separately. Both the first-layer decoder and the second-layer decoder can be constructed using convolutional neural networks or self-attention mechanisms with the same structure, but the dimensions of the input and output are different.

[0190] The model training uses Squared Generalized Cosine Similarity (SGCS) as the loss function. Based on the backpropagation principle, the gradient is calculated according to the loss between the original downlink CSI data and the reconstructed data output by the decoder, and the weights of the encoder and decoder are updated until the preset number of iterations is reached, or the accuracy of SGCS no longer improves.

[0191] Optionally, the encoder corresponds to the first target model described above, and the decoder corresponds to the second target model described above.

[0192] The second type is based on the GAN model structure.

[0193] The generative adversarial network used for compression includes an encoder, a generator, and a discriminator. The encoder is responsible for compressing the data, the generator is responsible for decompressing and reconstructing the data, and the discriminator ensures the quality of the data reconstructed by the generator through adversarial training. Optionally, the encoder corresponds to the first target model mentioned above, and the generator corresponds to the second target model mentioned above.

[0194] Encoder network design: The first layer encoder is used to compress the downlink CSI from each cooperative network device separately, and the second layer encoder is used to jointly compress the downlink CSI from all cooperative network devices. Both the first layer encoder and the second layer encoder can use convolutional neural networks or self-attention networks with the same structure, but the input and output dimensions are different.

[0195] Generator Network: The first-layer generator is used to jointly decompress the downlink CSI from each cooperating network device, and the second-layer generator is used to decompress the downlink CSI from each cooperating network device separately. Both the first-layer generator and the second-layer generator can be constructed using the same convolutional neural network or self-attention mechanism, but the dimensions of the input and output are different.

[0196] Discriminator Network: Both the first-layer discriminator and the second-layer discriminator can be built based on convolutional neural networks or self-attention mechanisms to distinguish between the data generated by the generator and the real downlink CSI (i.e., the downlink CSI reported from the terminal).

[0197] The model training uses SGCS as the loss function. During the model training process, an alternating training mode is used, that is, first fix the discriminator and train the generator; then fix the generator and train the discriminator, until the preset number of iterations is reached, or the SGCS accuracy no longer improves.

[0198] 5. Model Distribution

[0199] In one scenario, depending on the model training location, model distribution based on the Auto-encoder architecture includes the following cases:

[0200] When the model is trained on the terminal, the terminal needs to send the decoder part to each cooperating network device;

[0201] When the model is trained by the terminal manufacturer, the terminal manufacturer needs to deploy the encoder model to the terminal and update it regularly. As for how to deploy the decoder model to network devices, the terminal manufacturer can first deploy the decoder to the terminal, and then the terminal sends the decoder part to each collaborating network device; or deploy it to the network device through pre-configuration or factory settings.

[0202] When the model is trained on the network device side, the network device responsible for training the model needs to send the trained encoder model to the terminal and send the decoder model to other cooperating network devices.

[0203] In another scenario, model distribution based on GAN structures includes the following cases:

[0204] When the model is trained on the terminal, the terminal needs to send the generator part to each cooperating network device;

[0205] When the model is trained by the terminal manufacturer, the terminal manufacturer needs to deploy the encoder model to the terminal and update it regularly. Regarding how the generator model is deployed to network devices, the terminal manufacturer can first deploy the generator model to the terminal, and then the terminal sends the generator part to each collaborating network device; or deploy it to the network device through pre-configuration or factory settings.

[0206] When the model is trained on the network device side, the network device responsible for training the model needs to send the trained encoder model to the terminal and send the generator model to other cooperating network devices.

[0207] 6. Model Deduction

[0208] The terminal inputs the estimated downlink CSI from each cooperating network device into the encoder of the trained Auto-encoder model or the encoder of the GAN model to jointly compress the downlink CSI from each cooperating network device, and feeds back the compressed downlink CSI to each cooperating network device. Each cooperating network device decompresses the received jointly compressed CSI based on the decoder of the Auto-encoder model or the generator of the GAN to obtain the downlink CSI of each cooperating network device.

[0209] II. The first target model is the model obtained through training a one-sided model.

[0210] Alternatively, the training process for the model used for compression (also known as the compressor model) is as follows:

[0211] 1. Input to the model used for compression

[0212] The terminal estimates the downlink CSI from each cooperating network device, the number of transmit antenna ports of each cooperating network device, and the number of subbands occupied by the transmitted signal.

[0213] 2. Output of the model used for compression

[0214] The CSI is a jointly compressed CSI from downlink CSIs from various cooperating network devices.

[0215] 3. Data acquisition for the compression model

[0216] Specifically, the model is trained on the terminal side or by the terminal manufacturer, including:

[0217] The terminal estimates the downlink CSI from each cooperating network device based on the downlink CSI-RS sent by each cooperating network device;

[0218] The terminal stores downlink CSIs from each cooperating network device according to time.

[0219] The terminal acquires tag data information, wherein the tag data includes: first tag data and second tag data. The first tag data is information obtained by compressing the third downlink CSI of multiple network devices using the principle of compressed sensing, and the second tag data is information obtained by compressing the first tag data information using the principle of compressed sensing.

[0220] Optionally, the terminal can also report the stored training data and label data to the terminal manufacturer, who can then train the model and deploy it to the terminal.

[0221] 4. Model training for compression

[0222] Terminals or terminal manufacturers utilize the collected training data and tag data information, employing methods such as... Figure 4 The two-layer compression network structure shown can be implemented in two layers. The first layer (also known as the first-layer compressor) and the second layer (also known as the second-layer compressor) can both use the same convolutional neural network or self-attention mechanism network, but the input and output dimensions are different. Based on the backpropagation principle, the gradient is calculated according to the output of the second-layer compressor and the loss of the label data, and the weights of the first-layer compressor and the second-layer compressor are updated until the preset number of iterations is reached, or the accuracy of SGCS no longer improves.

[0223] 5. Model Deduction

[0224] The terminal inputs the estimated downlink CSI from each cooperating network device into the trained compressor model, and performs joint compression on the downlink CSI of each cooperating network device to obtain the jointly compressed downlink CSI.

[0225] The training process for the decompressor model corresponding to the compressor model specifically includes:

[0226] 1. Input to the decompressor model

[0227] The input data consists of the downlink CSI data of each cooperating network device jointly compressed based on the compressed sensing principle, and the compression ratio of the compressed sensing.

[0228] 2. Output of the decompressor model

[0229] Downlink CSI from various collaborative network devices;

[0230] 3. Data Acquisition for the Decompressor Model

[0231] Optionally, the model is trained on the network device side;

[0232] Optionally, the methods for obtaining training data include:

[0233] The terminal reports data that is jointly compressed using compressed sensing of the downlink CSI of each cooperating network device.

[0234] Each cooperating network device obtains the downlink CSI of other cooperating network devices through Backhaul, and performs joint compression of the downlink CSI of each cooperating network device based on compressed sensing and using the same compression matrix as the terminal side.

[0235] Optionally, the methods for obtaining tag data include:

[0236] The base station obtains the downlink CSI of other cooperating network devices through Backhaul;

[0237] The terminal reports the downlink CSI of each cooperating network device.

[0238] 4. Decompressor Model Training

[0239] The network device uses the collected training data and label data information. The first-layer decompressor and the second-layer decompressor can both adopt the same structure of convolutional neural network or self-attention mechanism network, but the input and output dimensions are different. Based on the backpropagation principle, the gradient is calculated according to the output of the second-layer decompressor and the loss of the label data, and the weights of the first-layer decompressor and the second-layer decompressor are updated until the preset number of iterations is reached, or the accuracy of SGCS no longer improves.

[0240] 5. Model Deduction

[0241] The network device inputs the jointly compressed downlink CSI from each cooperating network device into the trained compressor model for decompression, thereby obtaining the downlink CSI from each cooperating network device.

[0242] It should be noted that, after obtaining the jointly compressed target information based on the first target model, optionally, in one implementation, sending the target information to at least two network devices includes one of the following:

[0243] A11. Send target information to at least two network devices via directional beams;

[0244] Optionally, when the terminal operates in the FR2 band, due to increased path loss, the terminal needs to use a narrower transmit beam. The terminal uses the same time-frequency resources to transmit the target information to each cooperating network device via spatial division, such as... Figure 5 As shown, three base stations (BS) provide services to the terminal. The base stations send downlink CSI-RS respectively. The terminal measures the downlink CSI corresponding to the three base stations and sends the target information obtained by jointly compressing the downlink CSI corresponding to the three base stations to the three base stations through directional beams.

[0245] A12. Transmit target information to at least two network devices via an omnidirectional antenna;

[0246] Optionally, when the terminal operates in the FR1 band, it does not need to use a narrower transmit beam. Instead, the terminal transmits the target information to each cooperating network device on a time-frequency resource negotiated by all cooperating network devices, such as... Figure 6 As shown, three base stations (BS) provide services to the terminal. The base stations send downlink CSI-RS respectively. The terminal measures the downlink CSI corresponding to the three base stations and sends the target information obtained by jointly compressing the downlink CSI corresponding to the three base stations to the three base stations through an omnidirectional antenna.

[0247] Optionally, in one implementation, the method further includes:

[0248] The terminal receives indication information sent by network devices, the indication information being used to indicate whether the terminal performs joint compression and joint feedback on the first downlink CSI from multiple network devices in cooperative communication mode;

[0249] The step of compressing the plurality of first downlink CSIs to obtain target information includes:

[0250] When the instruction information instructs the terminal to perform joint compression and joint feedback of the first downlink CSI from multiple network devices in the cooperative communication mode, the multiple first downlink CSIs are compressed to obtain the target information.

[0251] Optionally, the network device can send indication information to the terminal through the Physical downlink control channel (PDCCH). For example, the network device transmits a control bit on the PDCCH, such as the JointCSIFeedback indication bit, with a value of 0 or 1, to notify the terminal whether to perform joint compression and joint feedback of the first downlink CSI from multiple network devices in cooperative communication mode. Optionally, an alternative expression is: to notify the terminal whether, in cooperative communication mode, the terminal feeds back the downlink CSI to each cooperating network device separately, or whether the terminal feeds back the downlink CSI after joint compression of the cooperating network devices.

[0252] It should be noted that, in order to facilitate the differentiation of downlink CSI by network devices, the network devices need to know which other cooperating network devices' downlink CSIs are included in the jointly compressed downlink CSI they receive. Optionally, in one implementation, the method further includes:

[0253] The terminal sends the identification information of the network device associated with the target information to at least two network devices.

[0254] In other words, in this implementation, the terminal notifies the network device of the network device's identification information.

[0255] Alternatively, in another implementation, the control center connected to each collaborative network device notifies each collaborative network device of the identification information of the current terminal of the collaborative network device; that is, the identification information of the collaborative network device is provided by the control center to each network device.

[0256] It should be noted that at least one embodiment of this application mines the common sparsity characteristics of CSIs among different cooperating network devices under distributed MIMO. The AI ​​model used for compression includes a two-layer structure. The first layer is used to compress downlink CSIs from the cooperating network devices separately, and the second layer is used to jointly compress the compressed downlink CSIs from each cooperating network device to remove the correlation between the downlink CSIs of each cooperating network device. Then, the terminal feeds back the compressed CSIs to each cooperating network device, and each cooperating network device decompresses the received compressed downlink CSIs. The model used for decompression also includes a two-layer structure. The first layer is used to decompress the jointly compressed data, and the second layer is used to decompress the downlink CSIs of each cooperating base station separately. This allows each network device to simultaneously obtain the CSI between itself and the terminal, as well as the CSI between other cooperating network devices and the terminal. The embodiments of this application can achieve joint CSI feedback and compression under cooperating network devices with low feedback overhead, and without increasing the latency of CSI feedback.

[0257] The technical solutions provided in this application can be applied to various systems, especially 5G systems. For example, applicable systems include Global System for Mobile Communication (GSM), Code Division Multiple Access (CDMA), Wideband Code Division Multiple Access (WCDMA) General Packet Radio Service (GPRS), Long Term Evolution (LTE), LTE Frequency Division Duplex (FDD), LTE Time Division Duplex (TDD), Long Term Evolution Advanced (LTE-A), Universal Mobile Telecommunication System (UMTS), Worldwide Interoperability for Microwave Access (WiMAX), and 5G New Radio (NR). All of these systems include terminals (also called terminal equipment) and network equipment. The systems may also include a core network component, such as Evolved Packet System (EPS) and 5G system (5GS).

[0258] The terminal involved in the embodiments of this application can also be called a terminal device, which can be a device that provides voice and / or data connectivity to a user, a handheld device with wireless connectivity, or other processing devices connected to a wireless modem, etc. The name of the terminal device may also differ in different systems; for example, in a 5G system, the terminal device can be called User Equipment (UE). Wireless terminal devices can communicate with one or more core networks (CNs) via a Radio Access Network (RAN). Wireless terminal devices can be mobile terminal devices, such as mobile phones (or "cellular" phones) and computers with mobile terminal devices, for example, portable, pocket-sized, handheld, computer-embedded, or vehicle-mounted mobile devices, which exchange voice and / or data with the radio access network. Examples include Personal Communication Service (PCS) phones, cordless phones, Session Initiated Protocol (SIP) phones, Wireless Local Loop (WLL) stations, Personal Digital Assistants (PDAs), and other devices. Wireless terminal equipment can also be referred to as a system, subscriber unit, subscriber station, mobile station, mobile station, remote station, access point, remote terminal, access terminal, user terminal, user agent, or user device, but this application does not limit the terminology.

[0259] The network device involved in this application embodiment can be a base station, which may include multiple cells providing services to terminals. Depending on the specific application, a base station may also be called an access point, or a device in an access network that communicates with a wireless terminal device through one or more sectors on the air interface, or other names. The network device can be used to exchange received air frames with Internet Protocol (IP) packets, acting as a router between the wireless terminal device and the rest of the access network, where the rest of the access network may include an Internet Protocol (IP) communication network. The network device can also coordinate the attribute management of the air interface. For example, the network equipment involved in the embodiments of this application can be a base transceiver station (BTS) in a Global System for Mobile communications (GSM) or Code Division Multiple Access (CDMA), a NodeB in a Wide-band Code Division Multiple Access (WCDMA) system, an evolved Node B (eNB or e-NodeB) in a long term evolution (LTE) system, a 5G base station (gNB) in a next generation system, a Home evolved Node B (HeNB), a relay node, a femto, a pico, etc., and is not limited in the embodiments of this application. In some network structures, the network equipment may include centralized unit (CU) nodes and distributed unit (DU) nodes, and the centralized unit and distributed unit may also be geographically separated.

[0260] Network devices and terminal devices can each use one or more antennas for multiple-input multiple-output (MIMO) transmission. MIMO transmission can be single-user MIMO (SU-MIMO) or multiple-user MIMO (MU-MIMO). Depending on the configuration and number of antenna combinations, MIMO transmission can be 2D-MIMO, 3D-MIMO, FD-MIMO, or massive-MIMO, and can also be diversity transmission, precoding transmission, or beamforming transmission, etc.

[0261] like Figure 7 As shown, this application embodiment provides a channel state information acquisition method, executed by a network device, including:

[0262] Step S701: Receive target information sent by the terminal. The target information is obtained by compression of multiple first downlink channel state information (CSI), and each first downlink CSI corresponds to a network device.

[0263] Step S702: Decompress the target information to obtain multiple decompressed first downlink CSIs.

[0264] Optionally, the step of decompressing the target information to obtain multiple decompressed first downlink CSIs includes:

[0265] The target information is input into the second target model to obtain multiple decompressed first downlink CSIs output by the second target model;

[0266] The second target model is a two-layer decompression model. The first-layer decompression model decompresses the target information to obtain the fourth downlink CSI, and the second-layer multiple decompression models decompress the fourth downlink CSI respectively.

[0267] The second target model is obtained by training a bilateral model and the first training data; or, the second target model is obtained by training a unilateral model and the second training data, wherein the second training data includes: training data of the first layer decompression model, label data of the first layer decompression model, training data of the second layer decompression model, and label data of the second layer decompression model, wherein the training data of the first layer decompression model is the second label data information, the label data of the first layer decompression model is the first label data information, the training data of the second layer decompression model is the first label data information, and the label data of the second layer decompression model is the third downlink CSI from multiple network devices;

[0268] The first training data includes at least one of the following: the third downlink CSI corresponding to multiple network devices, the number of transmit antenna ports of multiple network devices, the number of subbands occupied by the transmit signals of multiple network devices, and tag data information;

[0269] The first tag data is information obtained by compressing the third downlink CSI of multiple network devices using the principle of compressed sensing, and the second tag information is information obtained by compressing the first tag data using the principle of compressed sensing.

[0270] Optionally, the bilateral model includes at least one of the following:

[0271] Autoencoder model, convolutional neural network model, generative adversarial network model, Transformer structure model.

[0272] Optionally, when the second target model is obtained by training using a bilateral model and the first training data, the method further includes at least one of the following:

[0273] The first target model is sent to the terminal. The first target model is a model used for multiple first downlink CSI compressions. The first target model is obtained by the network device through training using the first training data.

[0274] The second target model sent by the receiving terminal.

[0275] Optionally, the method further includes:

[0276] The terminal is sent an indication message, which is used to indicate whether the terminal performs joint compression and joint feedback on the first downlink CSI from multiple network devices in the cooperative communication mode.

[0277] Optionally, the method further includes one of the following:

[0278] The receiving terminal sends the identification information of the network device associated with the target information;

[0279] The terminal receives identification information of the network device that provides cooperative communication for the terminal, sent by the control center.

[0280] It should be noted that all the implementation methods in the above embodiments are applicable to the embodiments of the channel state information acquisition method applied to the network device side, and can achieve the same technical effect, so they will not be described again here.

[0281] like Figure 8 As shown, this application embodiment provides a channel state information transmission device 800, applied to a terminal, including:

[0282] The first acquisition unit 801 is used to acquire multiple first downlink channel state information (CSI), each first downlink CSI corresponding to a network device.

[0283] The second acquisition unit 802 is used to compress the plurality of first downlink CSIs to acquire target information;

[0284] The first sending unit 803 is used to send the target information to at least two network devices, wherein the at least two network devices provide coherent joint transmission for the terminal.

[0285] Optionally, the second acquisition unit 802 is configured to:

[0286] The plurality of first downlink CSIs are input into the first target model to obtain the target information output by the first target model;

[0287] The first target model is a two-layer compression model. The first layer compression model compresses multiple first downlink CSIs respectively, and the second layer compression model compresses multiple second downlink CSIs output by the first layer compression model. The first target model is obtained by training on first training data, which includes at least one of the following: third downlink CSIs corresponding to multiple network devices, the number of transmit antenna ports of multiple network devices, the number of subbands occupied by the transmit signals of multiple network devices, and tag data information.

[0288] The third downlink CSI is the downlink CSI used for model training.

[0289] Optionally, the first target model is a model trained using a two-sided model, or the first target model is a model trained using a one-sided model.

[0290] Optionally, the bilateral model includes at least one of the following:

[0291] Autoencoder model, convolutional neural network model, generative adversarial network model, Transformer structure model.

[0292] Optionally, when the first target model is a model trained using a two-sided model, the apparatus further includes at least one of the following:

[0293] The second sending unit is used to send the second target model to multiple network devices. The second target model is a model used to decompress the target information. The second target model is obtained by the terminal through training using the first training data.

[0294] The second receiving unit is used to receive the first target model sent by the network device;

[0295] The fourth acquisition unit is used to acquire the pre-configured first target model.

[0296] Optionally, when the first target model is a model trained by a bilateral model, the label data information included in the first training data is the third downlink CSI corresponding to multiple network devices respectively.

[0297] Optionally, when the first target model is a model trained by a one-sided model, the label data information included in the first training data includes first label data and second label data. The first label data is information obtained by compressing the third downlink CSI of multiple network devices using the compressed sensing principle, and the second label information is information obtained by compressing the first label data information using the compressed sensing principle.

[0298] Optionally, the first transmitting unit 803 is configured to implement one of the following:

[0299] Send target information to at least two network devices using a directional beam;

[0300] The target information is transmitted to at least two network devices via an omnidirectional antenna.

[0301] Optionally, the device further includes:

[0302] The third receiving unit is used to receive indication information sent by the network device. The indication information is used to indicate whether the terminal performs joint compression and joint feedback on the first downlink CSI from multiple network devices in the cooperative communication mode.

[0303] The second acquisition unit 802 is used for:

[0304] When the instruction information instructs the terminal to perform joint compression and joint feedback of the first downlink CSI from multiple network devices in the cooperative communication mode, the multiple first downlink CSIs are compressed to obtain the target information.

[0305] Optionally, the device further includes:

[0306] The third sending unit is used to send the identification information of the network device associated with the target information to at least two network devices.

[0307] It should be noted that this device embodiment corresponds one-to-one with the above method embodiments. All implementation methods in the above method embodiments are applicable to this device embodiment and can achieve the same technical effect.

[0308] It should be noted that the division of units in the embodiments of this application is illustrative and only represents one logical functional division. In actual implementation, other division methods may be used. Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units described above can be implemented in hardware or as software functional units.

[0309] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a processor-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0310] like Figure 9 As shown, this application embodiment also provides a terminal, including a processor 900, a transceiver 910, a memory 920, and a program stored in the memory 920 and executable on the processor 900; wherein the transceiver 910 is connected to the processor 900 and the memory 920 via a bus interface, and the processor 900 is used to read the program in the memory and execute the following processes:

[0311] Acquire multiple First Downlink Channel State Information (CSI) entries, with each First Downlink CSI corresponding to a network device.

[0312] The multiple first downlink CSIs are compressed to obtain target information;

[0313] The target information is transmitted to at least two network devices via a transceiver, and the at least two network devices provide coherent joint transmission for the terminal.

[0314] Transceiver 910 is used to receive and send data under the control of processor 900.

[0315] Among them, Figure 9In this context, the bus architecture can include any number of interconnected buses and bridges, specifically linking various circuits of one or more processors represented by processor 900 and memory represented by memory 920 together. The bus architecture can also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. The bus interface provides an interface. The transceiver 910 can be multiple components, including transmitters and receivers, providing a unit for communicating with various other devices over a transmission medium, including wireless channels, wired channels, optical fibers, etc. For different user equipment, the user interface 930 can also be an interface capable of connecting external or internal devices, including but not limited to keypads, displays, speakers, microphones, joysticks, etc.

[0316] The processor 900 is responsible for managing the bus architecture and general processing, while the memory 920 can store the data used by the processor 900 during operation.

[0317] Optionally, the processor 900 can be a CPU (Central Processing Unit), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array), or CPLD (Complex Programmable Logic Device), and the processor can also adopt a multi-core architecture.

[0318] The processor executes any of the methods described in the embodiments of this application according to the obtained executable instructions by calling a computer program stored in memory. The processor and memory may also be physically separated.

[0319] Optionally, the processor is configured to read the computer program in the memory and perform the following operations:

[0320] The plurality of first downlink CSIs are input into the first target model to obtain the target information output by the first target model;

[0321] The first target model is a two-layer compression model. The first layer compression model compresses multiple first downlink CSIs respectively, and the second layer compression model compresses multiple second downlink CSIs output by the first layer compression model. The first target model is obtained by training on first training data, which includes at least one of the following: third downlink CSIs corresponding to multiple network devices, the number of transmit antenna ports of multiple network devices, the number of subbands occupied by the transmit signals of multiple network devices, and tag data information.

[0322] The third downlink CSI is the downlink CSI used for model training.

[0323] Optionally, the first target model is a model trained using a two-sided model, or the first target model is a model trained using a one-sided model.

[0324] Optionally, the bilateral model includes at least one of the following:

[0325] Autoencoder model, convolutional neural network model, generative adversarial network model, Transformer structure model.

[0326] Optionally, if the first target model is a model trained using a bilateral model, the processor, when reading the computer program from the memory, further performs at least one of the following operations:

[0327] The second target model is sent to multiple network devices. The second target model is a model used to decompress the target information. The second target model is obtained by the terminal through training using the first training data.

[0328] The first target model sent by the network device;

[0329] Obtain the pre-configured first target model.

[0330] Optionally, when the first target model is a model trained by a bilateral model, the label data information included in the first training data is the third downlink CSI corresponding to multiple network devices respectively.

[0331] Optionally, when the first target model is a model trained by a one-sided model, the label data information included in the first training data includes first label data and second label data. The first label data is information obtained by compressing the third downlink CSI of multiple network devices using the compressed sensing principle, and the second label information is information obtained by compressing the first label data information using the compressed sensing principle.

[0332] Optionally, the processor is configured to read a computer program from the memory and perform one of the following operations:

[0333] Send target information to at least two network devices using a directional beam;

[0334] The target information is transmitted to at least two network devices via an omnidirectional antenna.

[0335] Optionally, the processor, for reading the computer program in the memory, further performs the following operations:

[0336] The terminal receives indication information sent by network devices through a transceiver. The indication information is used to indicate whether the terminal should perform joint compression and joint feedback on the first downlink CSI from multiple network devices in cooperative communication mode.

[0337] When the instruction information instructs the terminal to perform joint compression and joint feedback of the first downlink CSI from multiple network devices in the cooperative communication mode, the multiple first downlink CSIs are compressed to obtain the target information.

[0338] Optionally, the processor, for reading the computer program in the memory, further performs the following operations:

[0339] The target information is transmitted via a transceiver to at least two network devices, along with the identification information of the network device associated with it.

[0340] It should be noted that the terminal provided in this application embodiment can implement all the method steps implemented in the above method embodiment and can achieve the same technical effect. Here, the parts that are the same as those in the method embodiment and the beneficial effects will not be described in detail.

[0341] This application also provides a computer-readable storage medium storing a computer program thereon, wherein the computer program, when executed by a processor, implements the steps of a channel state information transmission method applied to a terminal. The processor-readable storage medium can be any available medium or data storage device accessible to the processor, including but not limited to magnetic storage (e.g., floppy disks, hard disks, magnetic tapes, magneto-optical disks (MO), etc.), optical storage (e.g., CDs, DVDs, BDs, HVDs, etc.), and semiconductor storage (e.g., ROMs, EPROMs, EEPROMs, non-volatile memory (NAND flash), solid-state drives (SSDs), etc.).

[0342] like Figure 10 As shown, this application embodiment provides a channel state information acquisition device 1000, applied to a network device, including:

[0343] The first receiving unit 1001 is used to receive target information sent by the terminal. The target information is obtained by compression of multiple first downlink channel state information (CSI) and each first downlink CSI corresponds to a network device.

[0344] The third acquisition unit 1002 is used to decompress the target information and acquire multiple decompressed first downlink CSIs.

[0345] Optionally, the third acquisition unit 1002 is used for:

[0346] The target information is input into the second target model to obtain multiple decompressed first downlink CSIs output by the second target model;

[0347] The second target model is a two-layer decompression model. The first-layer decompression model decompresses the target information to obtain the fourth downlink CSI, and the second-layer multiple decompression models decompress the fourth downlink CSI respectively.

[0348] The second target model is obtained by training a bilateral model and the first training data; or, the second target model is obtained by training a unilateral model and the second training data, wherein the second training data includes: training data of the first layer decompression model, label data of the first layer decompression model, training data of the second layer decompression model, and label data of the second layer decompression model, wherein the training data of the first layer decompression model is the second label data information, the label data of the first layer decompression model is the first label data information, the training data of the second layer decompression model is the first label data information, and the label data of the second layer decompression model is the third downlink CSI from multiple network devices;

[0349] The first training data includes at least one of the following: the third downlink CSI corresponding to multiple network devices, the number of transmit antenna ports of multiple network devices, the number of subbands occupied by the transmit signals of multiple network devices, and tag data information;

[0350] The first tag data is information obtained by compressing the third downlink CSI of multiple network devices using the principle of compressed sensing, and the second tag information is information obtained by compressing the first tag data using the principle of compressed sensing.

[0351] Optionally, the bilateral model includes at least one of the following:

[0352] Autoencoder model, convolutional neural network model, generative adversarial network model, Transformer structure model.

[0353] Optionally, when the second target model is obtained by training using a bilateral model and the first training data, the apparatus further includes at least one of the following:

[0354] The fourth sending unit is used to send the first target model to the terminal. The first target model is a model for performing multiple first downlink CSI compressions. The first target model is obtained by the network device through training using the first training data.

[0355] The fourth receiving unit is used to receive the second target model sent by the terminal.

[0356] Optionally, the device further includes:

[0357] The fifth sending unit is used to send indication information to the terminal, the indication information being used to indicate whether the terminal performs joint compression and joint feedback on the first downlink CSI from multiple network devices in the cooperative communication mode.

[0358] Optionally, the device further includes one of the following:

[0359] The fifth receiving unit is used to receive the identification information of the network device associated with the target information sent by the terminal;

[0360] The sixth receiving unit is used to receive the identification information of the network device that provides cooperative communication for the terminal, sent by the control center.

[0361] It should be noted that this device embodiment corresponds one-to-one with the above method embodiments. All implementation methods in the above method embodiments are applicable to this device embodiment and can achieve the same technical effect.

[0362] It should be noted that the division of units in the embodiments of this application is illustrative and only represents one logical functional division. In actual implementation, other division methods may be used. Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units described above can be implemented in hardware or as software functional units.

[0363] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a processor-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0364] like Figure 11 As shown, this application embodiment also provides a network device, including a processor 1100, a transceiver 1110, a memory 1120, and a program stored in the memory 1120 and executable on the processor 1100; wherein the transceiver 1110 is connected to the processor 1100 and the memory 1120 via a bus interface, wherein the processor 1100 is used to read the program in the memory and execute the following process: wherein the processor is used to read the computer program in the memory and perform the following operations:

[0365] The target information is received by the transceiver terminal. The target information is obtained by compression of multiple first downlink channel state information (CSI) and each first downlink CSI corresponds to a network device.

[0366] The target information is decompressed to obtain multiple decompressed first downlink CSIs.

[0367] Transceiver 1110 is used to receive and send data under the control of processor 1100.

[0368] Among them, Figure 11In this context, the bus architecture can include any number of interconnected buses and bridges, specifically linking various circuits of one or more processors represented by processor 1100 and memory represented by memory 1120 together. The bus architecture can also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. The bus interface provides an interface. The transceiver 1110 can be multiple components, including transmitters and receivers, providing a unit for communicating with various other devices over a transmission medium, including wireless channels, wired channels, optical fibers, etc. For different user equipment, the user interface 1130 can also be an interface capable of connecting external or internal devices, including but not limited to keypads, displays, speakers, microphones, joysticks, etc.

[0369] The processor 1100 is responsible for managing the bus architecture and general processing, and the memory 1120 can store the data used by the processor 1100 when performing operations.

[0370] Optionally, the processor 1100 can be a CPU (Central Processing Unit), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array), or CPLD (Complex Programmable Logic Device), and the processor can also adopt a multi-core architecture.

[0371] The processor executes any of the methods described in the embodiments of this application according to the obtained executable instructions by calling a computer program stored in memory. The processor and memory may also be physically separated.

[0372] Optionally, the processor is configured to read the computer program in the memory and perform the following operations:

[0373] The target information is input into the second target model to obtain multiple decompressed first downlink CSIs output by the second target model;

[0374] The second target model is a two-layer decompression model. The first-layer decompression model decompresses the target information to obtain the fourth downlink CSI, and the second-layer multiple decompression models decompress the fourth downlink CSI respectively.

[0375] The second target model is obtained by training a bilateral model and the first training data; or, the second target model is obtained by training a unilateral model and the second training data, wherein the second training data includes: training data of the first layer decompression model, label data of the first layer decompression model, training data of the second layer decompression model, and label data of the second layer decompression model, wherein the training data of the first layer decompression model is the second label data information, the label data of the first layer decompression model is the first label data information, the training data of the second layer decompression model is the first label data information, and the label data of the second layer decompression model is the third downlink CSI from multiple network devices;

[0376] The first training data includes at least one of the following: the third downlink CSI corresponding to multiple network devices, the number of transmit antenna ports of multiple network devices, the number of subbands occupied by the transmit signals of multiple network devices, and tag data information;

[0377] The first tag data is information obtained by compressing the third downlink CSI of multiple network devices using the principle of compressed sensing, and the second tag information is information obtained by compressing the first tag data using the principle of compressed sensing.

[0378] Optionally, the bilateral model includes at least one of the following:

[0379] Autoencoder model, convolutional neural network model, generative adversarial network model, Transformer structure model.

[0380] Optionally, when the second target model is acquired through training using a bilateral model and the first training data, the processor, for reading the computer program in the memory, further performs at least one of the following operations:

[0381] The first target model is sent to the terminal. The first target model is a model used for multiple first downlink CSI compressions. The first target model is obtained by the network device through training using the first training data.

[0382] The second target model sent by the receiving terminal.

[0383] Optionally, the processor, for reading the computer program in the memory, further performs the following operations:

[0384] The transceiver sends an indication message to the terminal, which is used to indicate whether the terminal should perform joint compression and joint feedback on the first downlink CSI from multiple network devices in cooperative communication mode.

[0385] Optionally, the processor, for reading the computer program in the memory, further performs one of the following operations:

[0386] The transceiver receives the identification information of the network device associated with the target information sent by the terminal;

[0387] The transceiver receives identification information of the network devices that provide cooperative communication for the terminal, sent by the control center.

[0388] It should be noted that the network device provided in this application embodiment can implement all the method steps implemented in the above method embodiment and can achieve the same technical effect. Here, the parts that are the same as those in the method embodiment and the beneficial effects will not be described in detail.

[0389] This application also provides a computer-readable storage medium storing a computer program thereon, wherein the computer program, when executed by a processor, implements the steps of a method for acquiring channel state information applied to a network device. The processor-readable storage medium can be any available medium or data storage device accessible to the processor, including but not limited to magnetic storage (e.g., floppy disks, hard disks, magnetic tapes, magneto-optical disks (MO), etc.), optical storage (e.g., CDs, DVDs, BDs, HVDs, etc.), and semiconductor storage (e.g., ROMs, EPROMs, EEPROMs, non-volatile memory (NAND flash), solid-state drives (SSDs)).

[0390] This application also provides a computer program product, including computer instructions. When these computer instructions are executed by a processor, they implement the various processes in the above method embodiments and achieve the same technical effects. To avoid repetition, they will not be described again here.

[0391] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0392] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-executable instructions. These computer-executable instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0393] These processor-executable instructions may also be stored in a processor-readable memory that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the processor-readable memory produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0394] These processors can execute instructions that can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable device for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0395] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for transmitting channel state information, characterized in that, Applied to terminals, including: Acquire multiple First Downlink Channel State Information (CSI) entries, with each First Downlink CSI corresponding to a network device. The multiple first downlink CSIs are compressed to obtain target information; The target information is sent to at least two network devices, which provide coherent joint transmission for the terminal.

2. The method according to claim 1, characterized in that, The step of compressing the plurality of first downlink CSIs to obtain target information includes: The plurality of first downlink CSIs are input into the first target model to obtain the target information output by the first target model; The first target model is a two-layer compression model. The first layer compression model compresses multiple first downlink CSIs respectively, and the second layer compression model compresses multiple second downlink CSIs output by the first layer compression model. The first target model is obtained by training on first training data, which includes at least one of the following: third downlink CSIs corresponding to multiple network devices, the number of transmit antenna ports of multiple network devices, the number of subbands occupied by the transmit signals of multiple network devices, and tag data information. The third downlink CSI is the downlink CSI used for model training.

3. The method according to claim 2, characterized in that, The first target model is a model trained using a two-sided model, or the first target model is a model trained using a one-sided model.

4. The method according to claim 3, characterized in that, The bilateral model includes at least one of the following: Autoencoder model, convolutional neural network model, generative adversarial network model, Transformer structure model.

5. The method according to claim 3 or 4, characterized in that, When the first target model is a model trained using a bilateral model, the method further includes at least one of the following: The second target model is sent to multiple network devices. The second target model is a model used to decompress the target information. The second target model is obtained by the terminal through training using the first training data. The first target model sent by the network device; Obtain the pre-configured first target model.

6. The method according to claim 3, characterized in that, When the first target model is a model trained by a bilateral model, the label data information included in the first training data is the third downlink CSI corresponding to multiple network devices respectively.

7. The method according to claim 3, characterized in that, When the first target model is a model obtained by training a one-sided model, the label data information included in the first training data includes first label data and second label data. The first label data is information obtained by compressing the third downlink CSI of multiple network devices using the compressed sensing principle, and the second label information is information obtained by compressing the first label data information using the compressed sensing principle.

8. The method according to claim 1, characterized in that, Sending the target information to at least two network devices includes one of the following: Send target information to at least two network devices using a directional beam; The target information is transmitted to at least two network devices via an omnidirectional antenna.

9. The method according to claim 1, characterized in that, Also includes: The terminal receives indication information sent by network devices, the indication information being used to indicate whether the terminal performs joint compression and joint feedback on the first downlink CSI from multiple network devices in cooperative communication mode; The step of compressing the plurality of first downlink CSIs to obtain target information includes: When the instruction information instructs the terminal to perform joint compression and joint feedback of the first downlink CSI from multiple network devices in the cooperative communication mode, the multiple first downlink CSIs are compressed to obtain the target information.

10. The method according to claim 1, characterized in that, Also includes: Send the identification information of the network device associated with the target information to at least two network devices.

11. A method for obtaining channel state information, characterized in that, Applied to network devices, including: The target information sent by the receiving terminal is obtained by compression of multiple first downlink channel state information (CSI), and each first downlink CSI corresponds to a network device. The target information is decompressed to obtain multiple decompressed first downlink CSIs.

12. The method according to claim 11, characterized in that, The step of decompressing the target information to obtain multiple decompressed first downlink CSIs includes: The target information is input into the second target model to obtain multiple decompressed first downlink CSIs output by the second target model; The second target model is a two-layer decompression model. The first-layer decompression model decompresses the target information to obtain the fourth downlink CSI, and the second-layer multiple decompression models decompress the fourth downlink CSI respectively. The second target model is obtained by training a bilateral model and the first training data; or, the second target model is obtained by training a unilateral model and the second training data, wherein the second training data includes: training data of the first layer decompression model, label data of the first layer decompression model, training data of the second layer decompression model, and label data of the second layer decompression model, wherein the training data of the first layer decompression model is the second label data information, the label data of the first layer decompression model is the first label data information, the training data of the second layer decompression model is the first label data information, and the label data of the second layer decompression model is the third downlink CSI from multiple network devices; The first training data includes at least one of the following: the third downlink CSI corresponding to multiple network devices, the number of transmit antenna ports of multiple network devices, the number of subbands occupied by the transmit signals of multiple network devices, and tag data information; The first tag data is information obtained by compressing the third downlink CSI of multiple network devices using the principle of compressed sensing, and the second tag information is information obtained by compressing the first tag data using the principle of compressed sensing.

13. The method according to claim 12, characterized in that, The bilateral model includes at least one of the following: Autoencoder model, convolutional neural network model, generative adversarial network model, Transformer structure model.

14. The method according to claim 12 or 13, characterized in that, When the second target model is obtained by training a bilateral model and the first training data, the method further includes at least one of the following: The first target model is sent to the terminal. The first target model is a model used for multiple first downlink CSI compressions. The first target model is obtained by the network device through training using the first training data. The second target model sent by the receiving terminal.

15. The method according to claim 11, characterized in that, Also includes: The terminal is sent an indication message, which is used to indicate whether the terminal performs joint compression and joint feedback on the first downlink CSI from multiple network devices in the cooperative communication mode.

16. The method according to claim 11, characterized in that, It also includes the following: The receiving terminal sends the identification information of the network device associated with the target information; The terminal receives identification information of the network device that provides cooperative communication for the terminal, sent by the control center.

17. A terminal, characterized in that, Includes memory, transceiver, and processor: A memory for storing computer programs; a transceiver for sending and receiving data under the control of the processor; and a processor for reading the computer programs from the memory and performing the following operations: Acquire multiple First Downlink Channel State Information (CSI) entries, with each First Downlink CSI corresponding to a network device. The multiple first downlink CSIs are compressed to obtain target information; The target information is transmitted to at least two network devices via a transceiver, and the at least two network devices provide coherent joint transmission for the terminal.

18. The terminal according to claim 17, characterized in that, The processor is configured to read the computer program in the memory and perform the following operations: The plurality of first downlink CSIs are input into the first target model to obtain the target information output by the first target model; The first target model is a two-layer compression model. The first layer compression model compresses multiple first downlink CSIs respectively, and the second layer compression model compresses multiple second downlink CSIs output by the first layer compression model. The first target model is obtained by training on first training data, which includes at least one of the following: third downlink CSIs corresponding to multiple network devices, the number of transmit antenna ports of multiple network devices, the number of subbands occupied by the transmit signals of multiple network devices, and tag data information. The third downlink CSI is the downlink CSI used for model training.

19. The terminal according to claim 18, characterized in that, The first target model is a model trained using a two-sided model, or the first target model is a model trained using a one-sided model.

20. The terminal according to claim 19, characterized in that, The bilateral model includes at least one of the following: Autoencoder model, convolutional neural network model, generative adversarial network model, Transformer structure model.

21. The terminal according to claim 19 or 20, characterized in that, When the first target model is a model trained using a bilateral model, the processor, for reading the computer program in the memory, further performs at least one of the following operations: The second target model is sent to multiple network devices. The second target model is a model used to decompress the target information. The second target model is obtained by the terminal through training using the first training data. The first target model sent by the network device; Obtain the pre-configured first target model.

22. The terminal according to claim 19, characterized in that, When the first target model is a model trained by a bilateral model, the label data information included in the first training data is the third downlink CSI corresponding to multiple network devices respectively.

23. The terminal according to claim 19, characterized in that, When the first target model is a model obtained by training a one-sided model, the label data information included in the first training data includes first label data and second label data. The first label data is information obtained by compressing the third downlink CSI of multiple network devices using the compressed sensing principle, and the second label information is information obtained by compressing the first label data information using the compressed sensing principle.

24. The terminal according to claim 17, characterized in that, The processor is configured to read a computer program from the memory and perform one of the following operations: Send target information to at least two network devices using a directional beam; The target information is transmitted to at least two network devices via an omnidirectional antenna.

25. The terminal according to claim 17, characterized in that, The processor, for reading the computer program in the memory, also performs the following operations: The terminal receives indication information sent by network devices through a transceiver. The indication information is used to indicate whether the terminal should perform joint compression and joint feedback on the first downlink CSI from multiple network devices in cooperative communication mode. When the instruction information instructs the terminal to perform joint compression and joint feedback of the first downlink CSI from multiple network devices in the cooperative communication mode, the multiple first downlink CSIs are compressed to obtain the target information.

26. The terminal according to claim 17, characterized in that, The processor, for reading the computer program in the memory, also performs the following operations: The target information is transmitted via a transceiver to at least two network devices, along with the identification information of the network device associated with it.

27. A network device, characterized in that, Includes memory, transceiver, and processor: A memory for storing computer programs; a transceiver for sending and receiving data under the control of the processor; and a processor for reading the computer programs from the memory and performing the following operations: The target information is received by the transceiver terminal. The target information is obtained by compression of multiple first downlink channel state information (CSI) and each first downlink CSI corresponds to a network device. The target information is decompressed to obtain multiple decompressed first downlink CSIs.

28. The network device according to claim 27, characterized in that, The processor is configured to read the computer program in the memory and perform the following operations: The target information is input into the second target model to obtain multiple decompressed first downlink CSIs output by the second target model; The second target model is a two-layer decompression model. The first-layer decompression model decompresses the target information to obtain the fourth downlink CSI, and the second-layer multiple decompression models decompress the fourth downlink CSI respectively. The second target model is obtained by training a bilateral model and the first training data; or, the second target model is obtained by training a unilateral model and the second training data, wherein the second training data includes: training data of the first layer decompression model, label data of the first layer decompression model, training data of the second layer decompression model, and label data of the second layer decompression model, wherein the training data of the first layer decompression model is the second label data information, the label data of the first layer decompression model is the first label data information, the training data of the second layer decompression model is the first label data information, and the label data of the second layer decompression model is the third downlink CSI from multiple network devices; The first training data includes at least one of the following: the third downlink CSI corresponding to multiple network devices, the number of transmit antenna ports of multiple network devices, the number of subbands occupied by the transmit signals of multiple network devices, and tag data information; The first tag data is information obtained by compressing the third downlink CSI of multiple network devices using the principle of compressed sensing, and the second tag information is information obtained by compressing the first tag data using the principle of compressed sensing.

29. The network device according to claim 28, characterized in that, The bilateral model includes at least one of the following: Autoencoder model, convolutional neural network model, generative adversarial network model, Transformer structure model.

30. The network device according to claim 28 or 29, characterized in that, When the second target model is acquired through training using a bilateral model and the first training data, the processor, for reading the computer program in the memory, further performs at least one of the following operations: The first target model is sent to the terminal. The first target model is a model used for multiple first downlink CSI compressions. The first target model is obtained by the network device through training using the first training data. The second target model sent by the receiving terminal.

31. The network device according to claim 27, characterized in that, The processor, for reading the computer program in the memory, also performs the following operations: The transceiver sends an indication message to the terminal, which is used to indicate whether the terminal should perform joint compression and joint feedback on the first downlink CSI from multiple network devices in cooperative communication mode.

32. The network device according to claim 27, characterized in that, The processor, for reading the computer program in the memory, also performs one of the following operations: The transceiver receives the identification information of the network device associated with the target information sent by the terminal; The transceiver receives identification information of the network devices that provide cooperative communication for the terminal, sent by the control center.

33. A channel state information transmission device, applied to a terminal, characterized in that, include: The first acquisition unit is used to acquire multiple first downlink channel state information (CSI), each first downlink CSI corresponding to a network device. The second acquisition unit is used to compress the plurality of first downlink CSIs to acquire target information; The first sending unit is used to send the target information to at least two network devices, wherein the at least two network devices provide coherent joint transmission for the terminal.

34. A channel state information acquisition device, applied to network equipment, characterized in that, include: The first receiving unit is used to receive target information sent by the terminal. The target information is obtained by compression of multiple first downlink channel state information (CSI) and each first downlink CSI corresponds to a network device. The third acquisition unit is used to decompress the target information and acquire multiple decompressed first downlink CSIs.

35. A processor-readable storage medium, characterized in that, The processor-readable storage medium stores a computer program for causing the processor to perform the method according to any one of claims 1 to 16.

Images