WO2023016166A1 - 码本传输方法、终端、基站和存储介质 - Google Patents

码本传输方法、终端、基站和存储介质 Download PDF

Info

Publication number
WO2023016166A1
WO2023016166A1 PCT/CN2022/104950 CN2022104950W WO2023016166A1 WO 2023016166 A1 WO2023016166 A1 WO 2023016166A1 CN 2022104950 W CN2022104950 W CN 2022104950W WO 2023016166 A1 WO2023016166 A1 WO 2023016166A1
Authority
WO
WIPO (PCT)
Prior art keywords
channel
codebook
attention mechanism
network
parameters
Prior art date
Application number
PCT/CN2022/104950
Other languages
English (en)
French (fr)
Inventor
李伦
高争光
吴昊
肖华华
刘磊
鲁照华
Original Assignee
中兴通讯股份有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2023016166A1 publication Critical patent/WO2023016166A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B17/00—Monitoring; Testing
    • H04B17/30—Monitoring; Testing of propagation channels
    • H04B17/373—Predicting channel quality or other radio frequency [RF] parameters
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B17/00—Monitoring; Testing
    • H04B17/30—Monitoring; Testing of propagation channels
    • H04B17/391—Modelling the propagation channel
    • 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
    • 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
    • H04B7/06—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L1/00—Arrangements for detecting or preventing errors in the information received
    • H04L1/02—Arrangements for detecting or preventing errors in the information received by diversity reception
    • H04L1/06—Arrangements for detecting or preventing errors in the information received by diversity reception using space diversity
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L1/00—Arrangements for detecting or preventing errors in the information received
    • H04L1/12—Arrangements for detecting or preventing errors in the information received by using return channel

Definitions

  • the embodiments of the present application relate to the field of wireless communication, and in particular, to a codebook transmission method, a terminal, a base station, and a storage medium.
  • CSI Channel State Information
  • the base station obtains CSI through feedback from the user equipment (UE) side, and the CSI received by the base station will be attenuated and reflected due to the influence of the transmission environment, causing the base station to The accuracy of CSI obtained from the side is low.
  • UE user equipment
  • the embodiment of the present application provides a codebook transmission method, the method includes the following steps: acquiring the channel parameters of the channel; performing feature extraction processing on the channel parameters according to the preset attention mechanism, and generating the first Codebook; perform encoding processing on the first codebook according to a preset encoding network to generate a second codebook for the channel; send the second codebook to the base station side, where the second codebook is used and causing the base station side to generate channel state information of the channel.
  • the embodiment of the present application also provides a codebook transmission method.
  • the method includes the following steps: receiving the second codebook of the channel sent by the terminal side, wherein the second codebook is determined by the terminal side according to the channel parameter generation; use a preset decoding network to decode the second codebook to generate a third codebook; use a preset attention mechanism to extract features from the third codebook; analyze the feature extraction
  • the third codebook acquires channel parameters of the channel; generates channel state information of the channel according to the channel parameters.
  • the embodiment of the present application also proposes a terminal, including: an acquisition module, used to acquire channel parameters of a channel; an attention mechanism module, used to perform feature extraction processing on the channel parameters according to a preset attention mechanism, and generate the The first codebook of the channel; the coding module is used to encode the first codebook according to the preset coding network to generate the second codebook of the channel; the transmission module is used to send the second codebook A codebook to the base station side, where the second codebook is used to enable the base station side to generate channel state information of the channel.
  • the embodiment of the present application also proposes a base station, including: a receiving module, configured to receive a second codebook of a channel transmitted by the terminal side, wherein the second codebook is generated by the terminal side according to channel parameters of the channel
  • the decoding module is used to decode the second codebook using a preset decoding network to generate a third codebook;
  • the attention mechanism module is used to perform decoding on the third codebook using a preset attention mechanism Feature extraction;
  • an analysis module configured to analyze the third codebook after feature extraction to obtain channel parameters of the channel, and generate channel state information of the channel according to the channel parameters.
  • Embodiments of the present application also provide a computer-readable storage medium storing a computer program, wherein the computer program implements the codebook transmission method described in any one of the above claims when executed by a processor.
  • FIG. 1 is a flowchart of a codebook transmission method provided in an embodiment of the present application
  • Figure 1a is a schematic structural diagram of the attention mechanism provided by the embodiment of the present application.
  • Figure 1b is a schematic structural diagram of the attention mechanism provided by the embodiment of the present application.
  • Figure 1c is a schematic structural diagram of the attention mechanism provided by the embodiment of the present application.
  • Figure 1d is a structural diagram of the spatial attention mechanism provided by the embodiment of the present application.
  • Figure 1e is a structural diagram of the channel attention mechanism provided by the embodiment of the present application.
  • FIG. 2 is a flowchart of a codebook transmission method provided in an embodiment of the present application
  • FIG. 3 is a flowchart of a codebook transmission method provided in an embodiment of the present application.
  • FIG. 4 is a flowchart of a codebook transmission method provided in an embodiment of the present application.
  • FIG. 5 is a flowchart of a codebook transmission method provided in an embodiment of the present application.
  • FIG. 6 is a flowchart of a codebook transmission method provided in an embodiment of the present application.
  • FIG. 7 is a schematic structural diagram of a terminal provided in an embodiment of the present application.
  • Fig. 7a is a schematic structural diagram of a terminal provided in an embodiment of the present application.
  • FIG. 7b is a schematic structural diagram of a terminal provided in an embodiment of the present application.
  • FIG. 8 is a schematic structural diagram of a base station provided in an embodiment of the present application.
  • FIG. 8a is a schematic structural diagram of a base station provided in an embodiment of the present application.
  • Fig. 8b is a schematic structural diagram of a base station provided in an embodiment of the present application.
  • the main purpose of the embodiments of this application is to propose a codebook transmission method, terminal, base station, and storage medium, aiming at improving the quality of the codebook information generated by the terminal side, so that the accuracy of the CSI obtained by the base station side according to the codebook information is accurate. degree increase.
  • the embodiment of the present application relates to a codebook transmission method, which is applied to the terminal side of the wireless communication system.
  • N t ports which may be physical antennas or logical antennas
  • N r ports on the terminal side.
  • Step 101 acquire channel parameters of the channel.
  • the channel parameters of the channel acquired by the terminal side include channel basis vectors, the number of channel basis vectors, the maximum number of ports, channel type information, codebook type information, codebook number information, bandwidth indication, frequency domain units, etc.
  • the channel type information may include channel scene, channel angle spread, channel delay spread and other information representing the channel type
  • the channel basis vector may be the basis vector used to transform the channel, and it may be used to represent the distance between the base station side and the terminal side
  • the channel base vector is a frequency domain base vector
  • the channel base vector is a time domain base vector
  • the channel base vector is N F *1 Discrete Fourier Transform (DFT for short) vector, or a vector composed of multiple DFT vectors
  • the number of channel basis vectors N is the number of channel basis vectors
  • the number of channel basis vectors is the frequency corresponding to the number of frequency domain units.
  • the maximum number of ports can be the number of communication antennas or the maximum number of communication ports on the base station side
  • the channel type indication can be information indicating the channel type (such as city macro, city micro, indoor, factory, office areas, shopping malls, squares, subways, etc.)
  • the bandwidth indication can be information indicating the communication bandwidth (such as the number of subbands, the number of physical resource blocks, and the size of the bandwidth)
  • the frequency domain unit can be the basic unit of the frequency domain.
  • the terminal side can also generate the channel matrix of the channel according to the channel parameters of the channel.
  • Step 102 perform feature extraction processing on channel parameters according to a preset attention mechanism, and generate a first codebook of the channel.
  • the channel parameters acquired by the terminal side are the description information of the channel matrix, and the terminal side first processes the input channel parameters to generate the channel matrix of the channel, or receives the downlink pilot channel to estimate the terminal's
  • H is an N ⁇ M complex matrix
  • the generated channel matrix contains spatial dimension information and channel dimension information
  • the spatial dimension information includes height dimension information and width dimension information (by N r , N t )
  • the channel dimension information is disassembled into two real numbers by the real part and imaginary part of the channel matrix
  • the wireless channel wireless frequency domain channel and wireless time domain channel the wireless frequency domain channel is usually composed of port space dimension information, frequency Domain granularity dimension information, time dimension information, channel dimension information and other dimensions
  • the wireless time domain channel is usually composed of port space dimension information, time domain granularity dimension information, channel dimension information and time dimension information and other dimensions
  • the channel dimension of the channel matrix The information can also be replaced with any one of port space dimension information, granularity dimension information or time dimension information, wherein, when the channel matrix contains time dimension information, the obtained channel matrix is a four-dimensional matrix, and the remaining dimension information is a three-dimensional matrix;
  • the terminal side will input the generated channel matrix into the attention
  • the attention mechanism used in this application can be a spatial attention mechanism (Spatial Attention, referred to as SA) or a channel attention mechanism (Channel Attention, referred to as CA) as shown in Figure 1a, as shown in Figure 1b
  • SA spatial Attention
  • CA channel attention mechanism
  • the attention mechanism of the channel first and then the space or the attention mechanism of the space first and then the channel as shown in Figure 1c, and the input AM includes the space dimension information HW and the channel dimension information C
  • the space dimension information includes the height dimension information H and the width Dimension information W, that is to say, the input to the attention mechanism must be at least one three-dimensional information containing three dimensions (the input time dimension information is four-dimensional information), and the channel dimension information can be the first dimension or the last dimension, namely It can be expressed as C ⁇ H ⁇ W or H ⁇ W ⁇ C.
  • SA is to redistribute the key feature weight ratio on the spatial dimension information of the input data, and perform the maximum pooling operation and the average pooling operation on the channel dimension information of the input channel parameters, and the maximum pooling operation is to obtain the channel dimension information
  • the maximum value of all values, the average pooling operation is to average all the values of the channel dimension information, so that two tensors with a size of H ⁇ W ⁇ 1 can be obtained, and the two are stacked together according to the channel dimension to form H ⁇ W ⁇ 2 tensor.
  • CA is to redistribute the weight ratio of key features on the channel dimension information of the input data, and perform the maximum pooling operation and the average pooling operation on the spatial dimension information of the input channel information, and the maximum pooling operation takes the spatial dimension information
  • the average pooling is to average all the values of the spatial dimension information, so that two tensors with a size of 1 ⁇ 1 ⁇ C can be obtained, and the two are sent to the fully connected network or convolution Neural network, it should be noted that the size of the output data after the fully connected network or convolutional neural network should be kept the same as the input data size.
  • Step 103 Perform encoding processing on the first codebook according to a preset encoding network to generate a second codebook for the channel.
  • the second codebook can be obtained. Among them, the first codebook The number of elements corresponding to the second codebook and the second codebook is different.
  • the number of elements in the second codebook is generally less than the number of elements corresponding to the first codebook.
  • the number of elements in the second codebook is the same as the number of elements in the first codebook.
  • the ratio of the number of elements is called the compression ratio.
  • the encoding network is used to compare the first codebook
  • the processing is actually to perform a compression operation on the first codebook, so that when the terminal side transmits the second codebook to the base station side, a part of overhead can be reduced.
  • Step 104 sending the second codebook to the base station side, where the second codebook is used to enable the base station side to generate channel state information of the channel.
  • the terminal side after the terminal side generates the second codebook, it can send the generated second codebook to the base station side in any communication mode. Preferably, before transmitting the second codebook to the base station side, It is also possible to perform operations such as quantization, encoding, and modulation on the second codebook, and then transmit it to the base station side. After receiving the second codebook, the base station side can perform corresponding inverse operations on the second codebook to generate the first codebook. codebook, and then perform an attention mechanism operation on the first codebook, extract channel parameters from it, and then restore channel state information according to the channel parameters.
  • the channel parameters of the channel are obtained; the channel parameters are subjected to feature extraction processing according to the preset attention mechanism, and the first codebook of the channel is generated; the first codebook is generated according to the preset coding network.
  • the codebook is encoded to generate a second codebook of the channel; the second codebook is sent to the base station side, wherein the second codebook is used to enable the base station side to generate the channel state information of the channel;
  • the encoding network is used for encoding processing, so that the key information of the channel parameters in the second codebook generated by the terminal side is more significant, and the key information of the channel parameters is weakened during the transmission process.
  • the implementation of the present application relates to a codebook transmission method, which is applied on the terminal side, as shown in Figure 2, and specifically includes the following steps:
  • Step 201 acquire channel parameters of the channel.
  • this step is substantially the same as step 101 provided in the embodiment of the present application, and details are not repeated here.
  • Step 202 receiving the attention mechanism activation parameter, attention mechanism parameter set and network parameter set sent by the base station side.
  • the attention mechanism activation parameters received by the terminal side are used by the terminal to determine whether to enable the attention mechanism, the attention mechanism parameter set is used by the terminal to select the corresponding attention mechanism parameters according to its own situation, and the network parameter set is used for The terminal selects corresponding network parameters according to its own conditions.
  • the terminal can also receive the attention mechanism activation parameters and network parameters first, and whether the terminal meets the attention mechanism activation parameters according to its own situation.
  • the terminal side returns the attention mechanism to the base station side.
  • the force mechanism activation indication the base station side returns the attention mechanism parameter set according to the attention mechanism activation indication, if the terminal side does not need to enable the attention mechanism, it does not need to feedback the attention mechanism activation indication, or the attention mechanism parameter index is 0(0 indicates that the attention mechanism is not enabled).
  • Step 203 judge whether to enable the attention mechanism according to the channel type information and the attention mechanism enabling parameters.
  • the channel type information obtained by the terminal side includes three types of information: channel scenario, channel angle extension, and channel delay extension. According to at least one of the three types of information contained in the channel type information, it is judged whether it is consistent with the attention sent by the base station side. Whether the corresponding information contained in the activation parameters of the force mechanism match, when the information on the terminal side and the base station side match, it is deemed that the attention mechanism can be enabled.
  • Step 204 when the attention mechanism is enabled, select the attention mechanism parameters matching the channel type information from the attention mechanism parameter set, and generate the attention mechanism according to the attention mechanism parameters.
  • the attention mechanism parameter set includes two types of attention mechanism combination parameters and attention mechanism structure parameters.
  • the attention mechanism combination parameters include the four attention mechanism combinations shown in Figure 1a- Figure 1c.
  • Select an appropriate attention mechanism combination according to the channel type information, and the attention mechanism structure parameters include maximum pooling parameters, average pooling parameters, network weight parameters, convolution kernel parameters, and activation functions to clarify the attention mechanism
  • the parameters of the structure after specifying which attention mechanism combination to use on the terminal side, can select the relevant parameters used by the attention mechanism in the attention mechanism combination from the attention mechanism structure parameters according to the channel type information, and according to The relevant parameters of the selected attention mechanism and the combination of the attention mechanism generate the attention mechanism on the terminal side.
  • Step 205 selecting a first network parameter set that matches the channel type information from the network parameter sets.
  • the network parameter set includes various network parameter combinations, and the terminal side will select the first network parameter set that meets the requirements of the terminal side according to the channel type information after receiving the network parameter set.
  • Step 206 Select network parameters matching the codebook type information and codebook number information from the first network parameter set, and generate a coding network according to the network parameters.
  • the acquired channel parameters on the terminal side also include two types of information: codebook type information and codebook number information.
  • the codebook type information specifies the codebook style that the terminal side needs to generate, and the codebook number information
  • the information specifies the number of codebooks that need to be generated by the terminal side. Therefore, the terminal side also needs to further screen the first network parameter set generated in step 205 according to the codebook type information and codebook number information to meet the requirements of the terminal side.
  • Network parameters where the network parameters include compression rate, activation function, number of network layers, network layer mapping, network layer weights, network layer bias, network layer weight normalization coefficient, network layer attention mechanism parameters, etc. , after the network parameters are determined, the coding network used by the terminal side can be generated according to all the information in the network parameters.
  • Step 207 perform feature extraction processing on the channel parameters according to the attention mechanism, and generate a first codebook of the channel.
  • this step is substantially the same as step 102 provided in the embodiment of the present application, and details are not repeated here. .
  • Step 208 Perform encoding processing on the first codebook according to the encoding network to generate a second codebook for the channel.
  • this step is substantially the same as step 103 provided in the embodiment of the present application, and details are not repeated here. .
  • Step 209 sending the second codebook to the base station side, where the second codebook is used to enable the base station side to generate channel state information of the channel.
  • this step is substantially the same as step 104 mentioned in the embodiment of the present application, and details are not repeated here.
  • the embodiment of the present application can also communicate with the base station side before transmitting the second codebook, so that the terminal side can select the attention mechanism and coding network supported by the base station side and meet the channel requirements, and then It is ensured that the optimal feature extraction and encoding of channel parameters can be performed, and the quality of the generated second codebook can be improved.
  • the implementation of the present application relates to a codebook transmission method, which is applied on the terminal side, as shown in Figure 3, and specifically includes the following steps:
  • Step 301 acquire channel parameters of the channel.
  • this step is substantially the same as step 101 provided in the embodiment of the present application, and details are not repeated here.
  • Step 302 receiving the attention mechanism activation parameter, attention mechanism parameter set and network parameter set sent by the base station side.
  • this step is substantially the same as step 202 provided in the embodiment of the present application, and details are not repeated here.
  • Step 303 judging whether to enable the attention mechanism according to the channel type information and the attention mechanism enabling parameters.
  • this step is substantially the same as step 203 provided in the embodiment of the present application, and details are not repeated here.
  • Step 304 when the attention mechanism is enabled, select the attention mechanism parameters matching the channel type information from the attention mechanism parameter set, and generate the attention mechanism according to the attention mechanism parameters.
  • this step is substantially the same as step 204 provided in the embodiment of the present application, and details are not repeated here.
  • Step 305 selecting a first network parameter set that matches the channel type information from the network parameter sets.
  • this step is substantially the same as step 205 provided in the embodiment of the present application, and details are not repeated here.
  • Step 306 Select network parameters matching the codebook type information and codebook number information from the first network parameter set, and generate a coding network according to the network parameters.
  • this step is substantially the same as step 206 provided in the embodiment of the present application, and details are not repeated here.
  • Step 307 Obtain the attention mechanism index and the network index corresponding to the attention mechanism parameter and the network parameter from the attention mechanism parameter set and the network parameter set.
  • the attention mechanism parameter set also includes the relative attention mechanism parameter index.
  • the network parameter set also includes the network parameter index corresponding to the network parameter.
  • the network parameter index is actually the encoding network parameter index of the encoding network used by the terminal side.
  • the network parameter set sent by the base station may only contain a set of network parameters. After receiving it, the terminal side directly uses this set of network parameters to generate a coded network, and the network parameter index returned by the terminal side is 1. (1 indicates that the base station side is instructed to use the decoding network corresponding to the network parameter index).
  • Step 308 sending the attention mechanism index and the network index to the base station side.
  • the attention mechanism parameter index and the network parameter index need to be sent to the base station side, so that the base station side can clarify the attention mechanism and encoding network used by the terminal side. .
  • Step 309 perform feature extraction processing on the channel parameters according to the attention mechanism, and generate a first codebook of the channel.
  • this step is substantially the same as step 102 mentioned in the embodiment of the present application, and details are not repeated here.
  • Step 310 Perform encoding processing on the first codebook according to the encoding network to generate a second codebook for the channel.
  • this step is substantially the same as step 103 mentioned in the embodiment of the present application, and details are not repeated here.
  • Step 311 sending the second codebook to the base station side, where the second codebook is used to enable the base station side to generate channel state information of the channel.
  • this step is substantially the same as step 104 mentioned in the embodiment of the present application, and details are not repeated here.
  • the attention mechanism index of the attention mechanism used by the terminal side and the network index of the encoding network can also be sent to the base station side, which can make the base station side clear how the base station side is Channel parameters are processed, so that the base station side can accurately obtain information from the second codebook.
  • the implementation of the present application relates to a codebook transmission method, as shown in Figure 4, specifically including the following steps:
  • Step 401 acquire channel parameters of the channel.
  • this step is substantially the same as step 101 mentioned in the embodiment of the present application, and details are not repeated here.
  • Step 402 generating a channel matrix of the channel according to the channel parameters.
  • this step is roughly the same as the method for generating the channel matrix mentioned in step 102 mentioned in the embodiment of the present application, and will not be repeated here.
  • Step 403 process the channel matrix to obtain a precoding matrix.
  • the frequency-domain granularity dimension information or time-domain granularity information of the channel matrix H can also be operated to obtain the precoding matrix.
  • the precoding matrix can be, but not limited to, a frequency-domain compressed channel matrix, Time domain compressed channel matrix or channel state eigenvector matrix.
  • the port space dimension information is not limited to the first dimension, it can also be in the second dimension, then at this time H is a complex matrix of H c ⁇ N.
  • the real part and imaginary part of a complex number can also be disassembled into two real numbers as channel dimension information, and the channel dimension information can be in the first dimension or the third dimension, that is, H is 2 ⁇ N ⁇ N c or N ⁇ N c ⁇ 2 real number matrix
  • the spatial dimension information of the time-domain compression matrix is the height dimension information and width dimension information composed of N and N c .
  • the acquisition of the frequency-domain compressed channel matrix is as follows: assuming that the channel matrix H has M PRB physical resource blocks, each channel of K consecutive PRB physical resource blocks constitutes a sub-band channel matrix, and the K sub-band channel matrix is taken in the receiving antenna dimension Any one of the PRB channels in the PRB channel is used as the channel of the subband, or the channel of the PRB corresponding to the maximum value of the K PRB channels is used as the channel of the subband, or the average of the K PRB channels is taken value channel as the channel for this subband.
  • the real part and imaginary part of a complex number can also be disassembled into two real numbers as the s-channel dimension information, and the channel dimension information can be in the first dimension or the third dimension, that is, H c is 2 ⁇ N t ⁇ N s Or a real number matrix of N t ⁇ N s ⁇ 2, and the spatial dimension information of the frequency domain compressed channel state information matrix is the height dimension information and width dimension information composed of N t and N s .
  • H has M PRBs, and each continuous channel of K PRBs constitutes a sub-band channel matrix.
  • K channels H 1 ... H k in each sub-band by Using their conjugate matrices H 1 H ... H k H , get K correlation matrices R 1 ...
  • R k add up the K correlation matrices and perform Singular Value Decomposition (SVD for short), and take the largest feature
  • the spatial dimension of the port is not limited to being in the first dimension, and it can also be in the second dimension.
  • V is a complex matrix of N s ⁇ N.
  • the real part and imaginary part of a complex number can also be disassembled into two real numbers as channel dimension information, and the channel dimension information can be in the first dimension or the third dimension, that is, V is 2 ⁇ N ⁇ N s or N ⁇ N s ⁇ 2 real number matrix, and the spatial dimension information of the channel state eigenvector matrix is the height dimension information and width dimension information composed of N t and N s .
  • the channel matrix and precoding matrix can be generated according to the channel parameters after the terminal side obtains the channel parameters, or the terminal side can input the channel parameters into the attention mechanism after obtaining the channel parameters , generated by the attention mechanism, but the methods used to generate the channel matrix and precoding matrix are the same.
  • Step 404 perform feature extraction processing on the precoding matrix according to a preset attention mechanism, and generate a first codebook of the channel.
  • this step is roughly the same as the method of feature extraction using the attention mechanism mentioned in step 102 mentioned in the embodiment of the present application, and will not be repeated here.
  • Step 405 Perform encoding processing on the first codebook according to a preset encoding network to generate a second codebook for the channel.
  • this step is substantially the same as step 103 mentioned in the embodiment of the present application, and details are not repeated here.
  • Step 406 sending the second codebook to the base station side, where the second codebook is used to enable the base station side to generate channel state information of the channel.
  • this step is substantially the same as step 104 mentioned in the embodiment of the present application, and details are not repeated here.
  • the channel matrix can also be generated, and the channel matrix can be compressed, and then the secondary compression can be performed through the encoding process, which can reduce the time required for the terminal side to transmit the second codebook. overhead.
  • the implementation of the present application relates to a codebook transmission method, as shown in Figure 5, applied on the base station side, specifically including the following steps:
  • Step 501 receiving a second codebook of a channel sent by the terminal side, wherein the second codebook is generated by the terminal side according to channel parameters of the channel.
  • the base station side receives the second codebook sent by the terminal side, and the received second codebook is generated by the terminal side according to channel parameters of the channel.
  • Step 502 using a preset decoding network to perform decoding processing on the second codebook to generate a third codebook.
  • the base station side will first use the decoding network to decode the second codebook to perform a restoration operation on the second codebook to restore it to the third codebook, where , the decoding network exists corresponding to the encoding network on the terminal side, and the third codebook is substantially the same as the first codebook mentioned in the embodiment of this application, because the second codebook is generated by compressing and encoding the first codebook Yes, the third codebook generated after decoding the second codebook at the base station should be substantially the same as the information contained in the first codebook.
  • Step 503 using a preset attention mechanism to perform feature extraction on the third codebook.
  • the attention mechanism is also used to perform feature extraction operations on the third codebook, so that the key information in the third codebook is more prominent.
  • Step 504 analyzing the third codebook after feature extraction to obtain channel parameters of the channel, and generating channel state information of the channel according to the channel parameters.
  • the base station can obtain the corresponding channel parameters from the third codebook, and then generate the channel state according to the information contained in the channel parameters information.
  • the implementation of the present application relates to a codebook transmission method, as shown in FIG. 6, which is applied on the base station side, and specifically includes the following steps:
  • Step 601 sending attention mechanism enabling parameters, attention mechanism parameter sets and network parameter sets to the terminal side, wherein the attention mechanism parameter set includes attention mechanism parameters and attention mechanism indexes corresponding to attention mechanism parameters, network parameter set It includes network parameters and network indexes corresponding to the network parameters.
  • the base station before receiving the second codebook, the base station first needs to negotiate with the terminal side, and the base station sends the attention mechanism activation parameters, attention mechanism parameter sets, and network parameter sets contained in itself to the terminal side, so as to For the terminal side to select appropriate attention mechanism parameters and network parameters according to its own situation, and then generate the corresponding attention mechanism and encoding network.
  • Step 602 receiving the attention mechanism index and the network index returned by the terminal side according to the attention mechanism enabling parameter, the attention mechanism parameter set and the network parameter set.
  • the terminal side after specifying the attention mechanism parameters and network parameters used by the terminal side, it can select the attention mechanism index and Network index, if the network parameter set sent by the base station side only contains a set of network parameters, the terminal side will directly use this set of network parameters to generate a coding network after receiving it, and the network parameter index returned by the terminal side is 1 (1 indicates that the base station side uses its own decoding network).
  • Step 603 selecting the attention mechanism parameter corresponding to the attention mechanism index from the attention mechanism parameter set, and generating the attention mechanism according to the attention mechanism parameter.
  • the base station side After receiving the attention mechanism index sent by the terminal side, the base station side will select attention mechanism parameters corresponding to the attention mechanism index, and then generate the attention mechanism used by the base station side.
  • Step 604 Select a network parameter corresponding to the network index from the network parameter set, and generate a decoding network according to the network parameter.
  • the base station after receiving the network index sent by the terminal side, the base station will select the network parameter corresponding to the network index from the network parameter set on the base station side, and then generate the decoding network used by the base station side.
  • the network parameter index received by the base station side is 1, which is used to instruct the base station side to use the decoding network corresponding to the network parameter index fed back by the terminal side.
  • Step 605 receiving the second codebook of the channel sent by the terminal side, where the second codebook is generated by the terminal side according to channel parameters of the channel.
  • this step is substantially the same as step 501 in the embodiment of the present application, and details are not repeated here.
  • Step 606 Use a preset decoding network to decode the second codebook to generate a third codebook.
  • this step is substantially the same as step 502 in the embodiment of the present application, and details are not repeated here.
  • Step 607 using a preset attention mechanism to perform feature extraction on the third codebook.
  • this step is substantially the same as step 503 in the embodiment of the present application, and details are not repeated here.
  • Step 608 analyzing the third codebook after feature extraction to obtain channel parameters of the channel, and generating channel state information of the channel according to the channel parameters.
  • this step is substantially the same as step 504 in the embodiment of the present application, and details are not repeated here.
  • the embodiment of the present application relates to a terminal, as shown in FIG. 7 , including:
  • the attention mechanism module 702 is used to perform feature extraction processing on the channel parameters according to the preset attention mechanism, and generate the first codebook of the channel;
  • An encoding module 703, configured to perform encoding processing on the first codebook according to a preset encoding network to generate a second codebook of the channel;
  • the transmission module 704 is configured to send the second codebook to the base station side, where the second codebook is used to enable the base station side to generate channel state information of the channel.
  • the position of the attention mechanism module on the terminal side is not limited, as shown in Figure 7, it can be located between the acquisition module and the encoding module, as shown in Figure 7a, it can be located in the encoding module, Alternatively, as shown in FIG. 7b , an attention mechanism module 7021 and an attention mechanism module 7022 may be set in the terminal to perform feature extraction on channel parameters twice, so that the key information of channel parameters is more prominent.
  • the embodiment of the present application relates to a base station, as shown in FIG. 8 , including:
  • the receiving module 801 is configured to receive a second codebook of a channel sent by the terminal side, where the second codebook is generated by the terminal side according to channel parameters of the channel;
  • Decoding module 802 configured to use a preset decoding network to decode the second codebook to generate a third codebook
  • An attention mechanism module 803, configured to extract features from the third codebook using a preset attention mechanism
  • the parsing module 804 is configured to parse the third codebook after feature extraction to obtain channel parameters of the channel, and generate channel state information of the channel according to the channel parameters.
  • the location of the attention mechanism module on the base station side is not limited, as shown in Figure 8, it can be located between the decoding module and the parsing module, as shown in Figure 8a, it can be located in the decoding module, Alternatively, as shown in FIG. 8b , an attention mechanism module 8031 and an attention mechanism module 8032 may be set in the base station to perform feature extraction on the first codebook twice, so that the recovered key information of channel parameters is more accurate.
  • the embodiments of the present application relate to a computer-readable storage medium storing a computer program.
  • the computer program is executed by the processor, the above-mentioned method embodiments are realized.
  • a storage medium includes several instructions to make a device ( It may be a single-chip microcomputer, a chip, etc.) or a processor (processor) to execute all or part of the steps of the methods described in the various embodiments of the present application.
  • the aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disc, etc., which can store program codes. .

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • Electromagnetism (AREA)
  • Quality & Reliability (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

本申请提出了一种码本传输方法、终端、基站和存储介质，涉及无线通信领域。码本传输方法包括：获取信道的信道参数；根据预设的注意力机制对所述信道参数进行特征提取处理，生成所述信道的第一码本；根据预设的编码网络对所述第一码本进行编码处理，生成所述信道的第二码本；发送所述第二码本至基站侧，其中，所述第二码本用于使所述基站侧生成所述信道的信道状态信息。

Description

码本传输方法、终端、基站和存储介质
相关申请的交叉引用
本申请基于申请号为“202110914141.4”、申请日为2021年8月10日的中国专利申请提出，并要求该中国专利申请的优先权，该中国专利申请的全部内容在此以引入方式并入本申请。
技术领域
本申请实施方式涉及无线通信领域，特别涉及一种码本传输方法、终端、基站和存储介质。
背景技术
从长期演进(Long Term Evolution，简称LTE)发展到现在的新无线接入技术(New Radio Access Technology，简称NR)，多天线技术一直是通信标准中重要的技术之一。随着通信的标准指标在不断地提升，如何更加准确地获取信道状态信息(Channel State Information，简称CSI)是多天线技术性能提升的关键所在。
然而，在多天线系统中，基站侧是通过用户设备侧(User Equipment，简称UE)反馈的来获取CSI的，基站侧所接收的CSI会由于传输环境的影响而产生衰减、反射等影响导致基站侧获取的CSI的准确度较低。
发明内容
本申请实施方式提供了一种码本传输方法，所述方法包括以下步骤：获取信道的信道参数；根据预设的注意力机制对所述信道参数进行特征提取处理，生成所述信道的第一码本；根据预设的编码网络对所述第一码本进行编码处理，生成所述信道的第二码本；发送所述第二码本至基站侧，其中，所述第二码本用于使所述基站侧生成所述信道的信道状态信息。
本申请实施方式还提供了一种码本传输方法，所述方法包括以下步骤：接收终端侧发送信道的第二码本，其中，所述第二码本由所述终端侧根据所述信道的信道参数生成；利用预设的解码网络对所述第二码本进行解码处理生成第三码本；利用预设的注意力机制对所述第三码本进行特征提取；解析进行特征提取后的所述第三码本获取所述信道的信道参数；根据所述信道参数生成所述信道的信道状态信息。
本申请实施方式还提出了一种终端，包括：获取模块，用于获取信道的信道参数；注意力机制模块，用于根据预设的注意力机制对所述信道参数进行特征提取处理，生成所述信道的第一码本；编码模块，用于根据预设的编码网络对所述第一码本进行编码处理，生成所述信道的第二码本；传输模块，用于发送所述第二码本至基站侧，其中，所述第二码本用于使所述基站侧生成所述信道的信道状态信息。
本申请实施方式还提出了一种基站，包括：接收模块，用于接收终端侧发送信道的第二 码本，其中，所述第二码本由所述终端侧根据所述信道的信道参数生成；解码模块，用于利用预设的解码网络对所述第二码本进行解码处理生成第三码本；注意力机制模块，用于利用预设的注意力机制对所述第三码本进行特征提取；解析模块，用于解析进行特征提取后的所述第三码本获取所述信道的信道参数，并根据所述信道参数生成所述信道的信道状态信息。
本申请实施方式还提出了一种计算机可读存储介质，存储有计算机程序，其特征在于，所述计算机程序被处理器执行时实现权利要求上述任一项所述码本传输方法。
附图说明
图1是本申请实施方式提供的码本传输方法的流程图；
图1a是本申请实施例提供的注意力机制的结构示意图；
图1b是本申请实施例提供的注意力机制的结构示意图；
图1c是本申请实施例提供的注意力机制的结构示意图；
图1d是本申请实施例提供的空间注意力机制的结构图；
图1e是本申请实施例提供的通道注意力机制的结构图；
图2是本申请实施方式提供的码本传输方法的流程图；
图3是本申请实施方式提供的码本传输方法的流程图；
图4是本申请实施方式提供的码本传输方法的流程图；
图5是本申请实施方式提供的码本传输方法的流程图；
图6是本申请实施方式提供的码本传输方法的流程图；
图7是本申请实施方式提供的终端的结构示意图；
图7a是本申请实施方式提供的终端的结构示意图；
图7b是本申请实施方式提供的终端的结构示意图；
图8是本申请实施方式提供的基站的结构示意图；
图8a是本申请实施方式提供的基站的结构示意图；
图8b是本申请实施方式提供的基站的结构示意图。
具体实施方式
本申请实施方式的主要目的在于提出一种码本传输方法、终端、基站和存储介质，旨在实现提高终端侧生成的码本信息的质量，使得基站侧根据码本信息所获取的CSI的准确度增加。
为使本申请实施方式的目的、技术方案和优点更加清楚，下面将结合附图对本申请的各实施方式进行详细的阐述。然而，本领域的普通技术人员可以理解，在本申请各实施方式中，为了使读者更好地理解本申请而提出了许多技术细节。但是，即使没有这些技术细节和基于以下各实施方式的种种变化和修改，也可以实现本申请所要求保护的技术方案。以下各个实施方式的划分是为了描述方便，不应对本申请的具体实现方式构成任何限定，各个实施方式在不矛盾的前提下可以相互结合相互引用。
本申请的实施方式涉及一种码本传输方法，应用在无线通信系统的终端侧，该无线通信系统，基站侧有N t个端口(可以是物理天线，或者逻辑天线)，终端侧有N r个端口，如图1所示，具体包括以下步骤：
步骤101，获取信道的信道参数。
具体地说，终端侧所获取信道的信道参数包括信道基矢量、信道基矢量数目、最大端口数目、信道类型信息、码本类型信息、码本个数信息、带宽指示、频域单元等，其中，信道类型信息可以包括信道场景、信道角度扩展、信道时延扩展等表示信道类型的信息，其中，信道基矢量可以是用于变换信道的基矢量，可以用它来表示基站侧到终端侧之间的信道特征，当信道为频域信道时，信道基矢量为频域基矢量，当信道为时域信道时，信道基矢量为时域基矢量，优选地，信道基矢量为N F*1离散傅里叶(Discrete Fourier Transform，简称DFT)矢量，或者由多个DFT矢量组成的矢量；信道基矢量数目N是信道基矢量的数量，信道基矢量数目是根据频域单元的数目对应的频域单元取值集合确定的，最大端口数目可以是通信天线的数量或者基站侧具有通信端口的最大数量，信道类型指示可以是指示信道类型的信息(比如城市宏，城市微，室内，工厂，办公区，商场，广场，地铁等)，带宽指示可以是指示通信带宽的信息(比如子带个数，物理资源块个数，带宽大小)，频域单元可以是频域的基本单元，可以是用于指示频域信道基础的信息，除此之外，终端侧还可以根据信道的信道参数生成该信道的信道矩阵。
步骤102，根据预设的注意力机制对信道参数进行特征提取处理，生成信道的第一码本。
具体地说，终端侧所获取的信道参数是信道矩阵的描述信息，终端侧首先会对输入的信道参数进行处理来生成信道的信道矩阵，或者通过接收下行的导频信道，估计出的终端的N r个端口的信道矩阵H k，H k为一个符号上的第k个频域粒度上的信道，为N r×N t的复数矩阵，k＝1，2，…，M，其中，M为正整数，频域粒度可以是子载波，物理资源块(Physical Resource Block，简称PRB)，物理资源块组(Physical Resource block Group，简称PRG)，或者，子带(一个子带包括至少一个PRB)，H为H k的组合，比如属于N r×N t×M的复数矩阵，或者将第k个频域粒度上的信道N r×N t复数矩阵拉成一个N×1的列向量，因此另一种信道矩阵H的形式为N×M的复数矩阵，其中N＝N r×N t，需要说明的是，这里不局限于端口的空间维度在第一个维度，它也可以在第二维度，那么此时，H是一个N×M的复数矩阵；所生成的信道矩阵中包含有空间维度信息和通道维度信息，空间维度信息包含高度维度信息和宽度维度信息(由N r、N t组成)，通道维度信息是信道矩阵的实部虚部所拆开成两个实数；而由于无线信道无线频域信道和无线时域信道，无线频域信道通常是由端口空间维度信息、频域粒度维度信息、时间维度信息、通道维度信息等维度构成，无线时域信道通常是由端口空间维度信息、时域粒度维度信息、通道维度信息及时间维度信息等维度构成，信道矩阵的通道维度信息还可以替换为端口空间维度信息、粒度维度信息或时间维度信息中的任意一种，其中，当信道矩阵包含时间维度信息时，所获取的信道矩阵是四维矩阵，其余维度信息为三维矩阵；除此之外，在获取到信道矩阵之后，终端侧会将所生成的信道矩阵输入到注意力机制中进行处理，而在将信道矩阵输入到注意力机制之前，还可以对信道矩阵进行处理来生成预编码矩阵，将预编码矩阵输入到注意力机制中进行处理，而终端侧所设置的注意力机制(Attention Mechanism，简称AM)均可以对信道矩阵或者预编码矩阵进行处理。
AM在对信道矩阵或预编码矩阵进行处理时，实际上处理的信息是信道矩阵或预编码矩阵中所包含的空间维度信息和通道维度信息(也可以是端口空间维度信息、粒度维度信息或时间维度信息)，本申请所使用的注意力机制可以是如图1a所示的空间注意力机制(Spatial Attention，简称SA)或通道注意力机制(Channel Attention，简称CA)、如图1b所示的先通道后空间的注意力机制或如图1c所示的先空间后通道的注意力机制，而输入AM的包括空间 维度信息HW和通道维度信息C，空间维度信息又包含高度维度信息H和宽度维度信息W，也就是说输入到注意力机制的至少得是一个包含三种维度的三维信息(输入的时间维度信息是四维信息)，通道维度信息可为第一个维度或者最后一个维度，即可表示为C×H×W或H×W×C。
其中，SA是对输入数据的空间维度信息上进行重点特征权重比例再分配，对输入的信道参数的通道维度信息分别进行最大池化操作与平均池化操作，最大池化操作即取通道维度信息所有数值的最大值，平均池化操作即对通道维度信息所有数值取平均值，由此可以得到两个大小为H×W×1的张量，将二者按通道维度叠放在一起构成H×W×2的张量。利用尺寸为k×k×2大小的卷积核对补零后的H×W×2张量进行卷积操作，其中k∈[1，min(H，W)]的正整数，生成大小为H×W×1的张量，并用激活函数对其处理，最终生成空间注意力特征图，并将其与输入的信道参数在通道维度信息上分别点对点相乘最终生成SA处理后的第一码本，整体流程如图1d所示。
其中，CA是对输入数据的通道维度信息上进行重点特征权重比例再分配，对输入的信道信息的空间维度信息分别进行最大池化操作与平均池化操作，最大池化操作即取空间维度信息上所有数值的最大值，平均池化即对空间维度信息所有数值取平均值，由此可以得到两个大小为1×1×C的张量，将二者分别送入全连接网络或者卷积神经网络，需要说明的是经过全连接网络或者卷积神经网络后输出数据的尺寸应与输入数据尺寸保持相同。将得到的两个尺寸为1×1×C的池化输出张量相加并用激活函数将其处理，最终生成通道注意力特征图，并将其与输入数据在空间维度信息上分别点对点相乘最终生成通道注意力机制处理后的输出数据，整体流程如图1e所示。
步骤103，根据预设的编码网络对第一码本进行编码处理，生成信道的第二码本。
具体地说，在终端侧的编码网络包括X个网络层，其中第i层包括L e,i个节点，包括至少一个网络层权值W e,i，0-1个网络层偏置b e,i，激活函数A e,i，i＝1，…，M；其中第i层可以是一个卷积层，也可以是一个全连接层，或者池化层，或者是几个卷积层组成的残差网络块，这些值都是这个编码网络的网络参数，使用该编码网络对注意力机制所生成的第一码本进行编码处理，便可以获取到第二码本，其中，第一码本和第二码本对应的元素个数是不同的，第二码本的元素个数一般元小于第一码本对应的元素个数，第二码本的元素个数和第一码本的元素个数的比值叫压缩率，比如第一码本总共包括1024个元素，而第二码本只有64个元素，那么压缩率为64/2048＝1/32，采用编码网络对第一码本进行处理，实际上就是对第一码本进行了一个压缩操作，使得终端侧在向基站侧传输第二码本时，可以减少一部分开销。
步骤104，发送第二码本至基站侧，其中，第二码本用于使基站侧生成信道的信道状态信息。
具体地说，终端侧在生成第二码本之后，便可以以任意一种通信方式将所生成的第二码本发送至基站侧，优选地，在将第二码本传输给基站侧之前，也可以对第二码本进行量化、编码、调制等操作，之后再将其传输给基站侧，基站侧在接收第二码本之后，便可以对第二码本进行相应的逆操作生成第一码本，之后对第一码本进行注意力机制操作之后，从中提取信道参数，之后根据信道参数来恢复信道状态信息。
本申请的实施方式，在终端侧，获取信道的信道参数；根据预设的注意力机制对信道参数进行特征提取处理，生成所述信道的第一码本；根据预设的编码网络对第一码本进行编码 处理，生成信道的第二码本；发送第二码本至基站侧，其中，第二码本用于使基站侧生成所述信道的信道状态信息；通过使用注意力机制对信道参数进行提取之后再使用编码网络进行编码处理，使得终端侧所生成的第二码本中的信道参数的关键信息更加显著，减弱信道参数的关键信息在传输过程中受到传输环境的影响，解决了现有技术中基站侧根据码本信息所获取的CSI准确度较低技术问题。
本申请的实施方式涉及一种码本传输方法，应用在终端侧，如图2所示，具体包括以下步骤：
步骤201，获取信道的信道参数。
具体地说，本步骤与本申请实施例所提供的步骤101大致相同，此处不一一赘述。
步骤202，接收基站侧发送的注意力机制启用参数、注意力机制参数集合和网络参数集合。
具体地说，终端侧所接收的注意力机制启用参数用于终端判断是否要启用注意力机制，注意力机制参数集合用于终端根据自身情况选取相对应的注意力机制参数，网络参数集合用于终端根据自身情况选取相对应的网络参数。
此处需要注意的是，终端也可以先接收注意力机制启用参数和网络参数，在终端根据自身情况是否满足注意力机制启用参数，当满足注意力机制启用参数之后，终端侧向基站侧返回注意力机制启用指示，基站侧再根据注意力机制启用指示返回注意力机制参数集合，如果终端侧无需启用注意力机制，则不需要反馈注意力机制启用指示，或者注意力机制参数索引为0(0表示不启用注意力机制)。
步骤203，根据信道类型信息和注意力机制启用参数判断是否启用注意力机制。
具体地说，终端侧所获取的信道类型信息包含信道场景、信道角度扩展和信道时延扩展三种信息，根据信道类型信息中包含的三种信息至少之一来判断是否与基站侧发送的注意力机制启用参数中所包含的相对应的信息是否匹配，当终端侧与基站侧信息匹配的情况下，视为可以启用注意力机制。
步骤204，当启用注意力机制时，从注意力机制参数集合中选取与信道类型信息匹配的注意力机制参数，并根据注意力机制参数生成注意力机制。
具体地说，注意力机制参数集合包含注意力机制组合参数和注意力机制结构参数两种类型，注意力机制组合参数包含图1a-图1c所示的4中注意力机制组合情况，终端侧先根据信道类型信息选择一种合适的注意力机制组合情况，而注意力机制结构参数包含最大池化参数、平均池化参数、网络权重参数、卷积核参数以及激活函数等用于明确注意力机制结构的参数，终端侧在明确使用哪种注意力机制组合情况之后，便可以根据信道类型信息从注意力机制结构参数中选取注意力机制组合情况中的注意力机制所使用的相关参数，并根据所选择的注意力机制的相关参数和注意力机制组合情况生成终端侧的注意力机制。
步骤205，从网络参数集合中选取与信道类型信息匹配的第一网络参数集合。
具体地说，网络参数集合中包含了多种网络参数组合，终端侧在接收到网络参数集合之后会根据信道类型信息选择出符合终端侧要求的第一网络参数集合。
步骤206，从第一网络参数集合中选取与码本类型信息和码本个数信息匹配的网络参数，并根据网络参数生成编码网络。
具体地说，终端侧在所获取的信道参数中还包含由码本类型信息和码本个数信息两种信 息，码本类型信息明确了终端侧所需要生成的码本样式，码本个数信息明确了终端侧所需要生成的码本个数，因此，终端侧还需要根据码本类型信息和码本个数信息从步骤205所生成的第一网络参数集合中进一步筛选去符合终端侧要求的网络参数，其中，网络参数包含压缩率，激活函数，网络层数，网络层映射，网络层权值，网络层偏置，网络层权值归一化系数，网络层注意力机制参数等等，在网络参数确定之后，便可以根据网络参数中的所有信息生成终端侧所使用的编码网络。
步骤207，根据注意力机制对信道参数进行特征提取处理，生成信道的第一码本。
具体地说，本步骤与本申请实施例所提供的步骤102大致相同，此处不一一赘述。。
步骤208，根据编码网络对第一码本进行编码处理，生成信道的第二码本。
具体地说，本步骤与本申请实施例所提供的步骤103大致相同，此处不一一赘述。。
步骤209，发送第二码本至基站侧，其中，第二码本用于使基站侧生成信道的信道状态信息。
具体地说，本步骤与本申请实施例提及的步骤104大致相同，此处不一一赘述。
本申请的实施例，在其他实施例的基础之上还可以与基站侧在传输第二码本前进行通信，使得终端侧可以选择基站侧支持且满足信道要求的注意力机制和编码网络，进而保证可以对信道参数最佳的特征提取和编码，提高生成的第二码本的质量。
本申请的实施方式涉及一种码本传输方法，应用在终端侧，如图3所示，具体包括以下步骤：
步骤301，获取信道的信道参数。
具体地说，本步骤与本申请实施例所提供的步骤101大致相同，此处不一一赘述。
步骤302，接收基站侧发送的注意力机制启用参数、注意力机制参数集合和网络参数集合。
具体地说，本步骤与本申请实施例所提供的步骤202大致相同，此处不一一赘述。
步骤303，根据信道类型信息和注意力机制启用参数判断是否启用注意力机制。
具体地说，本步骤与本申请实施例所提供的步骤203大致相同，此处不一一赘述。
步骤304，当启用注意力机制时，从注意力机制参数集合中选取与信道类型信息匹配的注意力机制参数，并根据注意力机制参数生成注意力机制。
具体地说，本步骤与本申请实施例所提供的步骤204大致相同，此处不一一赘述。
步骤305，从网络参数集合中选取与信道类型信息匹配的第一网络参数集合。
具体地说，本步骤与本申请实施例所提供的步骤205大致相同，此处不一一赘述。
步骤306，从第一网络参数集合中选取与码本类型信息和码本个数信息匹配的网络参数，并根据网络参数生成编码网络。
具体地说，本步骤与本申请实施例所提供的步骤206大致相同，此处不一一赘述。
步骤307，从注意力机制参数集合和网络参数集合中获取与注意力机制参数和网络参数对应的注意力机制索引和所述网络索引。
具体地说，注意力机制参数集合中除了包含注意力机制参数之外，还包含相对于的注意力机制参数索引，同样的，网络参数集合中还包含网络参数对应的网络参数索引，在终端侧选择了相对于的注意力机制参数和网络参数之后，还需要从注意力机制参数集合中和网络参数集合中选取与注意力机制参数和网络参数对应的注意力机制参数索引和网络参数索引，其 中，网络参数索引实际上是终端侧所使用的编码网络的编码网络参数索引。
此处需要注意的是：基站侧所发送的网络参数集合中可能只包含一组网络参数，终端侧接收到之后直接采用这一组网络参数生成编码网络，终端侧所返回的网络参数索引为1(1表示指示基站侧使用网络参数索引对应的解码网络)。
步骤308，将注意力机制索引和网络索引发送至基站侧。
具体地说，在注意力机制参数索引和网络参数索引明确之后，还需要将注意力机制参数索引和网络参数索引发送至基站侧，以使基站侧明确终端侧所使用的注意力机制和编码网络。
步骤309，根据注意力机制对信道参数进行特征提取处理，生成信道的第一码本。
具体地说，本步骤与本申请实施例提及的步骤102大致相同，此处不一一赘述。
步骤310，根据编码网络对第一码本进行编码处理，生成信道的第二码本。
具体地说，本步骤与本申请实施例提及的步骤103大致相同，此处不一一赘述。
步骤311，发送第二码本至基站侧，其中，第二码本用于使基站侧生成信道的信道状态信息。
具体地说，本步骤与本申请实施例提及的步骤104大致相同，此处不一一赘述。
本申请的实施例，在其他实施例的基础之上还可以将终端侧所使用的注意力机制的注意力机制索引和编码网络的网络索引发送至基站侧，可以使得基站侧明确基站侧是如何对信道参数进行处理的，从而基站侧可以准确的从第二码本中获取信息。
本申请的实施方式涉及一种码本传输方法，如图4所示，具体包括以下步骤：
步骤401，获取信道的信道参数。
具体地说，本步骤与本申请实施例提及的步骤101大致相同，此处不一一赘述。
步骤402，根据信道参数生成信道的信道矩阵。
具体地说，本步骤与本申请实施例提及的步骤102提及的生成信道矩阵的方法大致相同，此处不一一赘述。
步骤403，对信道矩阵进行处理获取预编码矩阵。
具体地说，在获取到信道矩阵之后，还可以对信道矩阵H的频域粒度维度信息或者时域粒度信息进行操作来获取预编码矩阵，预编码矩阵可以但不限于为频域压缩信道矩阵、时域压缩信道矩阵或信道状态特征向量矩阵。
其中，时域压缩信道矩阵的获取如下：假设时域信道矩阵H有N个天线端口数目，N＝Nt×Nr。同时时域信道有M个延时径，时延小的径信号功率强，时延大的径相对较小甚至趋于0，因此我们利用时域信道的稀疏性取矩阵的前N c个径的值作为代表,信道H可以压缩为H c,其矩阵大小为N×N c，N c为选取的时域径数,因此获得了时域压缩信道矩阵。需要说明的是这里不局限于端口空间维度信息在第一个维度，它也可以在第二维度，那么此时H是一个H c×N的复数矩阵。进一步地,复数的实部虚部也可以拆开成两个实数作为通道维度信息,通道维度信息可以在第一个维度或者第三个维度上,即H为2×N×N c或者N×N c×2的实数矩阵,而时域压缩矩阵的空间维度信息为N和N c组成的高度维度信息和宽度维度信息。
其中，频域压缩信道矩阵的获取如下：假设信道矩阵H有M个PRB物理资源块，每连续的K个PRB物理资源块的信道构成一个子带信道矩阵，在接收天线维度上取该K个PRB的信道中的任意一个PRB的信道作为该子带的信道，或者取该K个PRB的信道中的最大值对应PRB的信道作为该子带的信道，或者取该K个PRB的信道的平均值信道作为该子带的 信道。因此H可以压缩为H c，其矩阵大小为N t×N s，N s为子带的数目，其数值为N s＝M/K，因此获得了频域压缩信道矩阵，需要说明的是，这里不局限于端口空间维度信息在第一个维度，它也可以在第二维度，那么此时，H c是一个N s×N t的复数矩阵。进一步地，复数的实部虚部也可以拆开成两个实数作为s通道维度信息，通道维度信息可以在第一个维度或者第三个维度上，即H c为2×N t×N s或者N t×N s×2的实数矩阵，而频域压缩信道状态信息矩阵的空间维度信息为N t和N s组成的高度维度信息和宽度维度信息。
其中，信道状态特征向量矩阵的获取如下：H有M个PRB，每连续的K个PRB的信道构成一个子带信道矩阵，我们对每一个子带中的K个信道H 1…H k分别乘以它们的共轭矩阵H 1 H…H k H，得到K个相关矩阵R 1…R k，将该K个相关矩阵加和并做奇异值分解(Singular Value Decomposition，简称SVD)，取最大特征值对应的特征向量作为该子带的特征向量V i，其中i＝1…N s，将所有子带的特征向量堆叠在一起形成频域信道状态信息特征向量矩阵V，其矩阵大小为N t×N s，N s为子带的数目，其数值为N s＝M/K。需要说明的是，这里不局限于端口的空间维度在第一个维度，它也可以在第二维度，那么此时，V是一个N s×N的复数矩阵。进一步地，复数的实部虚部也可以拆开成两个实数作为通道维度信息，通道维度信息可以在第一个维度或者第三个维度上，即V为2×N×N s或者N×N s×2的实数矩阵，而信道状态特征向量矩阵的空间维度信息为N t和N s组成的高度维度信息和宽度维度信息。
此处需要注意的是，信道矩阵和预编码矩阵可以是终端侧在获取到信道参数之后就根据据信道参数生成，也可以是终端侧在获取信道参数之后，将信道参数输入到注意力机制中，由注意力机制生成，但所使用的生成信道矩阵和预编码矩阵的方法都是相同的。
步骤404，根据预设的注意力机制对预编码矩阵进行特征提取处理，生成信道的第一码本。
具体地说，本步骤与本申请实施例提及的步骤102提及的使用注意力机制进行特征提取的方法大致相同，此处不一一赘述。
步骤405，根据预设的编码网络对第一码本进行编码处理，生成信道的第二码本。
具体地说，本步骤与本申请实施例提及的步骤103大致相同，此处不一一赘述。
步骤406，发送第二码本至基站侧，其中，第二码本用于使基站侧生成信道的信道状态信息。
具体地说，本步骤与本申请实施例提及的步骤104大致相同，此处不一一赘述。
本申请的实施方式，在其他实施例的基础之上还可以生成信道矩阵，并对信道矩阵进行压缩操作，之后通过编码处理进行二次压缩，可以降低终端侧传输第二码本时所需的开销。
本申请的实施方式涉及一种码本传输方法，如图5所示，应用在基站侧，具体包括以下步骤：
步骤501，接收终端侧发送信道的第二码本，其中，第二码本由终端侧根据信道的信道参数生成。
具体地说，基站侧接收终端侧发送的第二码本，所接收的第二码本是终端侧根据信道的信道参数生成的。
步骤502，利用预设的解码网络对第二码本进行解码处理生成第三码本。
具体地说，基站侧在接收到第二码本之后，会先使用解码网络对第二码本进行解码处理，来对第二码本进行一个还原操作，使其恢复到第三码本，其中，解码网络是与终端侧的编码 网络相对应存在的，第三码本与本申请实施例提及的第一码本大体相同，因为第二码本是对第一码本进行压缩编码处理生成的，在基站测对第二码本进行解码处理后的所生成的第三码本应当与第一码本所包含的信息大体相同。
步骤503，利用预设的注意力机制对第三码本进行特征提取。
具体地说，第三码本生成之后，同样采取注意力机制对第三码本进行特征提取操作，使得第三码本中的关键信息更加显著。
步骤504，解析进行特征提取后的第三码本获取信道的信道参数，并根据信道参数生成信道的信道状态信息。
具体地说，基站侧在对第三码本进行特征提取处理之后，就可以从第三码本中获取到相对应的信道参数，之后便可以根据信道参数中所包含的信息来生成信道的状态信息。
本申请的实施方式涉及一种码本传输方法，如图6所示，应用在基站侧，具体包括以下步骤：
步骤601，向终端侧发送注意力机制启用参数、注意力机制参数集合和网络参数集合，其中，注意力机制参数集合包括注意力机制参数和注意力机制参数对应的注意力机制索引，网络参数集合包括网络参数和网络参数对应的网络索引。
具体地说，基站侧在接收第二码本之前，首先要和终端侧进行协商，基站侧将自身所包含的注意力机制启用参数、注意力机制参数集合和网络参数集合发送至终端侧，以供终端侧根据自身情况来选取合适和注意力机制参数和网络参数，进而生成相对应的注意力机制和编码网络。
步骤602，接收终端侧根据注意力机制启用参数、注意力机制参数集合和网络参数集合返回的注意力机制索引和网络索引。
具体地说，终端侧在明确自身所使用的注意力机制参数和网络参数之后，可以从注意力机制参数结合和网络参数集合中选取与注意力机制参数和网络参数相对应的注意力机制索引和网络索引，若基站侧所发送的网络参数集合中只包含一组网络参数，终端侧接收到之后直接采用这一组网络参数生成编码网络，终端侧返回的网络参数索引为1(1表示指示基站侧使用自身的解码网络)。
步骤603，从注意力机制参数集合中选取与注意力机制索引对应的注意力机制参数，并根据注意力机制参数生成注意力机制。
具体地说，基站侧在接收到终端侧所发送的注意力机制索引之后，会选择与注意力机制索引对应的注意力机制参数，进而生成基站侧所使用的注意力机制。
步骤604，从网络参数集合中选取与网络索引对应的网络参数，并根据网络参数生成解码网络。
具体地说，基站侧在接收到终端侧所发送的网络索引之后，会从基站侧的网络参数集合中选择与网络索引对应的网络参数，进而生成基站侧所使用的解码网络，而在基站侧只有一组网络参数时，基站侧所接收到的网络参数索引为1，用于指示基站侧使用终端侧所反馈的网络参数索引对应的解码网络。
步骤605，接收终端侧发送信道的第二码本，其中，第二码本由终端侧根据信道的信道参数生成。
具体地说，本步骤与本申请实施例的步骤501大致相同，此处不一一赘述。
步骤606，利用预设的解码网络对第二码本进行解码处理生成第三码本。
具体地说，本步骤与本申请实施例的步骤502大致相同，此处不一一赘述。
步骤607，利用预设的注意力机制对第三码本进行特征提取。
具体地说，本步骤与本申请实施例的步骤503大致相同，此处不一一赘述。
步骤608，解析进行特征提取后的第三码本获取信道的信道参数，并根据信道参数生成信道的信道状态信息。
具体地说，本步骤与本申请实施例的步骤504大致相同，此处不一一赘述。
本申请的实施方式涉及一种终端，如图7所示，包括：
获取模块701，用于获取信道的信道参数；
注意力机制模块702，用于根据预设的注意力机制对信道参数进行特征提取处理，生成信道的第一码本；
编码模块703，用于根据预设的编码网络对第一码本进行编码处理，生成信道的第二码本；
传输模块704，用于发送第二码本至基站侧，其中，第二码本用于使基站侧生成信道的信道状态信息。
此处需要注意的是：注意力机制模块在终端侧的位置并不受限制，如图7所示其可以位于获取模块和编码模块之间，如图7a所示其可以位于编码模块之中，也可以如图7b所示，在终端中设置注意力机制模块7021和注意力机制模块7022，对信道参数进行两次特征提取，使得信道参数的关键信息更加显著。
本申请的实施方式涉及一种基站，如图8所示，包括：
接收模块801，用于接收终端侧发送信道的第二码本，其中，第二码本由终端侧根据信道的信道参数生成；
解码模块802，用于利用预设的解码网络对第二码本进行解码处理生成第三码本；
注意力机制模块803，用于利用预设的注意力机制对第三码本进行特征提取；
解析模块804，用于解析进行特征提取后的第三码本获取信道的信道参数，并根据信道参数生成信道的信道状态信息。
此处需要注意的是：注意力机制模块在基站侧的位置并不受限制，如图8所示其可以位于解码模块和解析模块之间，如图8a所示其可以位于解码模块之中，也可以如图8b所示，在基站中设置注意力机制模块8031和注意力机制模块8032，对第一码本进行两次特征提取，使得所恢复的信道参数的关键信息更加准确。
本申请实施方式涉及一种计算机可读存储介质，存储有计算机程序。计算机程序被处理器执行时实现上述方法实施方式。
即，本领域技术人员可以理解，实现上述实施方式方法中的全部或部分步骤是可以通过程序来指令相关的硬件来完成，该程序存储在一个存储介质中，包括若干指令用以使得一个设备(可以是单片机，芯片等)或处理器(processor)执行本申请各个实施方式所述方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。
本领域的普通技术人员可以理解，上述各实施方式是实现本申请的具体实施方式，而在 实际应用中，可以在形式上和细节上对其作各种改变，而不偏离本申请的精神和范围。

Claims (10)

  1. 一种码本传输方法，应用在终端侧，所述方法包括：
    获取信道的信道参数；
    根据预设的注意力机制对所述信道参数进行特征提取处理，生成所述信道的第一码本；
    根据预设的编码网络对所述第一码本进行编码处理，生成所述信道的第二码本；
    发送所述第二码本至基站侧，其中，所述第二码本用于使所述基站侧生成所述信道的信道状态信息。
  2. 根据权利要求1所述的码本传输方法，其中，所述注意力机制为空间注意力机制或通道注意力机制；
    所述根据预设的注意力机制对所述信道参数进行特征提取处理，生成所述信道的第一码本，包括：
    根据所述信道参数生成所述信道的信道矩阵，其中，所述信道矩阵包含空间维度信息和通道维度信息；
    当所述注意力机制为所述空间注意力机制时，对所述通道维度信息分别进行池化处理获取最大池化特征和平均池化特征，对所述最大池化特征和所述平均池化特征进行卷积处理获取空间卷积特征，使用激活函数对所述空间卷积特征进行处理获取空间注意力机制特征，将所述空间注意力机制特征与所述信道参数在所述通道维度信息上进行相乘获取所述第一码本；
    当所述注意力机制为所述通道注意力机制时，对所述空间维度信息分别进行所述池化处理获取所述最大池化特征和所述平均池化特征，将所述最大池化特征和所述平均池化特征输入预设的神经网络进行处理获取最大卷积特征和平均卷积特征，使用所述激活函数对所述最大卷积特征和所述平均卷积特征的和值进行处理获取通道注意力机制特征，将所述通道注意力机制特征与所述信道参数在所述空间维度信息上进行相乘获取所述第一码本。
  3. 根据权利要求2所述的码本传输方法，其中，所述根据所述信道参数生成所述信道的信道矩阵之后还包括对所述信道矩阵进行处理获取预编码矩阵。
  4. 根据权利要求1至3中任一项所述的码本传输方法，其中，所述信道参数包括信道类型信息、码本类型信息和码本个数信息；
    所述获取信道的信道参数之后还包括：
    接收所述基站侧发送的注意力机制启用参数、注意力机制参数集合和网络参数集合；
    根据所述信道类型信息和所述注意力机制启用参数判断是否启用所述注意力机制；
    当启用所述注意力机制时，从所述注意力机制参数集合中选取与所述信道类型信息匹配的注意力机制参数，并根据所述注意力机制参数生成所述注意力机制；
    从所述网络参数集合中选取与所述信道类型信息匹配的第一网络参数集合；
    从所述第一网络参数集合中选取与所述码本类型信息和所述码本个数信息匹配的网络参数，并根据所述网络参数生成所述编码网络。
  5. 根据权利要求4所述的码本传输方法，其中，所述注意力机制参数集合包括所述注意力机制参数和所述注意力机制参数对应的注意力机制索引，所述网络参数集合包括所述网络参数和所述网络参数对应的网络索引；
    所述获取所述注意力机制参数和所述网络参数之后还包括：
    从所述注意力机制参数集合和所述网络参数集合中获取与所述注意力机制参数和所述网络参数对应的所述注意力机制索引和所述网络索引；
    将所述注意力机制索引和所述网络索引发送至所述基站侧。
  6. 一种码本传输方法，应用在基站侧，所述方法包括：
    接收终端侧发送信道的第二码本，其中，所述第二码本由所述终端侧根据所述信道的信道参数生成；
    利用预设的解码网络对所述第二码本进行解码处理生成第三码本；
    利用预设的注意力机制对所述第三码本进行特征提取；
    解析进行特征提取后的所述第三码本获取所述信道的信道参数，并根据所述信道参数生成所述信道的信道状态信息。
  7. 根据权利要求6所述的码本传输方法，其中，所述方法还包括：
    向所述终端侧发送注意力机制启用参数、注意力机制参数集合和网络参数集合，其中，所述注意力机制参数集合包括注意力机制参数和所述注意力机制参数对应的注意力机制索引，所述网络参数集合包括所述网络参数和网络参数对应的网络索引；
    接收所述终端侧根据所述注意力机制启用参数、所述注意力机制参数集合和所述网络参数集合返回的所述注意力机制索引和所述网络索引；
    从所述注意力机制参数集合中选取与所述注意力机制索引对应的所述注意力机制参数，并根据所述注意力机制参数生成所述注意力机制；
    从所述网络参数集合中选取与所述网络索引对应的所述网络参数，并根据所述网络参数生成所述解码网络。
  8. 一种终端，包括：
    获取模块，用于获取信道的信道参数；
    注意力机制模块，用于根据预设的注意力机制对所述信道参数进行特征提取处理，生成所述信道的第一码本；
    编码模块，用于根据预设的编码网络对所述第一码本进行编码处理，生成所述信道的第二码本；
    传输模块，用于发送所述第二码本至基站侧，其中，所述第二码本用于使所述基站侧生成所述信道的信道状态信息。
  9. 一种基站，包括：
    接收模块，用于接收终端侧发送信道的第二码本，其中，所述第二码本由所述终端侧根据所述信道的信道参数生成；
    解码模块，用于利用预设的解码网络对所述第二码本进行解码处理生成第三码本；
    注意力机制模块，用于利用预设的注意力机制对所述第三码本进行特征提取；
    解析模块，用于解析进行特征提取后的所述第三码本获取所述信道的信道参数，并根据所述信道参数生成所述信道的信道状态信息。
  10. 一种计算机可读存储介质，存储有计算机程序，所述计算机程序被处理器执行时实现权利要求1至5中任一项所述的码本传输方法，或权利要求6至7中任一项所述的码本传输方法。
PCT/CN2022/104950 2021-08-10 2022-07-11 码本传输方法、终端、基站和存储介质 WO2023016166A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110914141.4 2021-08-10
CN202110914141.4A CN115706595A (zh) 2021-08-10 2021-08-10 码本传输方法、终端、基站和存储介质

Publications (1)

Publication Number Publication Date
WO2023016166A1 true WO2023016166A1 (zh) 2023-02-16

Family

ID=85179526

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/104950 WO2023016166A1 (zh) 2021-08-10 2022-07-11 码本传输方法、终端、基站和存储介质

Country Status (2)

Country Link
CN (1) CN115706595A (zh)
WO (1) WO2023016166A1 (zh)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101931498A (zh) * 2010-07-06 2010-12-29 华为技术有限公司 一种传输信道状态信息的方法及装置
CN103378889A (zh) * 2012-04-24 2013-10-30 株式会社Ntt都科摩 码本生成方法、码本生成装置以及初始码本生成方法
US20140086285A1 (en) * 2012-09-27 2014-03-27 Nokia Siemens Networks Oy Non-codebook based channel state information feedback
CN104243003A (zh) * 2013-06-06 2014-12-24 电信科学技术研究院 一种基于码本的信道状态信息的传输方法和装置
CN110912598A (zh) * 2019-11-22 2020-03-24 中原工学院 基于长短时注意力机制的大规模mimo系统csi反馈方法
CN111555781A (zh) * 2020-04-27 2020-08-18 天津大学 一种基于深度学习注意力机制的大规模mimo信道状态信息压缩及重建方法
CN111756457A (zh) * 2019-03-27 2020-10-09 华为技术有限公司 信道预测方法、装置及计算机存储介质

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101931498A (zh) * 2010-07-06 2010-12-29 华为技术有限公司 一种传输信道状态信息的方法及装置
CN103378889A (zh) * 2012-04-24 2013-10-30 株式会社Ntt都科摩 码本生成方法、码本生成装置以及初始码本生成方法
US20140086285A1 (en) * 2012-09-27 2014-03-27 Nokia Siemens Networks Oy Non-codebook based channel state information feedback
CN104243003A (zh) * 2013-06-06 2014-12-24 电信科学技术研究院 一种基于码本的信道状态信息的传输方法和装置
CN111756457A (zh) * 2019-03-27 2020-10-09 华为技术有限公司 信道预测方法、装置及计算机存储介质
CN110912598A (zh) * 2019-11-22 2020-03-24 中原工学院 基于长短时注意力机制的大规模mimo系统csi反馈方法
CN111555781A (zh) * 2020-04-27 2020-08-18 天津大学 一种基于深度学习注意力机制的大规模mimo信道状态信息压缩及重建方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
CATT: "Codebook enhancements for DL MIMO", 3GPP DRAFT; R1-112117, 3RD GENERATION PARTNERSHIP PROJECT (3GPP), MOBILE COMPETENCE CENTRE ; 650, ROUTE DES LUCIOLES ; F-06921 SOPHIA-ANTIPOLIS CEDEX ; FRANCE, vol. RAN WG1, no. Athens, Greece; 20110822, 16 August 2011 (2011-08-16), Mobile Competence Centre ; 650, route des Lucioles ; F-06921 Sophia-Antipolis Cedex ; France , XP050537291 *

Also Published As

Publication number Publication date
CN115706595A (zh) 2023-02-17

Similar Documents

Publication Publication Date Title
US11632154B2 (en) Methods and apparatuses for feedback reporting in a wireless communications network
US20240129011A1 (en) Methods and Apparatuses for Enhancement on Basis Subset Indication for Two-Codebook Based CSI Reporting
US11923975B2 (en) Device and method for compressing and/or decompressing channel state information
EP2475126B1 (en) Method, terminal and base station for processing channel state information
US20230075037A1 (en) Channel state information feedback method, device, terminal, and network side, and storage medium
CN111030742A (zh) 在wlan通信/无线通信中由第一无线站执行的波束成形的方法
KR20160118086A (ko) 무선 통신 시스템에서 채널 정보를 피드백하기 위한 장치 및 방법
US20220385343A1 (en) Information transmission method and apparatus
CN111865377A (zh) 指示和确定预编码矩阵的方法以及通信装置
WO2019157709A1 (zh) 信息获取方法、装置、设备和存储介质
CN111757382B (zh) 指示信道状态信息的方法以及通信装置
CN112751598A (zh) 一种预编码矩阵的处理方法和通信装置
WO2023016166A1 (zh) 码本传输方法、终端、基站和存储介质
CN111865372B (zh) 一种用于构建预编码矩阵的系数指示方法和通信装置
KR20090115829A (ko) 무선통신 시스템에서의 채널 정보 전송 장치 및 방법
CN111010218B (zh) 指示和确定预编码向量的方法以及通信装置
CN115022896A (zh) 信息上报方法、装置、第一设备及第二设备
CN101834706A (zh) 一种信道信息的码本量化反馈方法及系统
CN111435851A (zh) 一种信道状态信息发送及接收方法、终端设备和网络设备
WO2023030538A1 (zh) 信道状态信息的处理方法、终端、基站、计算机可读存储介质
KR20240016872A (ko) 통신 시스템에서 채널 상태 정보 송수신 방법 및 장치
WO2023071683A1 (zh) 用于反馈信道状态的方法和装置
EP4318990A1 (en) Network parameter set information transmission method and apparatus, terminal, base station, and medium
CN117639863A (zh) 一种信道状态信息csi上报的方法和装置
WO2021074822A1 (en) Differential and quantized side information transmission for type ii csi

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22855149

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE