A channel state information acquisition method based on conditional generative adversarial network

Abstract

The present invention belongs to the field of wireless communication technology, and specifically is a method for acquiring channel state information based on a conditional generative adversarial network. The present invention is based on CSI fusion and a conditional generative adversarial network (GAN), and serves large-scale MIMO scenarios; first, the CSI based on the uplink reference signal (SRS) and the downlink feedback channel CSI are fused, and the generator receives the fused CSI to generate a predicted CSI of another frequency band. The discriminator optimizes the generator performance by comparing the real CSI with the generated predicted CSI, thereby improving the accuracy of the prediction. Experimental results show that the present invention can effectively improve the acquisition accuracy of channel state information in the FDD system, realize the fusion of SRS and downlink feedback, and the conversion of CSI between frequency bands. The present invention is suitable for wireless communications, 5G / 6GMIMO systems and other application scenarios that require efficient transmission of CSI.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communications, and in particular relates to a method for acquiring channel state information. Background Art

[0002] In frequency division duplexing (FDD) massive multiple-input multiple-output (MIMO) systems, accurate downlink channel state information (CSI) is crucial for efficient beamforming, interference mitigation, and resource allocation. However, traditional high-precision CSI acquisition methods often rely on complex computations and expensive hardware resources, limiting their application in practical wireless communication systems. Therefore, developing a low-computational-complexity, high-precision downlink CSI acquisition method has become a research priority for optimizing FDD system performance.

[0003] In FDD systems, accurate CSI acquisition is challenging due to uplink and downlink frequency differences and channel non-reciprocity. Channel feedback can assist the transmitter by providing downlink CSI, but this increases system load. Although methods such as codebook design and channel quantization have been developed to reduce feedback overhead, they still struggle to meet accuracy requirements in high-dimensional CSI, impacting system capacity.

[0004] In recent years, machine learning, particularly Generative Adversarial Networks (GANs), has emerged as an effective method for mapping uplink and downlink channel information. Through adversarial learning between a generator and a discriminator, GANs can establish a nonlinear mapping relationship between uplink and downlink channels in FDD systems. This technology leverages uplink channel information to generate corresponding downlink channel state information (CSI). This improves CSI accuracy without increasing feedback bandwidth, thereby enhancing beamforming and resource allocation efficiency in FDD massive MIMO systems. Summary of the Invention

[0005] The purpose of the present invention is to provide a channel state information acquisition method based on a conditional generative adversarial network to improve the accuracy of FDD downlink CSI acquisition and achieve higher-precision channel estimation, thereby improving the overall performance and capacity of the FDD MIMO system.

[0006] The present invention provides a channel state information acquisition method based on a conditional generative adversarial network (CAN), which is based on CSI fusion and a conditional generative adversarial network (GAN) and serves large-scale MIMO scenarios. GAN uses the mutual adversarial learning between the generator and the discriminator to establish a nonlinear mapping relationship between uplink and downlink channels in the FDD system. The uplink channel information is used to generate the corresponding downlink channel state information (CSI), thereby improving the accuracy of CSI acquisition without increasing the feedback bandwidth, thereby improving the beamforming and resource allocation efficiency of the FDD large-scale MIMO system. The specific steps are:

[0007] Step 1: Set environmental parameters such as the scene, motion trajectory, and antenna layout, select a channel model, and configure key parameters; generate multipath channel data through experiments or simulations; and generate a channel covariance matrix based on the channel data;

[0008] Step 2: Map the data in the channel matrix to the three channels (red, green, and blue) of the RGB image. The fusion module fuses the uplink reference signal (SRS) channel information with the downlink channel information to generate fused CSI data.

[0009] Step 3: The generator model in the conditional generative adversarial network receives the fused CSI data and generates the predicted CSI for another frequency band.

[0010] Step 4: Use the discriminator model in the conditional generative adversarial network to compare the generated predicted CSI with the true CSI to optimize the discriminator performance. Based on the feedback from the discriminator, the generator is iteratively optimized to improve the accuracy of inter-band CSI conversion.

[0011] Step 5: Performance evaluation, specifically using normalized mean square error (NMSE) and cosine similarity to evaluate the model-generated images to comprehensively measure its pixel-level accuracy and subspace alignment quality.

[0012] The present invention is described in further detail below:

[0013] The specific process of step 1 is:

[0014] Consider establishing a MIMO wireless channel between a base station equipped with a ±45° dual-polarized uniform planar array (UPA) antenna and a user equipped with a single-polarized UPA. Assume that the channel between the user and the base station consists of L paths, each path corresponds to a direction, and the channel matrix is for:

[0015]

[0016] Among them, NBS and N UE are the number of antennas on the base station side and the user side respectively; L is the number of paths; α l is the complex path gain of path l, including path loss and phase; a BS (θ BS,l ,φ BS,l ) is the direction vector of the UPA array at the base station side; α UE (θ UE,l ,φ UE,l ) is the direction vector of the user-side UPA array, where θ is the elevation angle and φ is the azimuth angle, describing the signal propagation direction in the vertical and horizontal planes, respectively. The specific steps for constructing the array direction vector at the transmitter or receiver are as follows.

[0017] For ±45° dual-polarization UPA, there is N in the vertical direction. V antennas, there are N antennas in the horizontal direction H The antenna spacing in the vertical and horizontal directions is d V and d H . Its array response vector is defined as

[0018]

[0019] in represents the amplitude component, specifically, Respectively represent the field patterns (FP) in the vertical and horizontal directions of the +45° polarized antenna. Correspondingly, Represents the FP in the vertical and horizontal directions of the -45° polarized antenna, respectively. Represents the phase component and satisfies:

[0020]

[0021] Where 1≤i≤N V ,1≤j≤N H f∈{UL,DL}, where UL and DL represent the carrier frequencies of the uplink and downlink, respectively, and d v and d H The antenna spacing in the vertical and horizontal directions, N V and N H are the number of antennas in the horizontal and vertical directions, respectively. c represents the speed of light, c≈3.00×10 8 m / s.

[0022] The channel matrix H can be modeled as a complex Gaussian matrix, whose covariance matrix R H Expressed as:

[0023] R H =E[HHH ], (5)

[0024] In practice, R can be approximated by averaging multiple channel realizations. H .

[0025] In step 2, the fusion module is used to fuse the received uplink SRS and downlink PMI feedback channel information to generate more accurate and robust channel state information data. The specific process is as follows:

[0026] (1) The process of the base station obtaining quantized feedback channel state information is as follows:

[0027] ① The base station sends the CSI-RS reference signal s CSI-RS , the received signal model on the user side is:

[0028] y UE =H d Vs CSI-RS +w, (6)

[0029] in, represents the receiving vector, represents the downlink channel matrix, represents the outer weight matrix, is the noise at the receiving end.

[0030] ② The user side estimates the downlink equivalent channel H e , the corresponding equivalent channel matrix is:

[0031]

[0032] ③And calculate the corresponding covariance matrix R e ;

[0033]

[0034] Where T and J represent the number of frequency domain subcarriers and OFDM symbols, respectively, and P is the number of ports.

[0035] ④ The user side quantizes the downlink CSI based on the protocol to obtain the quantization codeword And feed back to the base station.

[0036] ⑤ Then obtain the PMI downlink feedback covariance information based on the feedback vector

[0037] ⑵ Fusion of uplink SRS and downlink PMI feedback channel state information

[0038] ① Mapping: such as Figure 1 , in order to implement the conditional generative adversarial network, we first get the uplink SRS channel covariance matrix and the channel covariance matrix of the downlink PMI quantization feedback Generate a pseudo-color image. This process involves separating the real and imaginary parts of the matrix and mapping them to the RGB channels of an image with dimensions N×N×3, where N is the number of antennas at the base station. In this representation, real values ​​correspond to the red channel, imaginary values ​​correspond to the green channel, and the blue channel is filled with a constant value.

[0039] ② Normalization: To facilitate convergence of the subsequent neural network training, all samples are normalized to between [-1, 1]. Due to the particularity of the channel matrix, the normalization method of image processing cannot be used. In this paper, the channel matrix is ​​uniformly divided by a constant (determined by the absolute value of the matrix) to scale it.

[0040] ③ Fusion: The image generated based on the uplink channel covariance matrix of SRS and the downlink feedback covariance matrix is ​​then fused in the third dimension to obtain the fused sample like Figure 2 The blue channel does not provide any information and is removed in actual training.

[0041] In step 3, the generator model in the conditional generative adversarial network receives the fused sample data and generates predicted channel covariance samples of another frequency band.

[0042] (1) Building a conditional generative adversarial network

[0043] like Figure 2 As shown in Figure 1, a conditional generative adversarial network (CGN) is a generative model that introduces conditional constraints to guide the generation of specific attributes or features of samples. The generator generates samples based on random noise and conditional variables, while the discriminator determines the authenticity of the generated samples and whether they meet the conditions. Through this adversarial training, the generated samples are not only realistic but also consistent with the input conditions.

[0044] ①Build a generator: such as Figure 3 As shown in the figure, the generator uses a U-Net network structure. U-Net is a classic generator structure that employs a symmetric encoder-decoder architecture and shares feature information between the encoder and decoder through skip connections. The encoder extracts progressively more abstract features, while the decoder restores image resolution through upsampling, while incorporating high-resolution features from the encoder to preserve details. The final output layer generates the target image through convolution. In conditional generative adversarial networks, using U-Net as the generator enables efficient feature extraction and restoration, resulting in a more accurate downlink channel covariance image.

[0045] ②Build the discriminator: Figure 4As shown in Figure 1, the discriminator uses a convolutional neural network (CNN) architecture to distinguish generated images from real images. It consists of multiple convolutional layers and activation functions, gradually extracting high-dimensional features from the image. Through downsampling, the discriminator captures local spatial information and ultimately outputs a scalar value through a fully connected layer, representing the probability that the input image is real.

[0046] (2) The conditional variable (the fused sample covariance matrix R fusion ) is input to the generator together with the random noise vector z, and the conditional variable is used to guide the generator to generate samples with specific characteristics.

[0047] The generator uses the input random noise and conditional variables to generate pseudo samples G(z|R fusion ), the purpose of generating samples is to be as close as possible to the real data distribution.

[0048] The discriminator receives the generated sample G(z|R fusion ) and the real sample R DL , combined with the conditional variable R fusion , judging whether the input comes from the true downlink covariance R DL Or a generator.

[0049] The discriminator outputs a probability value D(G(z|R fusion )|R fusion ) or D(R DL |R fusion ), indicating that the network judges the input samples (z|R fusion ) and R DL Under the given condition R fusion The probability value of whether it is a true sample.

[0050] In step 4, the network parameters are optimized and iteratively optimized using the adversarial loss function, alternating between the generator and the discriminator through backpropagation:

[0051]

[0052] The discriminator D estimates the probability of whether the data is a real downlink channel based on the input. The last term in the objective function represents the pixel-by-pixel difference between the generated image and the corresponding "real" image, and imposes a penalty on the generator based on the size of this difference.

[0053] The goal of the generator is to minimize the objective function, while the discriminator strives to maximize the objective function, forming the following minimum maximum optimization framework:

[0054] G * =arg minG max D L CGAN (G,D), (10)

[0055] Through this iterative optimization method, the generator and discriminator gradually improve their performance in the confrontation, and the sample distribution finally generated is close to the real data distribution.

[0056] The batch size is selected based on the dataset size and hardware conditions, and the training cycle is set based on convergence requirements. To improve training stability, the training balance between the generator and the discriminator can be dynamically adjusted. For example, the discriminator can be updated after every three training epochs. This approach prevents the discriminator from overfitting the training data, ensuring that the generator has sufficient time to improve the quality of generated images. The learning rate can also be adjusted dynamically, ultimately achieving optimal model performance through experimental tuning.

[0057] In step 5, performance evaluation is different from the performance evaluation of traditional image processing training. The present invention uses normalized mean squared error (NMSE) and the cosine of the one-dimensional subspace angle of the matrix to measure the similarity between the downlink channel covariance matrix generated by the generator and the true downlink channel covariance matrix.

[0058] The present invention first fuses the CSI based on the uplink reference signal (SRS) and the CSI of the downlink feedback channel. A generator receives the fused CSI and generates predicted CSI for another frequency band. A discriminator then compares the actual CSI with the generated predicted CSI to optimize the generator's performance and thereby improve prediction accuracy. Experimental results demonstrate that the present invention can effectively improve the accuracy of channel state information acquisition in FDD systems, enabling the fusion of SRS and downlink feedback, and the conversion of CSI between frequency bands. The present invention is suitable for wireless communications, 5G / 6G MIMO systems, and other applications requiring efficient CSI transmission.

[0059] Based on the traditional FDD downlink CSI acquisition method, the present invention combines the advantages of deep learning algorithms and proposes an innovative channel state information (CSI) fusion method based on conditional generative adversarial networks (CGAN), which can efficiently integrate and model uplink (UL) and downlink (DL) information, and realize accurate modeling of complex channel distribution. By combining uplink SRS and downlink PMI feedback information, the mapping error problem caused by the uplink and downlink frequency difference in the traditional method is overcome, and the performance degradation of the feedback method caused by quantization accuracy error in weak channel environment is avoided. In terms of performance, the present invention significantly reduces the normalized mean square error (NMSE), improves the cosine similarity, and shows higher channel information reconstruction accuracy and adaptability, especially in complex channels and large-scale MIMO scenarios. The present invention has wide applicability and robustness, provides a new solution for improving the efficiency of UL and DL information interaction, and is of great value both in theory and practice. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 This is an illustration of channel fusion in the method of the present invention.

[0061] Figure 2 This is a flowchart of the neural network prediction method of the present invention.

[0062] Figure 3 、 4 This is a structural diagram of the conditional generative adversarial network generator and discriminator in the method of the present invention.

[0063] Figure 5 、 6 This is a simulation comparison diagram of the present invention and existing designs in the method of the present invention. DETAILED DESCRIPTION

[0064] The present invention is further described below through specific implementation examples.

[0065] As an example, this experiment simulates a wireless channel scenario based on a multiple-input multiple-output communication system. The base station side uses a 2×8 dual-polarization uniform planar array antenna with a total of 32 antenna units and an antenna spacing of 0.5 wavelengths. It is fixedly installed at a height of 10 meters in the center of the scene. The user equipment is equipped with 4 omnidirectional antennas. 80% of the users are located indoors, randomly distributed in multiple buildings with room sizes of 10 meters × 10 meters and a height of 3 meters. 20% of the users are located in the outdoor area within a radius of 500 meters from the base station. The channel model is based on the 3GPP TR 38.901 standard. The outdoor users use the UMa scenario, taking into account both line of sight (LoS) and non-line of sight (NLoS) conditions. The user equipment moves randomly in the scene at a speed of 1 meter / second, with a path length of 10 meters. Channel multipath characteristics, including randomly generated angles of arrival (AoA) and departure (AoD), delay spread, Doppler shift, and path loss are modeled to comprehensively evaluate the performance of MIMO communication channels in dynamic environments. Channel data is generated using the QuaDRiGa channel simulation generator.

[0066] To simulate the feedback acquisition process, a PMI feedback mechanism based on the 3GPP Release 16 protocol was introduced. During this process, the user calculates the optimal downlink beamforming matrix based on uplink channel state information and feeds the associated PMI back to the base station. The base station combines the PMI feedback from the user with the uplink channel information to generate the downlink channel estimate covariance, which is used to enrich the input features of the generator.

[0067] The uplink channel covariance generated by QuaDRiGa and the downlink feedback channel covariance generated based on R16 quantization are extracted as complex channel matrices respectively. The generated image size is 32×32×3, which is fused using the method in step 2 and used as the input and target output data sets of the generator.

[0068] During training, a combined objective function based on adversarial loss and pixel-level L1 loss is used to optimize the performance of the generator and discriminator. The adversarial loss aims to ensure that the downlink channel images generated by the generator are as close as possible to the real data, while the L1 loss further improves the accuracy of the generated results by penalizing pixel-level errors. During training, the Adam optimizer is used with parameters set to β1 = 0.5 and β2 = 0.999, an initial learning rate of 0.0002, and a batch size of 64. The generator and discriminator are updated every three epochs to maintain a dynamic balance between the generator and the discriminator.

[0069] The experiment verifies the model by generating multiple sets of channel data of different sizes and evaluates its generalization ability and prediction accuracy in different scenarios. Figure 5and Figure 6 Figure 2 shows the performance of the proposed CF-CGAN framework on the test set, including results under the NMSE and cosine similarity metrics. Under different channel conditions, CF-CGAN significantly reduces NMSE and improves cosine similarity, fully demonstrating its ability to effectively integrate and model uplink and downlink information, and has significant advantages over the feedback method and SRS-Hilbert method. Specifically, due to the failure to consider the non-negativity of the angular power spectrum, SRS-Hilbert may cause mapping errors when reconstructing the downlink correlation channel matrix; while the feedback method is prone to performance degradation in weak channel environments due to quantization accuracy errors caused by a large number of antennas.

[0070] Experimental results show that the CSI fusion in the CGAN model designed in this paper overcomes the inherent limitations of traditional methods through an effective mapping method, exhibits stronger robustness and adaptability in complex channel environments, and provides an efficient solution to the downlink channel estimation problem in frequency division duplex (FDD) systems.

Claims

1. A channel state information acquisition method based on conditional generative adversarial network, characterized in that: Based on CSI fusion and conditional generative adversarial networks, it serves large-scale MIMO scenarios. GAN establishes a nonlinear mapping relationship between uplink and downlink channels in FDD systems through mutual adversarial learning between the generator and the discriminator. It uses uplink channel information to generate corresponding downlink channel state information, improving the accuracy of CSI acquisition without increasing feedback bandwidth, thereby enhancing the beamforming and resource allocation efficiency of FDD large-scale MIMO systems. The specific steps are as follows: Step 1: Set environmental parameters such as the scene, motion trajectory, and antenna layout, select a channel model, and configure key parameters; generate multipath channel data through experiments or simulations; and generate a channel covariance matrix based on the channel data; Step 2: Map the data in the channel matrix to the red, green, and blue channels of the RGB image, and fuse the uplink SRS channel information and the downlink channel information to generate the fused CSI data; including: ① Mapping: In order to realize the conditional generative adversarial network, first, from the uplink SRS channel covariance matrix and the channel covariance matrix of the downlink PMI quantization feedback Generate a pseudo-color image; this process involves separating the real and imaginary parts of the matrix and mapping them to the RGB channels of the image. The image dimensions are N×N×3, where N is the number of antennas on the base station side. The real values ​​correspond to the red channel, the imaginary values ​​correspond to the green channel, and the blue channel is filled with a constant value. ② Normalization: To facilitate convergence of subsequent neural network training, all samples are normalized to the range [-1, 1]. Specifically, the channel matrix is ​​uniformly divided by a constant to scale it. ③ Fusion: The images generated by the uplink channel covariance matrix and the downlink feedback covariance matrix based on SRS are fused in the third dimension to obtain the fused samples Step 3: The generator model in the conditional generative adversarial network receives the fused CSI data and generates the predicted CSI for another frequency band. Step 4: Use the discriminator model in the conditional generative adversarial network to compare the generated predicted CSI with the true CSI to optimize the discriminator performance. Based on the feedback from the discriminator, the generator is iteratively optimized to improve the accuracy of inter-band CSI conversion. Step 5: Performance evaluation, specifically using normalized mean square error and cosine similarity to evaluate the model-generated images to comprehensively measure its pixel-level accuracy and subspace alignment quality.

2. The channel state information acquisition method according to claim 1, wherein: The specific process of step 1 is: Consider establishing a MIMO wireless channel between a base station equipped with a ±45° dual-polarization uniform planar array antenna and a user equipped with a single-polarization UPA. Assume that the channel between the user and the base station consists of L paths, each path corresponds to a direction, and the channel matrix is for: Among them, N BS and N UE are the number of antennas on the base station side and the user side respectively; L is the number of paths; α l is the complex path gain of path l, including path loss and phase; is the direction vector of the UPA array at the base station side; is the direction vector of the user-side UPA array, θ is the pitch angle, is the azimuth angle, which is used to describe the propagation direction of the signal in the vertical and horizontal planes respectively; the specific steps for constructing the array direction vector at the transmitting or receiving end are as follows; For ±45° dual-polarization UPA, there is N in the vertical direction. V antennas, there are N antennas in the horizontal direction H The antenna spacing in the vertical and horizontal directions is d V and d H , whose array response vector is defined as represents the amplitude component, specifically, Represent the vertical and horizontal field patterns of the +45° polarized antenna, respectively. They represent the FP in the vertical and horizontal directions of the -45° polarized antenna respectively; Represents the phase component and satisfies: Where 1≤i≤N V ,1≤j≤N H , f∈{UL,DL}, where UL and DL represent the carrier frequencies of the uplink and downlink, respectively, v and d H The antenna spacing in the vertical and horizontal directions, N V and N H are the number of antennas in the horizontal and vertical directions, respectively, and c represents the speed of light; The channel matrix H is modeled as a complex Gaussian matrix, and its covariance matrix R H Expressed as: ， (5) R is approximated by averaging multiple channel realizations H .

3. The channel state information acquisition method according to claim 2, wherein: The fusion processing of the uplink SRS channel information and the downlink channel information in step 2 to generate fused CSI data includes: ① The base station sends the CSI-RS reference signal s CSI-RS , the received signal model on the user side is: y UE =H d Vs CSI-RS +w, (6) in, represents the receiving vector, represents the downlink channel matrix, represents the outer weight matrix, is the noise at the receiving end; ② The user side estimates the downlink equivalent channel H e , the corresponding equivalent channel matrix is: ③And calculate the corresponding covariance matrix R e ; Where T and J represent the number of frequency domain subcarriers and OFDM symbols, respectively, and P is the number of ports; ④ The user side quantizes the downlink CSI based on the protocol to obtain the quantization codeword And feed back to the base station; ⑤ Then obtain the PMI downlink feedback covariance information based on the feedback vector 4. The channel state information acquisition method according to claim 3, characterized in that: In step 3, the generator model in the conditional generative adversarial network receives the fused CSI data and generates the predicted CSI for another frequency band. The specific process is as follows: ⑴Build a conditional generative adversarial network The conditional generative adversarial network (CGN) is a generative model that introduces conditional constraints to guide the generation of specific attributes or features of samples. The generator generates samples based on random noise and conditional variables, while the discriminator determines the authenticity of the generated samples and whether they meet the conditions. Through adversarial training, the generated samples are not only realistic but also consistent with the input conditions. The specific process is as follows: ① Build the generator: The generator uses a U-Net network structure, specifically a symmetric encoder-decoder architecture, and shares feature information between encoding and decoding through skip connections. The encoder is responsible for extracting progressively abstract features, while the decoder restores the image resolution through upsampling, while combining the high-resolution features of the encoding stage to preserve details. The final output layer generates the target image through convolution operations. In the conditional generative adversarial network, using U-Net as the generator can efficiently extract and restore features, generating a more accurate downlink channel covariance image. ②Build the discriminator: The discriminator uses a convolutional neural network structure to distinguish generated images from real images. It consists of multiple convolutional layers and activation functions to gradually extract high-dimensional features of the image. Through downsampling operations, the discriminator captures local spatial information and ultimately outputs a scalar value through a fully connected layer, indicating the probability that the input image is real. ⑵The conditional variable is the fused sample covariance matrix R fusion Together with the random noise vector z, the conditional variable is input to the generator, and is used to guide the generator to generate samples with specific characteristics; The generator uses the input random noise and conditional variables to generate pseudo samples G(z|R fusion ), the purpose of generating samples is to be as close as possible to the real data distribution; The discriminator receives the generated sample G(z|R fusion ) and the real sample R DL , combined with the conditional variable R fusion , determine whether the input comes from real data or the generator; The discriminator outputs a probability value D(G(z|R fusion )|R fusion ) or D(R DL |R fusion ), indicating that the network judges the input samples (z|R fusion ) and R DL Under the given condition R fusion The probability value of whether it is a true sample.

5. The channel state information acquisition method according to claim 4, characterized in that: In step 4, the discriminator model in the conditional generative adversarial network is used to compare the generated predicted CSI with the actual CSI to optimize the discriminator performance. The generator is iteratively optimized based on the feedback from the discriminator to improve the accuracy of inter-band CSI conversion. The specific process is as follows: Using the adversarial loss function, the generator and discriminator are optimized alternately through back propagation, that is, the objective function is: The discriminator D estimates the probability of whether the data is a real downlink channel based on the input; the last term in the objective function represents the pixel-by-pixel difference between the generated image and the corresponding "real" image, and imposes a penalty term on the generator based on the size of this difference; The goal of the generator is to minimize the objective function, while the discriminator strives to maximize the objective function, forming the following minimum maximum optimization framework: G * =arg min G max D L CGAN (G,D), (10) Through this iterative optimization method, the generator and discriminator gradually improve their performance in the confrontation, and the sample distribution finally generated is close to the real data distribution.

6. The channel state information acquisition method according to claim 5, characterized in that: The performance evaluation described in step 5 uses the normalized mean square error and the cosine of the one-dimensional subspace angle of the matrix to measure the similarity between the downlink channel covariance matrix generated by the generator and the true downlink channel covariance matrix.

Images