WO2019157709A1 - Information obtaining method and apparatus, device, and storage medium - Google Patents

Abstract

The present application provides an information obtaining method and apparatus, a device, and a storage medium. The method comprises: performing inverse discrete Fourier transform (IDFT) on frequency domain channel information to obtain time domain channel information; performing two-stage amplitude quantization and phase quantization on the time domain channel information to obtain channel state information; and sending the channel state information to a network device. By means of the transform from a frequency domain to a time domain, a terminal side enables channel state information to be sparser, such that the channel state information feedback amount can be reduced. In addition, two-stage amplitude quantization can improve the accuracy of channel feedback on the basis of increasing a little feedback amount.

Description

Information acquisition method, device, device and storage medium Technical field

The present application relates to the field of communications, and in particular, to an information acquisition method, apparatus, device, and storage medium.

Background technique

Massive multi-input multi-output (Massive MIMO) technology is one of the hot research technologies in the communication industry. Massive MIMO technology configures a large number of uniformly spaced rectangular antenna arrays (URAs) at the base station to achieve higher spatial freedom, thus enabling more users to use multi-user multiple input multiple outputs (multi The user multi-input multi-output (MU-MIMO) technology improves cell throughput and greatly improves the performance of the Massive MIMO cell.

In a frequency division duplexing (FDD) system, uplink and downlink channels use different frequency points, and there is no reciprocity between uplink and downlink channels. The base station cannot obtain higher-precision downlink channel information through the uplink channel. Therefore, in the current Long Term Evolution (LTE) system, the downlink channel information is fed back to the base station (evolved node B, eNB) by the user equipment (UE). However, due to the uplink resource limitation, the channel information fed back by the UE is code-quantized. Therefore, the channel information fed back by the UE is not completely equivalent to the actual channel information, and the base station can effectively eliminate the user between the users by using the channel information fed back by the UE. The interference does not improve the throughput of the cell. Therefore, for the FDD system, reporting the channel information with higher accuracy by the UE is one of the key factors for improving the performance of the Massive MIMO cell.

In the evolution process of the protocol, with the increase of the antenna port and the refinement of the space, the basic codebook is getting larger and larger, but due to the complexity of the actual channel, in the case of feeding back a single codebook, the basis of the refinement is only The granularity of the codebook is not sufficient to accurately describe the channel. Therefore, in the LTE R14 and New Radio (NR) protocols, the concept of linear merging is proposed, that is, linear weighting by two or more different base codebooks. Combine to describe channel information.

In the current version of the R14 version, according to the URA dual-polarized antenna configuration, the LTE ClassA codebook uses a discrete fourier transform (DFT) codebook combined with the Kronecker product codebook format to support 32port. Table 1 shows the codebook style when the codebook is configured to 1 (Codebook-Config = 1).

Table 1

among them,

f=i _1,1 is the horizontal codebook selection index, m=i _1,2 is the vertical codebook selection index, and n=i ₂ is the phase quantization index of the two polarization directions,

The selection vector for the base codebook of the broadband, P is the number of ports.

Further, in order to better eliminate interference between users, the LTE Advanced Codebook and the NR Type II codebook adopt a method of linearly combining multiple basic codebooks to achieve the purpose of approaching real channel information. The specific channel information feedback is composed as follows:

W=W ₁ *W ₂ (6)

among them,

Wherein W ₁ is a selection vector of a basic codebook of a wideband, and W _{1 is} based on a single-polarized oversampling codebook of the existing R14 Class A, that is, W ₁ is the above-mentioned v _f,m , L is a basic combination of linear combinations This number. W ₂ is a linear weighted quantization weighting coefficient,

Broadband amplitude information representing quantized weighting coefficients,

The sub-band amplitude information indicating the quantization weighting coefficient, c _i,j,k represents the sub-band phase information of the quantization weighting coefficient, the subscript i represents the polarization, and k represents the selected basic codebook. The maximum number of Ranks supported by the current protocol is 2.

It can be seen from the W ₂ matrix that the technical scheme 2 uses the amplitude two-level quantization in the weight quantization, and the phase first-level quantization scheme has certain similarity with our scheme, but the difference is the technical scheme II. The codebook of ₁ selects beam splitting based on frequency domain wideband,

The quantization is based on the wideband codebook correlation coefficient, 3bit quantization, and the entire bandwidth is fed back only one set.

Quantification is in

Based on the 1 bit fine adjustment, different Subband subcarriers will feedback according to their own channels. Considering the strong correlation between the wideband amplitude and the subband amplitude in the actual channel, a two-level quantization scheme is used between the wideband and the subband to reduce the amount of feedback. The reason why our algorithm uses two-level quantization is that the power difference between different time delays is very different. Using two-level quantization can improve the feedback accuracy of low-power time-delay in a certain feedback amount, and better feedback channel information. .

The LTE Advanced Codebook and the NR Type II codebook are used. Taking the 20M bandwidth and the 16port system as an example, the number of subbands specified by the protocol is 13, according to the 3GPP TS 38.214V1.2.0 protocol, for different Ranks, channels. The size of the feedback payload is shown in Table 2:

Table 2

Rank	L	Payload Size(bit)
1	4	347
2	4	683

It can be seen from Table 2 that the size of the Payload is positively correlated with the number of the basic codebooks of the feedback L. According to the LTE Advanced Codebook and the NR Type II codebook, if the channel information feedback accuracy is improved, it is necessary to increase L, which will increase Continue to increase the Payload Size, thereby increasing the overhead of reporting channel information by the UE.

Summary of the invention

The embodiment of the present invention provides a method, an apparatus, a device, and a storage medium for acquiring information, which can solve the problem of improving channel information feedback precision in the existing channel feedback mechanism, and increasing the overhead of reporting channel information by the terminal.

In a first aspect, an embodiment of the present application provides a method for acquiring information, including:

Performing discrete Fourier transform (IDFT) transform on the frequency domain channel information to obtain time domain channel information;

Performing amplitude two-stage quantization and phase quantization on the time domain channel information to acquire channel state information;

Transmitting the channel state information to a network device.

In the above solution, the terminal performs IDFT transformation on the frequency domain channel information, acquires time domain channel information, performs amplitude two-stage quantization and phase quantization on the time domain channel information, acquires channel state information, and transmits channel state information to the network device, and the terminal The frequency domain to time domain transform makes the channel state information more sparse, and the channel state information feedback amount can be reduced. Moreover, the amplitude two-level quantization can improve the channel feedback accuracy on the basis of increasing the feedback amount.

Optionally, performing the two-stage quantization and phase quantization on the time domain channel information to obtain channel state information, including:

Obtaining the power of the time domain channel information of each path, and ordering the powers in descending order;

And performing amplitude two-stage quantization and phase quantization on the time domain channel information corresponding to the first M powers, and acquiring the channel state information.

In the above solution, the terminal acquires the power of the time domain channel information of each path, and sorts the powers in descending order, and the terminal performs amplitude two-stage quantization and phase quantization on the time domain channel information corresponding to the first M powers. Obtaining channel state information can reduce the feedback amount of channel state information and save transmission resources.

Optionally, the acquiring the power of the time domain channel information of each path includes:

Obtaining the power of the time domain channel information of the kth _path according to formula (1)

Where k _path is a multipath index, i _port is a port index, and N _DFT is a point of a delay path.

Optionally, performing the two-stage quantization and phase quantization on the time domain channel information to obtain channel state information, including:

Determining L codebooks for time domain channel information of each path, and determining a codebook selection matrix according to an L codebook of time domain channel information of each path;

And performing correlation processing on the channel information in the first polarization direction of each path, the channel information in the second polarization direction, and the corresponding L codebooks, respectively, to obtain correlation coefficients of each path;

Obtaining a first weighting coefficient matrix according to the correlation coefficient of each path;

Performing amplitude two-stage quantization and phase quantization on the first weighting coefficient matrix to obtain a codebook weighting coefficient matrix;

And obtaining the channel state information according to the codebook selection matrix and the codebook weighting coefficient matrix.

Optionally, performing the two-level quantization of the weighting coefficient matrix, including:

Selecting, from the first weighting coefficient matrix, a first element having the largest weighting coefficient in each row to form a second weighting coefficient matrix;

Selecting, from the second matrix of weighting coefficients, a second element having the largest weighting coefficient;

And normalizing each element in the second weighting coefficient matrix by using the second element to obtain a third weighting coefficient matrix;

And quantizing the third weighting coefficient matrix by using the first quantization bit to obtain a first level amplitude quantization result;

And normalizing the correlation coefficient of each path according to the first-stage amplitude quantization result, and acquiring a normalized correlation coefficient of each path;

The normalized correlation coefficient of each path is quantized by using the second quantization bit to obtain a second-stage amplitude quantization result.

Optionally, performing phase quantization on the weighting coefficient matrix includes:

A phase quantization method is performed on each element in the first weighting coefficient matrix by using a preset phase modulation method to obtain a phase quantization matrix.

In the above solution, L codebooks are determined for the time domain channel information of each path, and the codebook selection matrix is determined according to the L codebook of the time domain channel information of each path, and the first polarization of each path is respectively determined. The channel information of the direction, the channel information of the second polarization direction, and the corresponding L codebooks are correlated, the correlation coefficient of each path is obtained, and the first weighting coefficient matrix is obtained according to the correlation coefficient of each path, from the first weighting In the coefficient matrix, the first element having the largest weighting coefficient in each row is selected to form a second weighting coefficient matrix, and the second element having the largest weighting coefficient is selected from the second weighting coefficient matrix, and the second element is used in the second weighting coefficient matrix. Each element is normalized to obtain a third weighting coefficient matrix, and the third weighting coefficient matrix is quantized by using the first quantization bit to obtain a first-stage amplitude quantization result, according to the first-level amplitude quantization result for each path The correlation coefficient is normalized, the normalized correlation coefficient of each path is obtained, and the normalized correlation coefficient of each path is quantized by the second quantization bit to obtain the second level Quantizing the result, using a preset phase modulation method, performing phase quantization on each element in the first weighting coefficient matrix, acquiring a phase quantization matrix, and acquiring channel state information according to the codebook selection matrix and the codebook weighting coefficient matrix, Compared with NR Type II, the IDFT time domain transform reduces the number of feedback bits and improves the compression efficiency without reducing the accuracy of channel information.

Optionally, the codebook selection matrix is

Where b is the codebook corresponding to the time domain channel information of each path, and M is the number of time delay paths;

The correlation coefficient of each path is ω _i =[ω _i,1 ,...,ω _i,L ,ω _i,L+1 ,...,ω _i,2L ] _2L*1 ; wherein i is the time delay path index of;

The first weighting coefficient matrix is

The second matrix of weighting coefficients is

Wherein the first element is

The second element is

The first stage amplitude quantization result is

The second-stage amplitude quantization result of the i-th path is p _2,i =[p _2,i,1 ...p _2,i,2L ] _1*2L ;

The phase quantization matrix is

The codebook weighting coefficient matrix is ω _quan =[p ₁ ] _M*1 .[p ₂ ] _M*2L .[Φ] _M*2L .

Optionally, performing the discrete Fourier transform (IDFT) transform on the frequency domain channel information to obtain time domain channel information, including:

Performing the IDFT transformation on the information of each of the transceiver antenna links in the frequency domain channel information H according to the formula (2), acquiring the time domain channel information Ψ;

Ψ(i _port ,:) ₌ IDFT(H(i _port ,:)),i _port ∈{1,...,N _port } (2)

Where i _port is the port index.

Optionally, before performing the discrete Fourier transform (IDFT) transform on the frequency domain channel information to obtain the time domain channel information, the method further includes:

Receiving pilot information sent by the network device;

Performing channel estimation according to the pilot information, and acquiring the frequency domain channel information.

In the above solution, the terminal receives the pilot information sent by the network device, and the terminal performs channel estimation according to the pilot information, acquires the frequency domain channel information, and can obtain the frequency domain channel information more accurately, thereby ensuring the feedback mechanism of the channel state information. Reliability.

Optionally, if the antenna of the terminal device is configured as the at least two receiving links, the performing channel estimation according to the pilot information, acquiring the frequency domain channel information, including:

Obtaining channel feedback information of each resource block according to the pilot information;

Performing singular value decomposition on the channel feedback information of each resource block, obtaining corresponding feature vectors, and ordering the feature vectors according to the order of the feature values from large to small;

Determining a rank of the channel state information;

If the rank of the channel state information is 1, the channel information of at least two of the receiving links is integrated into the frequency domain channel information according to a feature vector with the largest feature value;

If the rank of the channel state information is 2, the channel information of the at least two receiving links is integrated into the first frequency domain channel information and the second frequency domain channel information, respectively, according to the feature vectors corresponding to the first two feature values. .

In the above solution, the terminal acquires channel feedback information of each resource block according to the pilot information, performs SVD on the channel feedback information of each resource block, obtains a corresponding feature vector, and pairs the features according to the feature values from large to small. Vector ordering, determining the rank of the channel state information; if the rank of the channel state information is 1, the channel information of at least two receiving links is integrated into the frequency domain channel information according to the feature vector having the largest eigenvalue; if the rank of the channel state information 2, the channel information of the at least two receiving links is integrated into the first frequency domain channel information and the second frequency domain channel information according to the feature vectors corresponding to the first two feature values, and the rank of the channel state information may be accurately The frequency domain channel information is obtained to report more accurate channel state information.

In a second aspect, the embodiment of the present application provides a method for acquiring information, including:

Receiving channel state information sent by the terminal;

Performing channel reconstruction according to the channel state information to acquire time domain channel information;

Performing a discrete Fourier transform on the time domain channel information to obtain frequency domain channel information.

Optionally, performing channel reconstruction according to the channel state information to obtain time domain channel information includes:

Zeroing the weights of the delay paths other than the delay paths corresponding to k ₁ , . . . , k _M , and performing channel reconstruction on the M time delay paths according to formula (3) to obtain the time domain channel information. H _time ;

Where B _i is the ith row of the codebook selection matrix, ω _{quan_i} is the ith row of the codebook weighting coefficient matrix, k _{i is the} time domain multipath delay indication, N _port is the port index, and L is the path of each path. The number of codebooks corresponding to the time domain channel information;

The codebook selection matrix is

The codebook weighting coefficient matrix is ω _quan =[p ₁ ] _M*1 .[p ₂ ] _M*2L .[Φ] _M*2L ;

P ₁ is the first-order amplitude quantization result,

p ₂ is the second-stage amplitude quantization result, and the second-stage amplitude quantization result of the ith path is p _2,i =[p _2,i,1 ...p _2,i,2L ] _1*2L ;

Φ is the phase quantization matrix,

Optionally, before the receiving the channel state information sent by the terminal, the method further includes:

The pilot information is transmitted to the terminal to cause the terminal to perform channel estimation according to the pilot information.

Optionally, performing channel reconstruction according to the channel state information to obtain time domain channel information includes:

Determining a rank of the channel state information;

If the rank of the channel state information is 1, channel reconstruction is performed according to formula (4), and time domain channel information is acquired.

If the rank of the channel state information is 2, channel reconstruction is performed according to formula (5), and time domain channel information is acquired.

among them,

Yes

The i-th column.

For the implementation of the apparatus and the beneficial effects of the second aspect, reference may be made to the description of the first aspect, and details are not described herein again.

In a third aspect, an embodiment of the present application provides an information acquiring apparatus, including:

a transform module, configured to perform inverse discrete Fourier transform IDFT transform on frequency domain channel information, to obtain time domain channel information;

a quantization module, configured to perform amplitude two-stage quantization and phase quantization on the time domain channel information, and acquire channel state information;

And a sending module, configured to send the channel state information to the network device.

Optionally, the channel state information includes time domain multipath delay indication information, codebook selection indication information, and codebook weighting coefficient information.

Optionally, the determining module is specifically configured to obtain power of time domain channel information of each path, and sort the powers in descending order; and perform amplitude on time domain channel information corresponding to the first M powers. Two-stage quantization and phase quantization acquire the channel state information.

Optionally, the determining, by the metric module, the power of the time domain channel information of each path, including:

The quantization module acquires the power of the time domain channel information of the kth _path according to formula (1)

Where k _path is a multipath index, i _port is a port index, and N _DFT is a point of a delay path.

Optionally, the quantization module performs amplitude two-stage quantization and phase quantization on the time domain channel information, and acquires channel state information, including:

The quantization module determines L codebooks for the time domain channel information of each path, and determines a codebook selection matrix according to the L codebook of the time domain channel information of each path; respectively, the first polarization direction of each path The channel information, the channel information of the second polarization direction, and the corresponding L codebooks are correlated to obtain a correlation coefficient of each path; and the first weighting coefficient matrix is obtained according to the correlation coefficient of each path And performing amplitude two-stage quantization and phase quantization on the first weighting coefficient matrix to obtain a codebook weighting coefficient matrix; and acquiring the channel state information according to the codebook selection matrix and the codebook weighting coefficient matrix.

Optionally, the quantization module performs amplitude two-level quantization on the weighting coefficient matrix, including:

The quantization module selects, from the first weighting coefficient matrix, a first element having the largest weighting coefficient in each row to form a second weighting coefficient matrix; and selecting a second element having the largest weighting coefficient from the second weighting coefficient matrix And normalizing each element in the second weighting coefficient matrix by using the second element to obtain a third weighting coefficient matrix; and using the first quantization bit to quantize the third weighting coefficient matrix to obtain the first Level 1 amplitude quantization result; normalizing the correlation coefficient of each path according to the first level amplitude quantization result, obtaining a normalized correlation coefficient of each path; using the second quantization bit pair for each The normalized correlation coefficient of the path is quantized to obtain the second-level amplitude quantization result.

Optionally, the quantizing module performs phase quantization on the weighting coefficient matrix, including:

The quantization module performs phase quantization on each element in the first weighting coefficient matrix by using a preset phase modulation method to obtain a phase quantization matrix.

Optionally, the codebook selection matrix is

Where b is the codebook corresponding to the time domain channel information of each path, and M is the number of time delay paths;

The correlation coefficient of each path is ω _i =[ω _i,1 ,...,ω _i,L ,ω _i,L+1 ,...,ω _i,2L ] _2L*1 ; wherein i is the time delay path index of;

The first weighting coefficient matrix is

The second matrix of weighting coefficients is

Wherein the first element is

The second element is

The first stage amplitude quantization result is

The second-stage amplitude quantization result of the i-th path is p _2,i =[p _2,i,1 ...p _2,i,2L ] _1*2L ;

The phase quantization matrix is

The codebook weighting coefficient matrix is ω _quan =[p ₁ ] _M*1 .[p ₂ ] _M*2L .[Φ] _M*2L .

Optionally, the transform module is specifically configured to perform the IDFT transform on the information of each of the transmit and receive antenna links in the frequency domain channel information H according to formula (2), to obtain the time domain channel information Ψ;

Ψ(i _port ,:)=IDFT(H(i _port ,:)),i _port ∈{1,...,N _port } (2)

Where i _port is the port index.

Optionally, the device further includes:

a receiving module, configured to receive pilot information sent by a network device;

And an acquiring module, configured to perform channel estimation according to the pilot information, and acquire the frequency domain channel information.

Optionally, if the antenna of the terminal device is configured as at least two receiving links, the acquiring module is specifically configured to acquire channel feedback information of each resource block according to the pilot information, and provide channel feedback information for each resource block. Performing singular value decomposition, obtaining corresponding feature vectors, and ordering the feature vectors according to the order of the feature values; determining the rank of the channel state information; if the rank of the channel state information is 1, according to The feature vector having the largest eigenvalue integrates channel information of at least two of the receiving links into the frequency domain channel information; if the rank of the channel state information is 2, the eigenvectors corresponding to the first two eigenvalues respectively Channel information of at least two of the receiving links is integrated into first frequency domain channel information and second frequency domain channel information.

The implementation principle and beneficial effects of the device provided by the third aspect can be referred to the description of the first aspect, and details are not described herein again.

In a fourth aspect, an embodiment of the present application provides an information acquiring apparatus, including:

a receiving module, configured to receive channel state information sent by the terminal;

An acquiring module, configured to perform channel reconstruction according to the channel state information, and acquire time domain channel information;

And a transform module, configured to perform discrete Fourier transform on the time domain channel information to obtain frequency domain channel information.

Optionally, the channel state information includes time domain multipath delay indication information, codebook selection indication information, and codebook weighting coefficient information.

Optionally, the acquiring module is specifically configured to perform zero-padding on the weights of the delay paths other than the delay paths corresponding to k ₁ , . . . , k _M , and respectively perform M delay paths according to formula (3). Channel reconstruction, acquiring the time domain channel information H _time ;

Where B _i is the ith row of the codebook selection matrix, ω _{quan_i} is the ith row of the codebook weighting coefficient matrix, k _{i is the} time domain multipath delay indication, N _port is the port index, and L is the path of each path. The number of codebooks corresponding to the time domain channel information;

The codebook selection matrix is

The codebook weighting coefficient matrix is ω _quan =[p ₁ ] _M*1 .[p ₂ ] _M*2L .[Φ] _M*2L ;

P ₁ is the first-order amplitude quantization result,

p ₂ is the second-stage amplitude quantization result, and the second-stage amplitude quantization result of the ith path is p _2,i =[p _2,i,1 ...p _2,i,2L ] _1*2L ;

Φ is the phase quantization matrix,

Optionally, the device further includes:

And a sending module, configured to send pilot information to the terminal, so that the terminal performs channel estimation according to the pilot information.

Optionally, the acquiring module is specifically configured to determine a rank of the channel state information; if the rank of the channel state information is 1, perform channel reconstruction according to formula (4), and acquire time domain channel information.

If the rank of the channel state information is 2, channel reconstruction is performed according to formula (5), and time domain channel information is acquired.

among them,

Yes

The i-th column.

For the implementation of the apparatus and the beneficial effects of the fourth aspect, reference may be made to the description of the first aspect, and details are not described herein again.

Optionally, in any one of the foregoing first to fourth aspects, the channel state information includes time domain multipath delay indication information, codebook selection indication information, and codebook weighting coefficient information.

In a fifth aspect, an embodiment of the present application provides a terminal, a processor, and a memory, where

The memory is for storing instructions for executing the memory stored instructions, the terminal for performing the method of any of the first aspects when the processor executes the instructions stored by the memory.

In a sixth aspect, an embodiment of the present application provides a network device, including: a processor and a memory, where

The memory is configured to store an instruction, the processor is configured to execute the memory stored instruction, and when the processor executes the memory stored instruction, the network device is configured to perform the method of any one of the second aspect .

In a seventh aspect, the embodiment of the present application provides a computer readable storage medium, where a computer program is stored thereon, and when the program is executed by the processor, the steps of the information acquiring method according to any one of the first aspect or the second aspect are implemented. .

DRAWINGS

FIG. 1 is a schematic diagram of an application scenario of an information acquiring method according to an embodiment of the present disclosure;

2 is an interaction flowchart of an information acquisition method according to an embodiment of the present invention;

FIG. 3 is a flowchart of a method for acquiring information according to an embodiment of the present invention;

FIG. 4 is a flowchart of a method for acquiring information according to another embodiment of the present invention;

FIG. 5 is a flowchart of a method for acquiring information according to another embodiment of the present invention;

FIG. 6 is a flowchart of a method for acquiring information according to another embodiment of the present invention;

FIG. 7 is a flowchart of a method for acquiring information according to another embodiment of the present invention;

FIG. 8 is a flowchart of a method for acquiring information according to another embodiment of the present invention;

FIG. 9 is a block diagram of an information acquiring apparatus according to an embodiment of the present application;

FIG. 10 is a block diagram of an information acquiring apparatus according to another embodiment of the present disclosure;

FIG. 11 is a block diagram of an information acquiring apparatus according to another embodiment of the present application;

FIG. 12 is a block diagram of an information acquiring apparatus according to another embodiment of the present disclosure;

FIG. 13 is a block diagram of a communication device according to an embodiment of the present application.

Detailed ways

FIG. 1 is a schematic diagram of an application scenario of an information acquiring method according to an embodiment of the present disclosure, where the application scenario includes a network device 1 and a terminal 2. In the embodiment of the present application, the terminal is also called a User Equipment (UE), which is a device that provides voice and/or data connectivity to the user, for example, a handheld device with an wireless connection function, and an in-vehicle device. Wait. Common terminals include, for example, mobile phones, tablets, notebook computers, PDAs, mobile internet devices (MIDs), wearable devices such as smart watches, smart bracelets, pedometers, and the like. A network device, also known as a radio access network (RAN) device, is a device that accesses a terminal to a wireless network, including but not limited to: an evolved Node B (eNB), and a wireless network control. Radio network controller (RNC), Node B (Node B, NB), Base Station Controller (BSC), Base Transceiver Station (BTS), home base station (for example, Home evolved NodeB, or Home Node B, HNB), BaseBand Unit (BBU). In addition, a Wifi Access Point (AP) or the like may also be included.

2 is an interaction flowchart of an information acquisition method according to an embodiment of the present invention. As shown in FIG. 2, the method includes the following steps:

Step 101: The terminal performs inverse Fourier Transform (IDFT) transform on the frequency domain channel information to obtain time domain channel information.

In this embodiment, the terminal may perform channel estimation, acquire frequency domain channel information of each resource block, perform IDFT transform on the frequency domain channel information, transform the frequency domain into a sparse time domain, and then extract time domain channel information.

Optionally, in this embodiment, the channel state information includes time domain multipath delay indication information, codebook selection indication information, and codebook weighting coefficient information. The time domain multipath delay indication information is used to indicate the selection of the delay path, the codebook selection indication information is used to indicate the selected codebook, and the codebook weighting coefficient information is used to indicate the first level of the amplitude quantization information, the second level. Amplitude quantization information and phase quantization information. The time domain multipath delay indication information, the codebook selection indication information, and the codebook weighting coefficient information may be in the form of a matrix.

Optionally, the network device may also send some configuration parameters to the terminal in advance. On the basis of the embodiment shown in FIG. 2, as shown in FIG. 3, before step 101, step 201 and step 202 may be further included:

Step 201: The terminal receives pilot information sent by the network device.

In this embodiment, the network device may add a channel state information (CSI) feedback mode. For example, a possible pilot information sent by the network device to the terminal is shown in Table 3:

table 3

RB number N RB	100
Number of horizontal ports N 1	8
Number of vertical ports N 2	2
Number of ports N port	2N 1 N 2
Horizontal port oversampling multiple O 1	4
Vertical port oversampling multiple O 2	4
Time-frequency conversion sampling point number N DFT	N RB
Feedback time domain points N path	M
Feedback code number of points N beam	L
First level amplitude quantization bit P 1	2
Second stage amplitude quantization bit P 2	3
Phase quantization bit Φ	3

In this embodiment, the network device may send the pilot information to the terminal after establishing a connection with the terminal, or the network device may send the pilot information after receiving the request message for requesting the pilot information sent by the terminal. The network device may also send the pilot information to the terminal periodically, or the network device may send the updated pilot information to the terminal after each update of the pilot information.

Step 202: The terminal performs channel estimation according to the pilot information, and acquires frequency domain channel information.

In this embodiment, after receiving the pilot information sent by the network device, the terminal performs channel estimation according to the pilot information, and acquires frequency domain channel information.

In the method provided by the embodiment, the terminal receives the pilot information sent by the network device, and the terminal performs channel estimation according to the pilot information to obtain the frequency domain channel information, so that the frequency domain channel information can be acquired more accurately, thereby ensuring feedback of the channel state information. The reliability of the mechanism.

Exemplarily, in an LTE system, the system bandwidth is 20M, the number of Resource Blocks (RBs) is N _RB = 100, and the base station side configures the dual-polarized antenna 32 to transmit 32, and the number of pilot ports is N _TX = 32. The antennas are arranged in 2 rows and 8 columns, and the UE is configured to receive 1 transmission and reception. The terminal can perform IDFT transformation on the information of each transceiver antenna link in the frequency domain channel information according to formula (2) to obtain time domain channel information. ;

Ψ(i _port ,:)=IDFT(H(i _port ,:)),i _port ∈{1,...,N _port } (2)

Wherein, Ψ (i _port, :) represents i _port downlink channel information to the time domain, i _port is a port index, N _port is the number of transmission antennas of the network device, in this embodiment, the dimensions of channel information in the time domain 32Port *100RB.

Exemplarily, in an LTE system, the system bandwidth is 20M, the number of Resource Blocks (RBs) is N _RB = 100, and the base station side configures the dual-polarized antenna 32 to transmit 32, and the number of pilot ports is N _TX = 32. The antennas are arranged in two rows and eight columns. Different from the above embodiment, if the UE is configured with at least two receiving links, the UE obtains the frequency domain channel information obtained by the channel estimation after receiving the CSI pilot information. The dimension is 2Ant*32Port*100RB. As shown in FIG. 4, the step “the terminal performs channel estimation according to the pilot information to obtain frequency domain channel information” includes:

Step 301: Acquire channel feedback information of each resource block according to the pilot information.

Step 302: Perform Singular Value Decomposition (SVD) on the channel feedback information of each resource block, obtain corresponding feature vectors, and sort the feature vectors according to the order of the feature values from largest to smallest.

Step 303: Determine the rank of the channel state information. If the rank of the channel state information is 1, perform step 304. If the rank of the channel state information is 2, perform step 305.

Step 304: Integrate channel information of at least two receiving links into frequency domain channel information according to a feature vector with the largest feature value.

In this embodiment, when the rank of the channel state information is 1, the terminal first performs SVD on the channel feedback information of each RB, selects the largest eigenvalue vector, and integrates the channel information of the two receiving antennas according to the largest eigenvector. Frequency domain channel information H', dimension is 32Port*100RB,

Step 305: Integrate channel information of at least two receiving links into first frequency domain channel information and second frequency domain channel information according to feature vectors corresponding to the first two feature values.

In this embodiment, when the rank of the channel state information is 2, the terminal performs SVD on the channel feedback information of the two receiving antennas, and selects the largest and second largest eigenvalue vectors to be integrated into the first frequency domain channel information H' and The second frequency domain channel information H" has a dimension of 32Port*100RB, and then decomposes and quantifies the H' and H" respectively.

In the information acquisition method provided by the embodiment, the terminal acquires channel feedback information of each resource block according to the pilot information, performs SVD on the channel feedback information of each resource block, obtains a corresponding feature vector, and according to the feature value from large to small. The order of the eigenvectors is sorted to determine the rank of the channel state information; if the rank of the channel state information is 1, the channel information of at least two receiving links is integrated into the frequency domain channel information according to the eigenvector with the largest eigenvalue; If the rank of the state information is 2, the channel information of the at least two receiving links is integrated into the first frequency domain channel information and the second frequency domain channel information according to the feature vectors corresponding to the first two eigenvalues, respectively, according to the channel state information. The rank accuracy of the frequency domain channel information is obtained in order to report more accurate channel state information.

Step 102: The terminal performs amplitude two-stage quantization and phase quantization on the time domain channel information to obtain channel state information.

In this embodiment, since the channel power difference is relatively large between different time delay paths, in order to more accurately quantize the codebook weighting coefficient information, the amplitude of the time domain channel information is used in this embodiment by a two-stage quantization method.

Optionally, as shown in FIG. 5, the step “the terminal performs amplitude two-stage quantization and phase quantization on the time domain channel information to obtain channel state information” includes steps 401 and 402:

Step 401: The terminal acquires the power of the time domain channel information of each path, and sorts the powers in descending order.

In this embodiment, the terminal may acquire the power of the time domain channel information of the kth _path according to formula (1).

Where k _path is the delay path index, i _port is the port index, and N _DFT is the number of points of the delay path.

Obtaining the power of time domain channel information for each time delay path

After that, according to the order from big to small, the power is put in order.

Step 402: The terminal performs amplitude two-stage quantization and phase quantization on the time domain channel information corresponding to the first M powers, and acquires channel state information.

In this embodiment, the terminal selects the first M powers, performs amplitude quantization and phase quantization on the time domain channel information corresponding to the selected first M powers, and acquires channel state information.

It should be noted that, in this embodiment, only the power of the time domain channel information is used.

Sorting in descending order, does not affect the ordering of time domain channel information of each time delay path, that is, the order of the time domain channel information corresponding to the selected first M powers is still the original order, and may be The index of the time domain channel information corresponding to the selected first M powers is labeled K={k ₁ , . . . , k _M }.

In this embodiment, the terminal acquires the power of the time domain channel information of each path, and sorts the powers in descending order, and the terminal performs amplitude two-stage quantization and phase on the time domain channel information corresponding to the first M powers. Quantization and acquisition of channel state information can reduce the feedback amount of channel state information and save transmission resources.

Further, taking FIG. 6 as an example, the implementation of the step of “amplifying two-stage quantization and phase quantization of the time domain channel information to obtain channel state information” is described in detail. As shown in FIG. 6 , the method includes the following steps:

Step 501: Determine L codebooks for time domain channel information of each path, and determine a codebook selection matrix according to an L codebook of time domain channel information of each path.

In the present embodiment, each column of this time-domain of the k _i of the M channel information path delay information

Decompose, using R14ClassA's single-polarization oversampled Discrete Fourier Transform (DFT) codebook, from which L codebooks that best match the channel of each delay path are selected for each time delay path. [b _i,1 ... b _i,L ], where i is the index of the time delay path. Traversing the time domain channel information of all delay paths to obtain a codebook selection matrix

Where L is the number of codebooks corresponding to each time delay path, and M is the number of time delay paths. b is the codebook corresponding to the time domain channel information of each path, and b is a complex number.

Step 502: Perform correlation processing on the channel information in the first polarization direction of each path, the channel information in the second polarization direction, and the corresponding L codebooks, respectively, to obtain correlation coefficients of each path.

For example, in an LTE system, the system bandwidth is 20M, N _RB = 100, the base station side is configured with dual-polarized antennas 32, 32 receives, the number of pilot ports is N _TX = 32, and the antennas are arranged in 2 rows and 8 columns, and the UE is configured. 1 × 1, because the code that best matches the channel selected for each latency path is present unipolar, so for the selected M time-domain channel information Ψ (:, k _i) the first column k _i Channel information of the first polarization direction (polarization direction 0)

And channel information of the second polarization direction (polarization direction 1)

Correlate with the selected L codebook [b _i,1 ... b _i,L ] and obtain the correlation coefficient ω _i =[ω _i,1 ,...,ω _i,L ,ω for each time delay path _{i, L+1} ,...,ω _i,2L ] _2L*1 .

Step 503: Acquire a first weighting coefficient matrix according to a correlation coefficient of each path.

In this embodiment, the time domain channel information of all the delay paths is traversed to obtain the first weighting coefficient matrix.

Among them, the elements in ω are all plural.

Step 504: Perform amplitude two-stage quantization and phase quantization on the first weighting coefficient matrix to obtain a codebook weighting coefficient matrix.

In this embodiment, since the channel power difference is relatively large between different time delay paths, in order to more accurately quantize the first weighting coefficient matrix ω, we perform the first level on the element amplitude in the first weighting coefficient matrix ω. Amplitude quantization and second quarter amplitude quantification.

Optionally, as shown in FIG. 7, the step of performing two-stage quantization on the first weighting coefficient matrix includes:

The first level of amplitude quantization includes steps 601 - 604, and the second level of amplitude quantization includes steps 605 and 606.

Step 601: Select, from the first weighting coefficient matrix, a first element having the largest weighting coefficient in each row to form a second weighting coefficient matrix.

In this embodiment, for the M time delay paths, the first element with the largest weighting coefficient in each time delay path is selected.

Obtaining a second matrix of weighting coefficients

Step 602: Select a second element having the largest weighting coefficient from the second weighting coefficient matrix.

In this embodiment, from the second weighting coefficient matrix

Select the largest second element

among them,

The elements in are all real numbers.

Step 603: Normalize each element in the second weighting coefficient matrix by using the second element to obtain a third weighting coefficient matrix.

In this embodiment, the second element is adopted

For each element in the second matrix of weighting coefficients

Normalization is performed to obtain a third matrix of weighting coefficients such that each element in the matrix of third weighting coefficients is less than or equal to one.

Step 604: Quantize the third weighting coefficient matrix by using the first quantization bit to obtain a first-level amplitude quantization result.

In this embodiment, the first quantization bit is used to indicate a quantization mode, and the first quantization bit may be predefined according to actual needs, for example, the first quantization bit is predefined to be 2 or 3. If the first quantization bit Q ₁ = 2 bits, the third quantization coefficient matrix is quantized by using the first quantization bit, and the first-stage amplitude quantization result is obtained as

Among them, the elements in p ₁ are all real numbers.

In this embodiment, in order to further reduce the feedback amount of the channel feedback information, the terminal and the network device may pre-negotiate the first-level amplitude quantization result p ₁ and the time domain multi-path delay indication information.

The corresponding relationship between the terminals only needs to report the time domain multipath delay indication information.

Just fine. Table 4 shows a first-stage amplitude quantization correspondence table in which the quantization bits are 2 bits.

Table 4

Step 605: Normalize the correlation coefficient of each path according to the first-stage amplitude quantization result, and obtain a normalized correlation coefficient of each path.

In this embodiment, according to the first-level quantization result, for all correlation coefficients in each path

Normalize and obtain the normalized correlation coefficient for each path.

Step 606: Quantize the normalized correlation coefficient of each path by using the second quantization bit to obtain a second-level amplitude quantization result.

In this embodiment, the second quantization bit is used to indicate another quantization mode, and the second quantization bit may be predefined according to actual needs, for example, the first quantization bit is defined as 2, the second quantization bit is 3, or The first quantization bit is defined to be 3, and the second quantization bit is 2.

If the second quantization bit is Q ₂ =3 bits, the normalized correlation coefficient of each path is quantized by using the second quantization bit Q ₂ , and the obtained second-order quantization result of the ith path is p _2,i =[p _2,i,1 ...p _2,i,2L ] _1*2L , then the second-order quantization result corresponding to the M time delay diameter is

Wherein, the elements in p ₂ are all real numbers.

In this embodiment, in order to further reduce the feedback amount of the channel feedback information, the terminal and the network device may pre-negotiate the second-level amplitude quantization result p _2,i and the time domain multipath delay indication information.

The corresponding relationship between the terminals only needs to report the time domain multipath delay indication information.

That is, where i is the index of the time delay path, 1 ≤ i ≤ M, j is the codebook index, and 1 ≤ j ≤ 2L. Table 5 shows a second-level amplitude quantization correspondence table in which the quantization bits are 3 bits.

table 5

Optionally, the step of “phase-quantizing the first weighting coefficient matrix” includes performing phase quantization on each element in the first weighting coefficient matrix by using a preset phase modulation method to obtain a phase quantization matrix.

In this embodiment, the preset phase modulation method may be 8 Phase Shift Keying (8PSK), and the terminal may adopt 8PSK to phase quantize all M*2L elements in the first weighting coefficient matrix ω. , get the phase quantization result as

Among them, the elements in Φ are all complex numbers.

Step 607: Acquire channel state information according to the codebook selection matrix and the codebook weighting coefficient matrix.

In this embodiment, in the channel state information reported by the terminal to the network device, the codebook selection indication information includes a codebook selection matrix B, and the codebook weighting coefficient information includes a codebook weighting coefficient matrix ω _quan , and a time domain multipath delay indication. The information includes the multipath delay index K.

Codebook selection matrix

Wherein, the elements in B are plural;

The first-order amplitude quantization result is the first-order quantization matrix of the codebook amplitude.

Wherein the elements in p ₁ are real numbers;

The second-order quantization result is the second-order quantization matrix of the codebook amplitude.

Wherein, the elements in p ₂ are real numbers;

Code phase quantization matrix

Wherein, the elements in Φ are all plural;

Codebook weighting coefficient matrix ω _quan =[p ₁ ] _M*1 .[p ₂ ] _M*2L .[Φ] _M*2L ;

The time domain multipath index K = {k ₁ , ..., k _M }.

The information obtaining method provided in this embodiment determines L codebooks for the time domain channel information of each path, and determines a codebook selection matrix according to the L codebook of the time domain channel information of each path, respectively for each path. The channel information in the first polarization direction, the channel information in the second polarization direction, and the corresponding L codebooks are correlated, the correlation coefficients of each path are obtained, and the first weighting coefficient matrix is obtained according to the correlation coefficient of each path. Selecting, from the first weighting coefficient matrix, a first element having the largest weighting coefficient in each row to form a second weighting coefficient matrix, selecting a second element having the largest weighting coefficient from the second weighting coefficient matrix, and using the second element to the second element Each element in the weighting coefficient matrix is normalized to obtain a third weighting coefficient matrix, and the third weighting coefficient matrix is quantized by using the first quantization bit to obtain a first-level amplitude quantization result, according to the first-level amplitude quantization result pair The correlation coefficient of each path is normalized, the normalized correlation coefficient of each path is obtained, and the normalized correlation coefficient of each path is quantized by the second quantization bit. Obtaining a second-stage amplitude quantization result, performing phase quantization on each element in the first weighting coefficient matrix by using a preset phase modulation method, acquiring a phase quantization matrix, and acquiring a channel according to the codebook selection matrix and the codebook weighting coefficient matrix The state information, because of the IDFT time domain transform, reduces the number of feedback bits and improves the compression efficiency without reducing the accuracy of the channel information compared to the NR Type II.

Step 103: The terminal sends the channel state information to the network device.

In this embodiment, the channel state information fed back by the UE may include the following information:

Time domain multipath delay indication information

Codebook selection indication

Codebook amplitude first order quantization matrix

Codebook amplitude secondary quantization matrix

Codebook phase quantization matrix

Step 104: The network device receives channel state information sent by the terminal.

In this embodiment, the channel state information is the same as the channel state information in the foregoing embodiment, and details are not described herein again.

Step 105: The network device performs channel reconstruction according to channel state information, and acquires time domain channel information.

In this embodiment, after receiving the channel state information, the network device performs channel reconstruction on each time delay path according to the channel state information in the time domain, and acquires time domain channel information.

Optionally, the step of: performing channel reconstruction according to channel state information to obtain time domain channel information includes: zeroing the weight of the delay path other than the delay path corresponding to k ₁ , . . . , k _M , and according to Equation (3) performs channel reconstruction on M time delay paths respectively, and acquires time domain channel information H _time (k _i );

Where B _i is the ith row of the codebook selection matrix, ω _{quan_i} is the ith row of the codebook weighting coefficient matrix, k _{i is the} time domain multipath delay indication, N _port is the port index, and L is the path of each path. The number of codebooks corresponding to the time domain channel information;

The codebook selection matrix is

The codebook weighting coefficient matrix is ω _quan =[p ₁ ] _M*1 .[p ₂ ] _M*2L .[Φ] _M*2L ;

P ₁ is the first-order amplitude quantization result,

p ₂ is the second-order amplitude quantization result, and the second-order amplitude quantization result of the ith path is p _{2, i} = [p _{2, i, 1} ... p _{2, i, 2L} ] _1*2L ;

Φ is the phase quantization matrix,

Optionally, if the terminal is configured as one transmit link and two receive links, as shown in FIG. 8, the step of performing channel reconstruction according to channel state information to obtain time domain channel information includes:

Step 701: Determine the rank of the channel state information. If the rank of the channel state information is 1, perform step 702. If the rank of the channel state information is 2, perform step 703.

Step 702: Perform channel reconstruction according to formula (4), and obtain time domain channel information.

In this embodiment, when the rank of the channel state information is 1, B ⁽¹⁾ is a codebook selection matrix, P ₁ ⁽¹⁾ is a codebook amplitude first-order quantization matrix, and P ₂ ⁽¹⁾ is a codebook amplitude. The secondary quantization matrix, Φ ⁽¹⁾ is the codebook phase quantization matrix, N _port is the number of transmission antennas of the network device, and M is the number of delay paths, wherein the values of the parameters can be referred to the terminal side embodiment, where No longer.

Step 703: Perform channel reconstruction according to formula (5), and obtain time domain channel information.

among them,

Yes

The i-th column.

The embodiment shown in FIG. 8 is different from the embodiment shown in the formula (3) in that the time domain channel information of the delay path corresponding to k ₁ , . . . , k _M is obtained in the embodiment shown in FIG. 8 , and k is not considered. _1, ..., k _M delay outside diameter, and the equation (3) in the embodiment shown, need to be K _1, ..., weights other than the path delay corresponding to k _M zero padding delay path.

Step 106: The network device performs discrete Fourier transform on the time domain channel information to obtain frequency domain channel information.

In this embodiment, the time domain channel information H _{time is} subjected to row DFT transform to obtain frequency domain channel information.

The frequency domain channel information is the downlink channel information received by the network device end after the terminal quantization feedback.

In the information acquisition method provided by the embodiment of the present application, the terminal performs IDFT transformation on the frequency domain channel information, acquires time domain channel information, performs amplitude two-stage quantization and phase quantization on the time domain channel information, acquires channel state information, and transmits channel state information. For the network device, the network device performs channel reconstruction according to the channel state information, acquires time domain channel information, performs discrete Fourier transform on the time domain channel information, acquires frequency domain channel information, and performs frequency domain to time domain transformation on the terminal side. The channel state information is made more sparse, the channel state information feedback amount can be reduced, and the amplitude of the channel feedback can be improved by increasing the amount of feedback by the two-level quantization of the amplitude.

FIG. 9 is a block diagram of an information acquiring apparatus according to an embodiment of the present disclosure. As shown in FIG. 9, the apparatus includes:

The transform module 11 is configured to perform inverse discrete Fourier transform IDFT transform on the frequency domain channel information to obtain time domain channel information.

The quantization module 12 is configured to perform amplitude two-stage quantization and phase quantization on the time domain channel information to obtain channel state information.

The sending module 13 is configured to send channel state information to the network device.

Optionally, the channel state information includes time domain multipath delay indication information, codebook selection indication information, and codebook weighting coefficient information.

Optionally, the quantization module 12 is specifically configured to obtain power of time domain channel information of each path, and order power according to a sequence from largest to smallest; and perform amplitude two-level quantization on time domain channel information corresponding to the first M powers. And phase quantization to obtain channel state information.

Optionally, the quantifying module 12 obtains the power of the time domain channel information of each path, including:

The quantization module 12 obtains the power of the time domain channel information of the kth _path according to formula (1).

Where k _path is a multipath index, i _port is a port index, and N _DFT is a point of a delay path.

Optionally, the quantization module 12 performs amplitude two-stage quantization and phase quantization on the time domain channel information to obtain channel state information, including:

The quantization module 12 determines L codebooks for the time domain channel information of each path, and determines a codebook selection matrix according to the L codebook of the time domain channel information of each path; respectively for the first polarization direction of each path Channel information, channel information of the second polarization direction, and corresponding L codebooks are correlated to obtain a correlation coefficient of each path; and acquiring a first weighting coefficient matrix according to a correlation coefficient of each path; and a first weighting coefficient matrix Amplitude two-stage quantization and phase quantization are performed to obtain a codebook weighting coefficient matrix; channel state information is obtained according to the codebook selection matrix and the codebook weighting coefficient matrix.

Optionally, the quantization module 12 performs amplitude two-level quantization on the weighting coefficient matrix, including:

The quantization module 12 selects, from the first weighting coefficient matrix, the first element with the largest weighting coefficient in each row to form a second weighting coefficient matrix; selects the second element with the largest weighting coefficient from the second weighting coefficient matrix; and adopts the second element Normalizing each element in the second weighting coefficient matrix to obtain a third weighting coefficient matrix; performing quantization on the third weighting coefficient matrix by using the first quantization bit to obtain a first-stage amplitude quantization result; according to the first-level amplitude The quantization result normalizes the correlation coefficient of each path to obtain the normalized correlation coefficient of each path; the second quantization bit is used to quantize the normalized correlation coefficient of each path to obtain the second-level amplitude quantization result .

Optionally, the quantization module 12 performs phase quantization on the weighting coefficient matrix, including:

The quantization module performs phase quantization on each element in the first weighting coefficient matrix by using a preset phase modulation method to obtain a phase quantization matrix.

Optionally, the codebook selection matrix is

Where b is the codebook corresponding to the time domain channel information of each path, and M is the number of time delay paths;

The correlation coefficient of each path is ω _i =[ω _i,1 ,...,ω _i,L ,ω _i,L+1 ,...,ω _i,2L ] _2L*1 ; where i is the index of the time delay path ;

The first weighting coefficient matrix is

The second weighting coefficient matrix is

Where the first element is

The second element is

The first level of quantized results is

The second-order amplitude quantization result of the i-th path is p _{2, i} = [p _{2, i, 1} ... p _{2, i, 2L} ] _1*2L ;

The phase quantization matrix is

The codebook weighting coefficient matrix is ω _quan =[p ₁ ] _M*1 .[p ₂ ] _M*2L .[Φ] _M*2L .

Optionally, the transform module 11 is specifically configured to perform IDFT transform on the information of each transceiver antenna link in the frequency domain channel information H according to formula (2) to obtain time domain channel information Ψ;

Ψ(i _port ,:)=IDFT(H(i _port ,:)),i _port ∈{1,...,N _port } (2)

Where i _port is the port index.

FIG. 10 is a block diagram of an information acquiring apparatus according to another embodiment of the present invention. As shown in FIG. 10, the apparatus further includes:

The receiving module 14 is configured to receive pilot information sent by the network device.

The obtaining module 15 is configured to perform channel estimation according to the pilot information, and acquire the frequency domain channel information.

Optionally, if the antenna of the terminal device is configured as at least two receiving links, the acquiring module 15 is specifically configured to obtain channel feedback information of each resource block according to the pilot information, and perform singular value on the channel feedback information of each resource block. Decomposing, obtaining corresponding feature vectors, and sorting the feature vectors according to the order of the feature values; determining the rank of the channel state information; if the rank of the channel state information is 1, the feature vector having the largest feature value will be Channel information of at least two of the receiving links is integrated into the frequency domain channel information; if the rank of the channel state information is 2, at least two of the receiving according to the feature vectors corresponding to the first two eigenvalues respectively The channel information of the link is integrated into the first frequency domain channel information and the second frequency domain channel information.

The apparatus shown in FIG. 9 and FIG. 10 can be used to implement the method on the terminal side in any of the embodiments of FIG. 2 to FIG. 8. The implementation principle and beneficial effects refer to the foregoing method embodiments, and details are not described herein again.

FIG. 11 is a block diagram of an information acquiring apparatus according to another embodiment of the present application. As shown in FIG. 11, the apparatus includes:

The receiving module 21 is configured to receive channel state information sent by the terminal.

The obtaining module 22 is configured to perform channel reconstruction according to channel state information to obtain time domain channel information.

The transform module 23 is configured to perform discrete Fourier transform on the time domain channel information to obtain frequency domain channel information.

Optionally, the channel state information includes time domain multipath delay indication information, codebook selection indication information, and codebook weighting coefficient information.

Optionally, the obtaining module 22 is specifically configured to perform zero-padding on the weights of the delay paths other than the delay paths corresponding to k ₁ , . . . , k _M , and perform channeling on the M delay paths according to formula (3). Reconstruction, obtaining time domain channel information H _time ;

Where B _i is the ith row of the codebook selection matrix, ω _{quan_i} is the ith row of the codebook weighting coefficient matrix, k _{i is the} time domain multipath delay indication, N _port is the port index, and L is the path of each path. The number of codebooks corresponding to the time domain channel information;

The codebook selection matrix is

The codebook weighting coefficient matrix is ω _quan =[p ₁ ] _M*1 .[p ₂ ] _M*2L .[Φ] _M*2L ;

P ₁ is the first-order amplitude quantization result,

p ₂ is the second-order amplitude quantization result, and the second-order amplitude quantization result of the ith path is p _{2, i} = [p _{2, i, 1} ... p _{2, i, 2L} ] _1*2L ;

Φ is the phase quantization matrix,

FIG. 12 is a block diagram of an information acquiring apparatus according to another embodiment of the present disclosure. As shown in FIG. 12, the apparatus further includes:

The sending module 24 is configured to send pilot information to the terminal, so that the terminal performs channel estimation according to the pilot information.

Optionally, the obtaining module 22 is specifically configured to determine a rank of channel state information.

If the rank of the channel state information is 1, channel reconstruction is performed according to formula (4), and time domain channel information is acquired.

If the rank of the channel state information is 2, channel reconstruction is performed according to formula (5), and time domain channel information is acquired.

among them,

Yes

The i-th column.

The apparatus shown in FIG. 11 and FIG. 12 can be used to implement the method on the network device side in any of the embodiments shown in FIG. 2 to FIG. 8. The implementation principle and beneficial effects refer to the foregoing method embodiments, and details are not described herein again. .

FIG. 13 is a block diagram of a communication device according to an embodiment of the present disclosure. As shown in FIG. 13, the communication device includes a processor 31 and a memory 32. The memory is used to store instructions for executing instructions stored in the memory.

Wherein, when the communication device is configured to implement the terminal function, when the processor 32 executes an instruction to store the memory, the communication device is configured to perform the method on the terminal side shown in any of the embodiments of FIG.

Alternatively, when the communication device is configured to implement a network device function, when the processor 32 executes an instruction to store the memory, the communication device is configured to perform the method on the network device side shown in any of the embodiments of FIGS.

The embodiment of the present application further provides a computer readable storage medium, on which a computer program is stored, and when the program is executed by the processor, the steps of the information acquisition method in any of the embodiments of FIG. 2 to FIG. 8 are implemented.

Claims (30)

1. An information acquisition method, comprising:

   Performing discrete Fourier transform (IDFT) transform on the frequency domain channel information to obtain time domain channel information;

   Performing amplitude two-stage quantization and phase quantization on the time domain channel information to acquire channel state information;

   Transmitting the channel state information to a network device.
2. The method according to claim 1, wherein the performing amplitude two-stage quantization and phase quantization on the time domain channel information to obtain channel state information comprises:

   Obtaining the power of the time domain channel information of each path, and ordering the powers in descending order;

   And performing amplitude two-stage quantization and phase quantization on the time domain channel information corresponding to the first M powers, and acquiring the channel state information.
3. The method according to claim 2, wherein the acquiring the power of the time domain channel information of each path comprises:

   Obtaining the power of the time domain channel information of the kth _path according to formula (1)

   

   

   Where k _path is a multipath index, i _port is a port index, and N _DFT is a point of a delay path.
4. The method according to any one of claims 1-3, wherein the performing amplitude two-stage quantization and phase quantization on the time domain channel information to obtain channel state information comprises:

   Determining L codebooks for time domain channel information of each path, and determining a codebook selection matrix according to an L codebook of time domain channel information of each path;

   And performing correlation processing on the channel information in the first polarization direction of each path, the channel information in the second polarization direction, and the corresponding L codebooks, respectively, to obtain correlation coefficients of each path;

   Obtaining a first weighting coefficient matrix according to the correlation coefficient of each path;

   Performing amplitude two-stage quantization and phase quantization on the first weighting coefficient matrix to obtain a codebook weighting coefficient matrix;

   And obtaining the channel state information according to the codebook selection matrix and the codebook weighting coefficient matrix.
5. The method according to claim 4, wherein said performing amplitude two-level quantization on said matrix of weighting coefficients comprises:

   Selecting, from the first weighting coefficient matrix, a first element having the largest weighting coefficient in each row to form a second weighting coefficient matrix;

   Selecting, from the second matrix of weighting coefficients, a second element having the largest weighting coefficient;

   And normalizing each element in the second weighting coefficient matrix by using the second element to obtain a third weighting coefficient matrix;

   And quantizing the third weighting coefficient matrix by using the first quantization bit to obtain a first level amplitude quantization result;

   And normalizing the correlation coefficient of each path according to the first-stage amplitude quantization result, and acquiring a normalized correlation coefficient of each path;

   The normalized correlation coefficient of each path is quantized by using the second quantization bit to obtain a second-stage amplitude quantization result.
6. The method according to claim 5, wherein said phase quantizing said matrix of weighting coefficients comprises:

   A phase quantization method is performed on each element in the first weighting coefficient matrix by using a preset phase modulation method to obtain a phase quantization matrix.
7. The method of claim 6 wherein:

   The codebook selection matrix is

   

   Where b is the codebook corresponding to the time domain channel information of each path, and M is the number of time delay paths;

   The correlation coefficient of each path is ω _i =[ω _i,1 ,...,ω _i,L ,ω _i,L+1 ,...,ω _i,2L ] _2L*1 ; wherein i is the time delay path index of;

   The first weighting coefficient matrix is

   

   The second matrix of weighting coefficients is

   

   Wherein the first element is

   

   The second element is

   

   The first stage amplitude quantization result is

   

   The second-stage amplitude quantization result of the i-th path is p _2,i =[p _2,i,1 ...p _2,i,2L ] _1*2L ;

   The phase quantization matrix is

   

   The codebook weighting coefficient matrix is ω _quan =[p ₁ ] _M*1 .[p ₂ ] _M*2L .[Φ] _M*2L .
8. The method according to any one of claims 1 to 7, wherein the performing inverse discrete Fourier transform IDFT transform on the frequency domain channel information to obtain time domain channel information comprises:

   Performing the IDFT transformation on the information of each of the transceiver antenna links in the frequency domain channel information H according to the formula (2), acquiring the time domain channel information Ψ;

   Ψ(i _port ,:)=IDFT(H(i _port ,:)),i _port ∈{1,...,N _port } (2)

   Where i _port is the port index.
9. The method according to any one of claims 1-8, wherein the method further comprises: performing an inverse discrete Fourier transform IDFT transform on the frequency domain channel information to obtain the time domain channel information, the method further comprising:

   Receiving pilot information sent by the network device;

   Performing channel estimation according to the pilot information, and acquiring the frequency domain channel information.
10. The method according to claim 9, wherein if the antenna of the terminal device is configured as at least two receiving links, the performing channel estimation according to the pilot information, acquiring the frequency domain channel information, includes:

    Obtaining channel feedback information of each resource block according to the pilot information;

    Performing singular value decomposition on the channel feedback information of each resource block, obtaining corresponding feature vectors, and ordering the feature vectors according to the order of the feature values from large to small;

    Determining a rank of the channel state information;

    If the rank of the channel state information is 1, the channel information of at least two of the receiving links is integrated into the frequency domain channel information according to a feature vector with the largest feature value;

    If the rank of the channel state information is 2, the channel information of the at least two receiving links is integrated into the first frequency domain channel information and the second frequency domain channel information, respectively, according to the feature vectors corresponding to the first two feature values. .
11. An information acquisition method, comprising:

    Receiving channel state information sent by the terminal;

    Performing channel reconstruction according to the channel state information to acquire time domain channel information;

    Performing a discrete Fourier transform on the time domain channel information to obtain frequency domain channel information.
12. The method according to claim 11, wherein the performing channel reconstruction according to the channel state information to obtain time domain channel information comprises:

    Zeroing the weights of the delay paths other than the delay paths corresponding to k ₁ , . . . , k _M , and performing channel reconstruction on the M time delay paths according to formula (3) to obtain the time domain channel information. H _time ;

    

    Where B _i is the ith row of the codebook selection matrix, ω _{quan_i} is the ith row of the codebook weighting coefficient matrix, k _{i is the} time domain multipath delay indication, N _port is the port index, and L is the path of each path. The number of codebooks corresponding to the time domain channel information;

    The codebook selection matrix is

    

    The codebook weighting coefficient matrix is ω _quan =[p ₁ ] _M*1 .[p ₂ ] _M*2L .[Φ] _M*2L ;

    P ₁ is the first-order amplitude quantization result,

    

    p ₂ is the second-stage amplitude quantization result, and the second-stage amplitude quantization result of the ith path is p _2,i =[p _2,i,1 ...p _2,i,2L ] _1*2L ;

    Φ is the phase quantization matrix,

    
13. The method according to claim 11 or 12, wherein before the receiving the channel state information sent by the terminal, the method further includes:

    The pilot information is transmitted to the terminal to cause the terminal to perform channel estimation according to the pilot information.
14. The method according to claim 11, wherein the performing channel reconstruction according to the channel state information to obtain time domain channel information comprises:

    Determining a rank of the channel state information;

    If the rank of the channel state information is 1, channel reconstruction is performed according to formula (4), and time domain channel information is acquired.

    

    

    If the rank of the channel state information is 2, channel reconstruction is performed according to formula (5), and time domain channel information is acquired.

    

    

    among them,

    

    

    Yes

    

    The i-th column.
15. An information acquiring device, comprising:

    a transform module, configured to perform inverse discrete Fourier transform IDFT transform on frequency domain channel information, to obtain time domain channel information;

    a quantization module, configured to perform amplitude two-stage quantization and phase quantization on the time domain channel information, and acquire channel state information;

    And a sending module, configured to send the channel state information to the network device.
16. The apparatus according to claim 15, wherein the quantization module is specifically configured to acquire power of time domain channel information of each path, and order the powers in descending order; The time domain channel information corresponding to the power is subjected to amplitude two-stage quantization and phase quantization to obtain the channel state information.
17. The apparatus according to claim 16, wherein the obtaining, by the quantization module, the power of the time domain channel information of each path comprises:

    The quantization module acquires the power of the time domain channel information of the kth _path according to formula (1)

    

    

    Where k _path is a multipath index, i _port is a port index, and N _DFT is a point of a delay path.
18. The device according to any one of claims 15-17, wherein the quantization module performs amplitude two-stage quantization and phase quantization on the time domain channel information, and acquires channel state information, including:

    The quantization module determines L codebooks for the time domain channel information of each path, and determines a codebook selection matrix according to the L codebook of the time domain channel information of each path; respectively, the first polarization direction of each path The channel information, the channel information of the second polarization direction, and the corresponding L codebooks are correlated to obtain a correlation coefficient of each path; and the first weighting coefficient matrix is obtained according to the correlation coefficient of each path And performing amplitude two-stage quantization and phase quantization on the first weighting coefficient matrix to obtain a codebook weighting coefficient matrix; and acquiring the channel state information according to the codebook selection matrix and the codebook weighting coefficient matrix.
19. The apparatus according to claim 18, wherein the quantization module performs amplitude two-level quantization on the matrix of weighting coefficients, including:

    The quantization module selects, from the first weighting coefficient matrix, a first element having the largest weighting coefficient in each row to form a second weighting coefficient matrix; and selecting a second element having the largest weighting coefficient from the second weighting coefficient matrix And normalizing each element in the second weighting coefficient matrix by using the second element to obtain a third weighting coefficient matrix; and using the first quantization bit to quantize the third weighting coefficient matrix to obtain the first Level 1 amplitude quantization result; normalizing the correlation coefficient of each path according to the first level amplitude quantization result, obtaining a normalized correlation coefficient of each path; using the second quantization bit pair for each The normalized correlation coefficient of the path is quantized to obtain the second-level amplitude quantization result.
20. The apparatus according to claim 19, wherein the quantizing module performs phase quantization on the matrix of weighting coefficients, comprising:

    The quantization module performs phase quantization on each element in the first weighting coefficient matrix by using a preset phase modulation method to obtain a phase quantization matrix.
21. The device of claim 20 wherein:

    The codebook selection matrix is

    

    Where b is the codebook corresponding to the time domain channel information of each path, and M is the number of time delay paths;

    The correlation coefficient of each path is ω _i =[ω _i,1 ,...,ω _i,L ,ω _i,L+1 ,...,ω _i,2L ] _2L*1 ; wherein i is the time delay path index of;

    The first weighting coefficient matrix is

    

    The second matrix of weighting coefficients is

    

    Wherein the first element is

    

    The second element is

    

    The first stage amplitude quantization result is

    

    The second-stage amplitude quantization result of the i-th path is p _2,i =[p _2,i,1 ...p _2,i,2L ] _1*2L ;

    The phase quantization matrix is

    

    The codebook weighting coefficient matrix is ω _quan =[p ₁ ] _M*1 .[p ₂ ] _M*2L .[Φ] _M*2L .
22. The apparatus according to any one of claims 15 to 21, wherein the transforming module is specifically configured to perform information about each of the transmit and receive antenna links in the frequency domain channel information H according to formula (2). Describe the IDFT transform to obtain the time domain channel information Ψ;

    Ψ(i _port ,:)=IDFT(H(i _port ,:)),i _port ∈{1,...,N _port } (2)

    Where i _port is the port index.
23. The device according to any one of claims 15 to 22, wherein if the antenna of the terminal device is configured as at least two receiving links, the device further includes:

    And an acquiring module, configured to acquire channel feedback information of each resource block according to the pilot information, perform singular value decomposition on the channel feedback information of each resource block, obtain a corresponding feature vector, and follow the feature value from large to small Sorting the feature vectors in sequence; determining a rank of the channel state information; if the rank of the channel state information is 1, integrating channel information of at least two of the receive links according to a feature vector having the largest feature value The frequency domain channel information; if the rank of the channel state information is 2, the channel information of at least two of the receiving links are respectively integrated into the first frequency domain channel information according to the feature vectors corresponding to the first two eigenvalues And second frequency domain channel information.
24. An information acquiring device, comprising:

    a receiving module, configured to receive channel state information sent by the terminal;

    An acquiring module, configured to perform channel reconstruction according to the channel state information, and acquire time domain channel information;

    And a transform module, configured to perform discrete Fourier transform on the time domain channel information to obtain frequency domain channel information.
25. The apparatus according to claim 24, wherein the obtaining module is specifically configured to zero-zero the weight of the delay path other than the delay path corresponding to k ₁ , . . . , k _M , and according to the formula (3) Performing channel reconstruction on the M time delay paths to obtain the time domain channel information H _time ;

    

    Where B _i is the ith row of the codebook selection matrix, ω _{quan_i} is the ith row of the codebook weighting coefficient matrix, k _{i is the} time domain multipath delay indication, N _port is the port index, and L is the path of each path. The number of codebooks corresponding to the time domain channel information;

    The codebook selection matrix is

    

    The codebook weighting coefficient matrix is ω _quan =[p ₁ ] _M*1 .[p ₂ ] _M*2L .[Φ] _M*2L ;

    P ₁ is the first-order amplitude quantization result,

    

    p ₂ is the second-stage amplitude quantization result, and the second-stage amplitude quantization result of the ith path is p _2,i =[p _2,i,1 ...p _2,i,2L ] _1*2L ;

    Φ is the phase quantization matrix,

    
26. The apparatus according to claim 24, wherein the obtaining module is specifically configured to determine a rank of the channel state information;

    If the rank of the channel state information is 1, channel reconstruction is performed according to formula (4), and time domain channel information is acquired.

    

    

    If the rank of the channel state information is 2, channel reconstruction is performed according to formula (5), and time domain channel information is acquired.

    

    

    among them,

    

    

    Yes

    

    The i-th column.
27. The apparatus according to any one of claims 1 to 14, wherein the channel state information comprises time domain multipath delay indication information, codebook selection indication information, and Codebook weighting factor information.
28. A terminal characterized by a processor and a memory,

    The memory is configured to store an instruction, the processor is configured to execute the memory stored instruction, and when the processor executes the memory stored instruction, the terminal is configured to perform the method of any one of claims 1 to 10. Methods.
29. A network device characterized by a processor and a memory,

    The memory is for storing instructions for executing the memory stored instructions, the network device for performing any of the claims 11 to 14 when the processor executes the instructions stored by the memory The method described.
30. A computer readable storage medium having stored thereon a computer program, characterized in that the program is executed by a processor to implement the steps of the information acquisition method according to any one of claims 1 to 14.

Images