CN113556158B - Large-scale MIMO intelligent CSI feedback method for Internet of vehicles - Google Patents
Large-scale MIMO intelligent CSI feedback method for Internet of vehicles Download PDFInfo
- Publication number
- CN113556158B CN113556158B CN202110823053.3A CN202110823053A CN113556158B CN 113556158 B CN113556158 B CN 113556158B CN 202110823053 A CN202110823053 A CN 202110823053A CN 113556158 B CN113556158 B CN 113556158B
- Authority
- CN
- China
- Prior art keywords
- matrix
- csi
- data
- internet
- vehicles
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 32
- 239000011159 matrix material Substances 0.000 claims abstract description 60
- 230000006870 function Effects 0.000 claims abstract description 8
- 230000000694 effects Effects 0.000 claims abstract description 7
- 238000011176 pooling Methods 0.000 claims abstract description 7
- 238000012549 training Methods 0.000 claims description 17
- 230000008569 process Effects 0.000 claims description 11
- 238000004891 communication Methods 0.000 claims description 7
- 230000004913 activation Effects 0.000 claims description 6
- 238000013461 design Methods 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 6
- 238000010586 diagram Methods 0.000 claims description 5
- 238000010606 normalization Methods 0.000 claims description 4
- 239000000654 additive Substances 0.000 claims description 3
- 230000000996 additive effect Effects 0.000 claims description 3
- 239000000969 carrier Substances 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 229940088594 vitamin Drugs 0.000 claims description 3
- 229930003231 vitamin Natural products 0.000 claims description 3
- 235000013343 vitamin Nutrition 0.000 claims description 3
- 239000011782 vitamin Substances 0.000 claims description 3
- 150000003722 vitamin derivatives Chemical class 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 2
- 238000010200 validation analysis Methods 0.000 claims description 2
- 230000006855 networking Effects 0.000 claims 1
- 238000013135 deep learning Methods 0.000 description 4
- 238000011084 recovery Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 238000005562 fading Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
- H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
- H04B7/0413—MIMO systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
- H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
- H04B7/0413—MIMO systems
- H04B7/0417—Feedback systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
- H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
- H04B7/0413—MIMO systems
- H04B7/0456—Selection of precoding matrices or codebooks, e.g. using matrices antenna weighting
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Radio Transmission System (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention provides a large-scale MIMO intelligent CSI feedback method for the Internet of vehicles. Aiming at the problems of high complexity, low feedback precision and high feedback cost of a Channel State Information (CSI) feedback method of the Internet of vehicles, the invention constructs a network structure combining an encoder of a User Equipment (UE) with a decoder of a Base Station (BS) in an end-to-end mode. And gradually shrinking the original CSI matrix by adopting blocks formed by a continuous average pooling layer, a common convolution layer (CNN) and a depth separable convolution layer (DSCNN) at the UE end to obtain a compressed codeword matrix. At the BS end, blocks formed by continuous upsampling layers, CNN and DSCNN are first used to gradually expand the compressed codeword matrix, so as to obtain an initial CSI reconstruction. Then, a residual convolution block is constructed by utilizing the function of gradual fine tuning of a residual learning network (ResNet) to gradually approximate the initially reconstructed CSI to the original CSI so as to achieve a better reconstruction effect.
Description
Technical field:
the invention relates to a large-scale MIMO intelligent CSI (Channel State Information) feedback method for the Internet of vehicles, in particular to a CSI feedback method based on light weight and low complexity.
The background technology is as follows:
in the internet of vehicles environment, the variation of the wireless propagation channel is complicated due to the high-speed movement of the vehicle. Research on the characteristics of a high-speed wireless channel is the basis of research on the communication technology of a high-speed environment, and the characteristics of the wireless channel generally mainly comprise time-frequency domain channel response and time-frequency domain channel correlation coefficients. On the one hand, in a high-speed mobile environment, the wireless channel at the moment can simultaneously show time domain and frequency domain selective fading characteristics due to the influence of multipath and Doppler. On the other hand, since the wireless channel propagated during the high-speed movement of the vehicle has the characteristics of fast fading, severe doppler effect, high data transmission rate, complex communication environment and the like, the channel at this time is not a generalized stationary random process, i.e. the wireless channel in the high-speed movement environment is a typical non-stationary channel. Due to these characteristics of wireless channels in high-speed mobile environments, there are also challenges in achieving reliable, stable, and fast mobile communications.
In order to improve the system performance, the downlink CSI needs to be accurately acquired at the transmitting end, and if all CSI is fed back, a serious burden is caused to the feedback link. Therefore, in order to reduce excessive feedback overhead, the receiving end only needs to feed back part of CSI to the transmitting end. In recent years, a deep learning method is gradually applied to CSI feedback, and many researchers have proposed some CSI feedback frameworks based on deep learning. The deep learning based real-time channel recovery scheme utilizes deep neural networks in training and prediction to reduce feedback overhead. Existing schemes such as the deep learning based channel recovery framework csanet use an encoder to convert the channel matrix into codewords at the user side and a decoder to reconstruct the CSI from the codewords at the base station side. Because the code word is smaller and can not identify the information of the channel matrix, the overfitting phenomenon is serious and the recovery effect is general. There is also an improved network architecture csanetplus that employs a larger convolution kernel and more refianenet blocks to improve the network's reconstruction of CSI. Although the CSI feedback frames can obtain better reconstruction effects than the methods based on compressed sensing and codebooks, the complex deployment of the CSI feedback frames at the user end and the high requirement on the computational power of the devices are still issues to be solved.
The invention comprises the following steps:
aiming at the problem of high complexity feedback overhead, the invention provides a large-scale MIMO intelligent CSI feedback method oriented to the Internet of vehicles, which is characterized by comprising the following steps:
s1, constructing a communication system model based on large-scale MIMO for the Internet of vehicles to obtain batch CSI truncated matrix data;
s2, designing a structure of an encoder in a lightweight CSI feedback framework;
s3, designing a structure of a decoder in the lightweight CSI feedback framework;
s4, the encoder and the decoder which are provided by the S2 and the S3 are used as the same end-to-end network for training to obtain a network model in a training scheme with reasonable design.
The intelligent CSI feedback method based on the large-scale MIMO system for the Internet of vehicles is characterized in that S1 comprises the following steps:
in a system based on massive MIMO facing the Internet of vehicles, consider a base station with N t (N t > 1) transmitting antennas, and the user end has a single receiving antenna. The system employs orthogonal frequency division multiplexing (Orthogonal Frequency DivisionMultiplexing, OFDM) and has N s Sub-carriers. The resulting receiver signal y can be described as:
wherein ,
represented as N s Vitamin reception vector->
Represented as N s The vector of the transmitted light is maintained,
represented as N s ×N t A channel matrix of dimensions +.>
Expressed as a channel vector on the ith subcarrier, < >>
wherein ui Representing the precoding vector of the i-th subcarrier,
represented as N s Additive white gaussian noise in the dimension.
To better design the precoding vector u i In the CSI feedback process, it is required thatObtaining a sufficiently accurate signal at the base station
I.e. a higher reconstruction effect is obtained with a lower feedback amount. Whereas the CSI matrix is sparse in the angular domain, the virtual angular domain matrix can be found by two discrete fourier transform (Discrete Fourier Transform, DFT) matrices:
wherein ,Ds Is N s ×N s DFT matrix of (D) t Is N t ×N t Is a DFT matrix of (c). Will be original N s ×N t The matrix is truncated to N' s ×N t (N′ s ＜N s ) Is a truncated matrix of (i) CSI truncated matrix
The intelligent CSI feedback method based on the large-scale MIMO system for the Internet of vehicles is characterized in that the S2 comprises the following steps:
for CSI truncation matrix
The matrix belongs to complex matrix, wherein the matrix contains real and imaginary data, and the real and imaginary data are converted into N 'for facilitating the processing and training of the data' s ×N t X 2. The transformed matrix is used as input of the encoder, and a ConvBN block (a characteristic diagram is output by a two-dimensional common convolution layer Conv2D with a 3×3 convolution kernel, a batch normalization layer (BatchNormalization, BN) and an activation function (LeakyReLU)). Then, the compressed codeword matrix S is obtained by four consecutive SACN blocks en . Each SACN block consists of an average pooling layer (AveragePooling 2D) with a pooling window size of 2 x 2 and a SEConvBN block. While each SEConvBN block is formed by a depth separable convolution (SeparableConv 2D), BN, andLeakyReLU.
The intelligent CSI feedback method based on the large-scale MIMO system for the Internet of vehicles is characterized in that the S3 comprises the following steps:
the resulting codeword matrix S is output at the encoder en It is then input to the decoder, first through 4 consecutive USCN blocks, each constituted by a two-dimensional UpSampling layer (UpSampling 2D) with a data interpolation window size of 2 x 2 and a ConvBN block. The up-sampling layer can make the input data dimension improved, and the principle is that the dimension improvement is completed through repeated interpolation process of the rows and columns of the data. Next, N 'is obtained by a ConvBN layer' s ×N t The x 2 output is then input to two continuous ConvBlock blocks, each of which is a residual convolution block constructed from a residual network format and is composed of three ConvBN blocks. The sum of the output after passing through one ConvBN layer and the input of the whole ConvBlock block is taken as the final output of the ConvBlock block. Finally N' s ×N t The x 2 output is passed through a common convolutional layer Conv2D of a 3 x 3 convolutional kernel (the activation function Sigmoid, the data can be adjusted to 0,1]) Obtaining the final reconstruction result
In summary, the beneficial effects of the invention are as follows:
the intelligent CSI feedback method based on the large-scale MIMO system for the Internet of vehicles, which is provided by the invention, has the advantages that the reconstruction quality of the channel matrix is good, the network complexity is reduced, and good performance is obtained. For the equipment with low computational power and low storage at the current stage, the CSI feedback method provided by the invention has more excellent deployment and performance.
Drawings
The foregoing and/or additional aspects and advantages of the invention will become apparent and may be better understood from the following description of embodiments taken in conjunction with the accompanying drawings in which:
FIG. 1 is a general flow chart of the present invention;
FIG. 2 is a diagram of a user side code structure;
fig. 3 is a structural design diagram of a base station decoder.
Detailed Description
Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are illustrative only and are not to be construed as limiting the invention.
In the present disclosure, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The invention provides a large-scale MIMO intelligent CSI feedback method for the Internet of vehicles, which has better performance in complexity, reconstruction quality and running time.
The invention is described in detail with reference to fig. 1, and mainly comprises the following steps:
step 1: starting.
Step 2: and constructing a communication system model based on large-scale MIMO and oriented to the Internet of vehicles, and establishing a CSI cut-off matrix.
In a system based on massive MIMO facing the Internet of vehicles, consider a base station with N t (N t > 1) transmitting antennas, and the user end has a single receiving antenna. The system employs orthogonal frequency division multiplexing (Orthogonal Frequency Division Multiplexing, OFDM) and has N s Sub-carriers. The resulting receiver signal y can be described as:
wherein ,
represented as N s Vitamin reception vector->
Represented as N s The vector of the transmitted light is maintained,
represented as N s ×N t A channel matrix of dimensions +.>
Expressed as a channel vector on the ith subcarrier, < >>
wherein ui Representing the precoding vector of the i-th subcarrier,
represented as N s Additive white gaussian noise in the dimension.
To better design the precoding vector u i In the CSI feedback process, a sufficiently accurate signal is obtained at the base station
I.e. a higher reconstruction effect is obtained with a lower feedback amount. Whereas the CSI matrix is sparse in the angular domain, the virtual angular domain matrix can be found by two discrete fourier transform (Discrete Fourier Transform, DFT) matrices:
wherein ,Ds Is N s ×N s DFT matrix of (D) t Is N t ×N t Is a DFT matrix of (c). Will be original N s ×N t The matrix is truncated to N' s ×N t (N′ s ＜N s ) Is a truncated matrix of (i) CSI truncated matrix
For CSI truncation matrix
The matrix belongs to complex matrix, wherein the matrix contains real and imaginary data, and the real and imaginary data are converted into N 'for facilitating the processing and training of the data' s ×N t X 2. The transformed matrix is used as an input to the encoder.
Step 3: the CSI truncated matrix encoder is designed, and the CSI truncated matrix is input, and the implementation process is as follows:
the feature map number is first output through a ConvBN layer (64 by a two-dimensional normal convolution layer Conv2D of 3 x 3 convolution kernels, a batch normalization layer (Batch Normalization) and an activation function (LeakyReLU). Then, the compressed codeword matrix S is obtained by four consecutive SACN blocks en The transformation of the number of feature maps into
Each SACN block consists of an average pooling layer (AveragePooling 2D) with a pooling window size of 2 x 2 and a SEConvBN block (unlike the ConvBN block, where the normal two-dimensional convolution layer (Conv 2D) is replaced with a depth separable convolution layer (seprobleconv 2D) of a 3 x 3 convolution kernel, where the seprobleconv 2D has a significantly reduced number of network parameters compared to the normal convolution layer, assuming that the matrix size of the input two convolution layers is (w, l, c), the convolution kernel size (k, k) and no offset term is calculated, the final number of output channels is m for the normal convolution block, there are how many input channels, the parameter amount of the normal convolution block is: param=m×k×k×c, and the depth separable convolution block is the channel convolution first (i.e. we can get the channel convolutionThe number of feature maps equal to the number of input channels) then, at 1×1 point-by-point convolution is performed for each output feature map, the parameter number is calculated to obtain param=k×k×c+1×1×c×m= (k×k+m) ×c.
Step 4: the codeword matrix decoder is designed and codeword matrices are input, and the implementation process is as follows:
outputting the encoder to obtain a codeword matrix S en The input decoder is first 4 consecutive USCN blocks, each constituted by an Up Sampling layer (Up Sampling 2D) with a data interpolation window size of 2 x 2 and a ConvBN block, each constituted by an Up Sampling2D layer and a SEConvBN block. The up-sampling layer can enable the dimension of the input data to be improved, and the dimension improvement is completed through repeated interpolation of the rows and columns of the data. The number of the characteristic diagrams is as follows:
next, N 'is obtained by a ConvBN layer' s ×N t The x 2 output is then input to two consecutive ConvBlock blocks, each of which is a residual convolution block constructed from the form of a residual network. Each ConvBlock block contains three ConvBN layers with feature map numbers of 8, 16,2, respectively. Finally N' s ×N t The x 2 output is passed through a common convolutional layer Conv2D of a 3 x 3 convolutional kernel (the activation function Sigmoid, the data can be adjusted to 0,1]) Obtaining the final reconstruction result->
Step 5: the training scheme is designed, a network model is trained through a large amount of data, and the specific implementation process is as follows:
the training configuration of the communication system parameters and the network based on the massive MIMO facing the Internet of vehicles is as follows: the base station antennas are arranged in a uniform linear array antenna arrangement (Uniform Linear Array, ULA) with an antenna spacing of half a wavelength. The number of the base station end antennas is N t =32, ofdm system subcarrier number N s =1024. Considering the sparse characteristic of massive MIMO we only get the frontN′ s =32 rows. I.e. H is a 32 x 32 complex matrix. The training, validation and test data set sizes are respectively: 100000, 30000 and 20000. The training batch data size (batch size) was 100, training round (epochs) was 1500 rounds, and the net learning rate (learning rate) was 0.001.
Step 6: and (5) ending.
In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
While embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that: many changes, modifications, substitutions and variations may be made to the embodiments without departing from the spirit and principles of the invention, the scope of which is defined by the claims and their equivalents.
Claims (2)
1. An intelligent CSI feedback method based on a large-scale MIMO system and oriented to the Internet of vehicles is characterized by comprising the following steps:
s1, constructing a communication system model based on large-scale MIMO for the Internet of vehicles to obtain batch CSI truncated matrix data; comprising the following steps:
in a system based on massive MIMO facing the Internet of vehicles, consider a base station with N t (N t > 1) transmitting antennas, and the user end is provided with a single receiving antenna; the system employs orthogonal frequency division multiplexing (Orthogonal Frequency Division Multiplexing, OFDM) and has N s Sub-carriers; the resulting receiver signal y can be described as:
wherein ,
represented as N s Vitamin reception vector->
Represented as N s The vector of the transmitted light is maintained,
represented as N s ×N t A channel matrix of dimensions +.>
Expressed as a channel vector on the ith subcarrier, < >>
wherein ui Representing the precoding vector of the i-th subcarrier,
represented as N s Additive white gaussian noise in dimension;
to better design the precoding vector u i In the CSI feedback process, a sufficiently accurate signal is obtained at the base station
Namely, a higher reconstruction effect is obtained through a lower feedback quantity; whereas the CSI matrix is sparse in the angular domain, the virtual angular domain matrix can be found by two discrete fourier transform (Discrete Fourier Transform, DFT) matrices:
wherein ,Ds Is N s ×N s DFT matrix of (D) t Is N t ×N t Is a DFT matrix of (2); will be original N s ×N t The matrix is truncated to N' s ×N t (N′ s ＜N s ) Is a truncated matrix of (i) CSI truncated matrix
S2, designing a structure of an encoder in a lightweight CSI feedback framework; comprising the following steps:
for CSI truncation matrix
The matrix belongs to complex matrix, wherein the matrix contains real and imaginary data, and the real and imaginary data are converted into N 'for facilitating the processing and training of the data' s ×N t X 2; the converted matrix is used as the input of an encoder, and firstly passes through a ConvBN block, wherein the ConvBN block consists of a two-dimensional common convolution layer Conv2D with a 3 multiplied by 3 convolution kernel, a batch normalization layer (BatchNormalization, BN) and an activation function (LeakyReLU) output characteristic diagram; then, the compressed codeword matrix S is obtained by four consecutive ACN blocks en The method comprises the steps of carrying out a first treatment on the surface of the Four consecutive ACN blocks constitute a SACN; each SACN block consists of an average pooling layer (AveragePooling 2D) with a pooling window size of 2 x 2 and a SEConvBN block; while each SEConvBN block consists of a depth separable convolution (sepabalecon 2D), BN and LeakyReLU;
s3, designing a structure of a decoder in the lightweight CSI feedback framework; comprising the following steps:
the resulting codeword matrix S is output at the encoder en Then input it into the decoder, first pass 4 consecutive USCN blocks, each USCN block is formed by a two-dimensional up-sampling layer (UpSampling 2D) with the size of 2×2 of data interpolation window and a ConvBN block; the up-sampling layer can enable the dimension of the input data to be improved, and the principle is that the dimension is improved through repeated interpolation of the rows and the columns of the data; next, N 'is obtained by a ConvBN layer' s ×N t The x 2 output is then input to two continuous ConvBlock blocks, each of which is a residual constructed according to the form of a residual networkThe convolution block is composed of three ConvBN blocks; taking the sum of the output after passing through one ConvBN layer and the input of the whole ConvBlock as the final output of the ConvBlock; finally N' s ×N t The output of x 2 is passed through a common convolution layer Conv2D of 3 x 3 convolution kernel to obtain the final reconstruction result
The Conv2D is an activation function Sigmoid, and can adjust data to [0,1 ]]；
And S4, training the encoder and the decoder which are proposed by the S2 and the S3 as the same end-to-end network by using a training scheme with reasonable design to obtain a network model.
2. The intelligent CSI feedback method based on a massive MIMO system for the internet of vehicles according to claim 1, wherein S4 comprises:
the training configuration of the parameters and the network of the MIMO communication system of the vehicle networking in the S1 is as follows: the base station antennas adopt a uniform linear array antenna arrangement mode (Uniform Linear Array, ULA), and the antenna spacing is half wavelength; the number of the base station end antennas is N t =32, ofdm system subcarrier number N s =1024; considering the sparse characteristic of massive MIMO we only take the first N' s =32 rows; i.e., H is a complex matrix of 32×32; the training, validation and test data set sizes are respectively: 100000, 30000 and 20000; the training batch data size (batch size) was 100, training round (epochs) was 1500 rounds, and the net learning rate (learning rate) was 0.001.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110823053.3A CN113556158B (en) | 2021-07-21 | 2021-07-21 | Large-scale MIMO intelligent CSI feedback method for Internet of vehicles |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110823053.3A CN113556158B (en) | 2021-07-21 | 2021-07-21 | Large-scale MIMO intelligent CSI feedback method for Internet of vehicles |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113556158A CN113556158A (en) | 2021-10-26 |
| CN113556158B true CN113556158B (en) | 2023-06-20 |
Family
ID=78103664
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110823053.3A Active CN113556158B (en) | 2021-07-21 | 2021-07-21 | Large-scale MIMO intelligent CSI feedback method for Internet of vehicles |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113556158B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114499770B (en) * | 2022-01-20 | 2023-07-14 | 西安交通大学 | Multi-user downlink CSI feedback method based on deep learning |
| CN114567359A (en) * | 2022-03-02 | 2022-05-31 | 重庆邮电大学 | CSI feedback method based on multi-resolution fusion convolution feedback network in large-scale MIMO system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108390706A (en) * | 2018-01-30 | 2018-08-10 | 东南大学 | A kind of extensive mimo channel state information feedback method based on deep learning |
| CN109672464A (en) * | 2018-12-13 | 2019-04-23 | 西安电子科技大学 | Extensive mimo channel state information feedback method based on FCFNN |
| CN110912598A (en) * | 2019-11-22 | 2020-03-24 | 中原工学院 | Large-scale MIMO system CSI feedback method based on long-time attention mechanism |
| CN112737985A (en) * | 2020-12-25 | 2021-04-30 | 东南大学 | Large-scale MIMO channel joint estimation and feedback method based on deep learning |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10992331B2 (en) * | 2019-05-15 | 2021-04-27 | Huawei Technologies Co., Ltd. | Systems and methods for signaling for AI use by mobile stations in wireless networks |
-
2021
- 2021-07-21 CN CN202110823053.3A patent/CN113556158B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108390706A (en) * | 2018-01-30 | 2018-08-10 | 东南大学 | A kind of extensive mimo channel state information feedback method based on deep learning |
| CN109672464A (en) * | 2018-12-13 | 2019-04-23 | 西安电子科技大学 | Extensive mimo channel state information feedback method based on FCFNN |
| CN110912598A (en) * | 2019-11-22 | 2020-03-24 | 中原工学院 | Large-scale MIMO system CSI feedback method based on long-time attention mechanism |
| CN112737985A (en) * | 2020-12-25 | 2021-04-30 | 东南大学 | Large-scale MIMO channel joint estimation and feedback method based on deep learning |
Non-Patent Citations (2)
| Title |
|---|
| A Lightweight Deep Network for Efficient CSI Feedback in Massive MIMO Systems;Yuyao Sun;《 IEEE Wireless Communications Letters》;全文 * |
| 基于深度学习的大规模MIMO系统信道状态信息反馈研究;刘振宇;《中国博士学位论文全文数据库（信息科技辑）》;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113556158A (en) | 2021-10-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112737985B (en) | Large-scale MIMO channel joint estimation and feedback method based on deep learning | |
| CN108390706B (en) | Large-scale MIMO channel state information feedback method based on deep learning | |
| CN110350958B (en) | CSI multi-time rate compression feedback method of large-scale MIMO based on neural network | |
| CN108847876B (en) | Large-scale MIMO time-varying channel state information compression feedback and reconstruction method | |
| CN111630787B (en) | MIMO multi-antenna signal transmission and detection technology based on deep learning | |
| Chen et al. | Deep learning-based implicit CSI feedback in massive MIMO | |
| Lu et al. | Bit-level optimized neural network for multi-antenna channel quantization | |
| CN113556158B (en) | Large-scale MIMO intelligent CSI feedback method for Internet of vehicles | |
| CN111464220A (en) | Channel state information reconstruction method based on deep learning | |
| CN108599820B (en) | Large-scale MIMO system channel estimation method based on block structure adaptive compression sampling matching tracking algorithm | |
| CN109560841A (en) | Extensive mimo system channel estimation methods based on improved distributed compression perception algorithm | |
| CN111713035B (en) | MIMO multi-antenna signal transmission and detection technology based on artificial intelligence | |
| CN108964726A (en) | A kind of extensive MIMO uplink transmission channels estimation method of low complex degree | |
| Korpi et al. | DeepRx MIMO: Convolutional MIMO detection with learned multiplicative transformations | |
| CN108259397B (en) | Large-scale MIMO system channel estimation method based on adaptive regularization subspace tracking compressed sensing algorithm | |
| CN114884549A (en) | Large-scale MIMO channel state information feedback method based on deep learning | |
| CN110289898A (en) | A kind of channel feedback method based on the perception of 1 bit compression in extensive mimo system | |
| CN112600596B (en) | Millimeter wave system channel feedback method based on tensor parallel compression | |
| CN113381790B (en) | AI-based environment knowledge assisted wireless channel feedback method | |
| CN116248156A (en) | Deep learning-based large-scale MIMO channel state information feedback and reconstruction method | |
| CN114157331B (en) | Large-scale MIMO channel state information feedback method based on pseudo complex value input | |
| CN115021787A (en) | Channel state information feedback method based on complex convolutional neural network | |
| CN115865145A (en) | Large-scale MIMO channel state information feedback method based on Transformer | |
| CN104618293B (en) | A kind of optimization method of the unitary transformation matrix of smooth singular value decomposition | |
| CN116192209A (en) | Gradient uploading method for air computing federal learning under MIMO channel |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |