Signal transmission method and apparatus

Abstract

This application provides a signal transmission method and apparatus, relating to the field of communication technology. In this method, an access network device acquires channel information samples and determines M' frequency domain units from M frequency domain units at the (x+T)th time unit based on the channel information samples. Measurement pilots are then transmitted to the terminal from these M' frequency domain units. The measurement pilots are used to measure CSI (Continuous Channel Indicator). The channel information samples include channel information from the xth time unit to the (x+T-1)th time unit, where M' < M. This method allows the access network device to determine M' frequency domain units from the M frequency domain units at the (x+T)th time unit to transmit the measurement pilots, achieving frequency domain dimensionality reduction of the measurement pilots. Compared to existing technologies that use M frequency domain units to transmit measurement pilots, this reduces the number of frequency domain units used for transmitting the measurement pilots, lowering the overhead of transmitting the measurement pilots and thus reducing the measurement pilot overhead for downlink channel matrix estimation.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a signal transmission method and apparatus. Background Technology

[0002] Massive multiple input multiple output (MIMO) technology, by configuring dozens or even hundreds of antenna ports (hereinafter referred to as ports) in access network equipment, can significantly increase spatial degrees of freedom, and significantly improve spectral efficiency and transmission rate. It has been listed as the fifth generation (5G) technology. th One of the core technologies of 5G (new radio, NR) mobile communication.

[0003] To fully realize the aforementioned potential of massive MIMO, it is necessary to accurately obtain the channel matrix of the space-frequency two-dimensional channel (hereinafter referred to as the space-frequency two-dimensional channel matrix). Normally, taking frequency division duplexing (FDD) massive MIMO as an example, such as... Figure 1 As shown, the process of obtaining the downlink space-frequency two-dimensional channel matrix is ​​as follows: the access network device uses multiple resource elements (REs) to send multiple channel state information-reference signals (CSI-RS) to the terminal. The terminal receives the CSI-RS and estimates the downlink space-frequency two-dimensional channel matrix based on the CSI-RS. It then determines the channel state information (CSI) of the downlink channel based on the downlink space-frequency two-dimensional channel matrix, quantizes the CSI, and feeds back the quantized CSI to the access network device through the uplink channel.

[0004] It should be noted that the CSI-RS overhead required for estimating the downlink space-frequency two-dimensional channel matrix is ​​proportional to the number of ports in the access network equipment and the number of resource blocks (RBs) considered, while the uplink feedback overhead is also proportional to the number of ports in the access network equipment and the number of RBs considered. For example, see [link to example]. Figure 2 As the number of ports increases, the CSI-RS overhead required for estimating the downlink space-frequency two-dimensional channel matrix also increases. In massive MIMO systems, the number of ports in access network equipment is typically very large. Simultaneously, to achieve higher spectral efficiency in future RB-level precoding, the number of RBs considered is usually also very high. Therefore, obtaining the downlink space-frequency two-dimensional channel matrix leads to significant time-frequency resource overhead.

[0005] To reduce the estimation overhead of the downlink space-frequency two-dimensional channel matrix, 3GPP Release 16 (R16) proposes a downlink space-frequency two-dimensional channel matrix compression feedback scheme to reduce uplink feedback overhead. In this scheme, the access network device first sends multiple CSI-RS to the terminal using multiple different (i.e., orthogonal) REs. The terminal estimates the downlink space-frequency two-dimensional channel matrix based on the received CSI-RS. Then, the terminal performs space-frequency two-dimensional compression on the downlink space-frequency two-dimensional channel matrix and feeds it back to the access network device, thereby providing the access network device with the elements with the largest amplitudes and their corresponding positions. For details, see [link to details]. Figure 3 By multiplying the N*M downlink space-frequency two-dimensional channel matrix by an N-column spatial compression matrix and an M-row frequency compression matrix, the downlink space-frequency two-dimensional channel matrix is ​​compressed into a matrix with a smaller number of rows and columns compared to the original downlink space-frequency two-dimensional channel matrix. This compressed matrix is ​​denoted as the feedback coefficient (or feedback matrix) and is fed back to the access network device. In the spatial domain, the channel is projected onto multiple oversampled discrete Fourier transform (DFT) basis beams, and the basis beams with the largest projection coefficients are selected. The indexes of these basis beams and their corresponding projection coefficients are fed back to the access network device. In the frequency domain, the channel is projected onto the DFT basis beams (delay taps), and the taps with the highest energy are selected. The tap positions and their corresponding projection coefficients are fed back to the access network device. The access network device reconstructs the downlink space-frequency two-dimensional channel matrix based on the received positions of the multiple basis beams and their corresponding projection coefficients, as well as the tap positions and their corresponding projection coefficients.

[0006] While the above approach can reduce uplink feedback overhead, the terminal needs to first estimate the complete downlink space-frequency two-dimensional channel matrix by measuring CSI-RS, and then perform space-frequency two-dimensional compression on the downlink space-frequency two-dimensional channel matrix before feeding it back to the access network equipment. Therefore, it cannot reduce the CSI-RS overhead of downlink space-frequency two-dimensional channel matrix estimation (CSI-RS overhead can also be called downlink CSI-RS overhead). Summary of the Invention

[0007] This application provides a signal transmission method and apparatus for reducing the CSI-RS overhead of downlink space-frequency two-dimensional channel matrix estimation.

[0008] To achieve the above objectives, this application provides the following technical solutions:

[0009] In a first aspect, a signal transmission method is provided, comprising: an access network device acquiring channel information samples, the channel information samples including channel information from the x-th time unit to the x+T-1-th time unit, where x and T are both integers greater than or equal to 1; the access network device determining M′ frequency domain units from M frequency domain units in the x+T-th time unit based on the channel information samples, the M′ frequency domain units being used to transmit measurement pilots, the measurement pilots being used to measure channel state information (CSI), where M′ and M are both integers greater than or equal to 1, and M′ < M; and the access network device transmitting the measurement pilots to a terminal in the M′ frequency domain units. The method provided in the first aspect allows the access network device to determine M′ frequency domain units from the M frequency domain units in the x+T-th time unit to transmit measurement pilots, achieving frequency domain dimensionality reduction of the measurement pilots. Compared to existing technologies that use M frequency domain units to transmit measurement pilots, this reduces the number of frequency domain units used to transmit measurement pilots, lowers the overhead of transmitting measurement pilots, and thus reduces the measurement pilot overhead for downlink channel matrix estimation.

[0010] In one possible implementation, the method further includes: the access network device sending first indication information to the terminal, the first indication information indicating the location information of M′ frequency domain units. This possible implementation enables the terminal to determine the frequency domain unit from which the measurement pilot is received.

[0011] In one possible implementation, the access network device transmits measurement pilots to the terminal across M′ frequency domain units. This includes: the access network device transmitting measurement pilots for N′ ports to the terminal in each of the M′ frequency domain units based on channel information samples, and transmitting second indication information to the terminal. The second indication information indicates the value of N′, where N′ is an integer greater than or equal to 1. This possible implementation allows the terminal to determine the number of ports from which the measurement pilots are transmitted.

[0012] In one possible implementation, the method further includes: the access network device receiving feedback information from the terminal, the feedback information indicating information about the measurement pilot signal received by the terminal in the (x+T)th time unit; and the access network device determining the downlink channel matrix in the (x+T)th time unit based on the feedback information. In this possible implementation, after measuring the measurement pilot, the terminal does not perform channel estimation and compression feedback, but directly feeds back the feedback information indicating information about the measurement pilot signal received by the terminal in the (x+T)th time unit to the access network device, thereby reducing the overhead of uplink feedback.

[0013] In one possible implementation, the method further includes: the access network device uses the channel information of p time units as channel information samples to calculate the downlink channel matrix of the x+T+1 time unit, where p is an integer greater than or equal to 1 and less than or equal to x+T.

[0014] In one possible implementation, the measurement pilot transmitted in each of the M′ frequency domain units is a measurement pilot pre-coded using a fourth matrix; wherein the fourth matrix satisfies the following condition: the column correlation of the seventh matrix obtained by multiplying the fourth matrix and the third matrix is ​​minimized. By using the fourth matrix to pre-code the measurement pilot in each of the M′ frequency domain units, spatial dimensionality reduction of the measurement pilot can be achieved.

[0015] In one possible implementation, the downlink channel matrix of a time unit represents the channel information of that time unit. Each time unit's downlink channel matrix is ​​an N*M matrix, where N is an integer greater than or equal to 1. The third matrix satisfies the following condition: the F-norm of the difference between the product of the third and fifth matrices and the sixth matrix is ​​minimized. The sixth matrix is ​​an N*(M*T) matrix, and its column vectors consist of all column vectors from the downlink channel matrices of time units x to x+T-1. The column vectors in the fifth matrix correspond one-to-one with those in the sixth matrix. A column vector in the fifth matrix is ​​a sparsed representation of its corresponding column vector in the sixth matrix. The number of non-zero elements in each column vector of the fifth matrix is ​​equal to S1, where S1 is an integer greater than or equal to 1 and less than N. This possible implementation provides a method for determining the third matrix.

[0016] In one possible implementation, the M′ frequency domain units are determined by a second matrix that satisfies the following conditions: the column correlation of the matrix obtained by multiplying the second matrix and the first matrix is ​​minimized; the second matrix is ​​a row decimation matrix; each row of the second matrix contains only one non-zero element; and the positions of the non-zero elements in different rows are different. This possible implementation provides a method for determining the second matrix.

[0017] In one possible implementation, the first matrix satisfies the following condition: the F-norm of the matrix difference between the product of the first matrix and the eighth matrix and the conjugate transpose of the ninth matrix is ​​minimized; wherein, the ninth matrix is ​​determined based on the channel information samples and the fourth matrix; the m-th column vector of the ninth matrix is ​​composed of T column vectors arranged from top to bottom, the s-th column vector among the T column vectors is the product of the m-th column of the s-th downlink channel matrix and the fourth matrix, the s-th downlink channel matrix is ​​the downlink channel matrix of the x+s-1 time unit from the x-th time unit to the x+T-1-th time unit, m is an integer greater than or equal to 1 and less than or equal to M, and s is an integer greater than or equal to 1 and less than or equal to T; wherein, the column vectors in the eighth matrix correspond one-to-one with the column vectors in the conjugate transpose of the ninth matrix, and one column vector in the eighth matrix is ​​a sparsed representation of the corresponding column vector in the conjugate transpose of the ninth matrix, and the number of non-zero elements in each column vector in the eighth matrix is ​​equal to S2, where S2 is an integer greater than or equal to 1 and less than M. This possible implementation provides a method for determining the first matrix.

[0018] In one possible implementation, the measurement pilot transmitted in the m′-th frequency domain cell out of the M′ frequency domain cells is the measurement pilot after precoding by the m′-th fourth submatrix among the M′ fourth submatrices in the fourth matrix. The fourth matrix includes M fourth submatrices corresponding one-to-one with the M frequency domain cells, and the M′ fourth submatrices are the fourth submatrices corresponding to the M′ frequency domain cells. The m-th fourth submatrix among the M fourth submatrices satisfies the following condition: the m-th seventh submatrix obtained by multiplying the m-th fourth submatrix with the m-th third submatrix in the third matrix has the minimum column correlation; the third matrix includes M third submatrices; m is an integer greater than or equal to 1 and less than or equal to M; and m′ is an integer greater than or equal to 1 and less than or equal to M′. By using each of the fourth submatrices in the fourth matrix to precode the measurement pilot in each of the M′ frequency domain cells, spatial dimensionality reduction of the measurement pilot can be achieved.

[0019] In one possible implementation, the downlink channel matrix of a time unit represents the channel information of that time unit. Each time unit's downlink channel matrix is ​​an N*M matrix, where N is an integer greater than or equal to 1. The m-th third sub-matrix among the M third sub-matrixes satisfies the following condition: the F-norm of the product of the m-th third sub-matrix and the m-th fifth sub-matrix is ​​minimized by the matrix difference of the m-th sixth sub-matrix. The m-th sixth sub-matrix is ​​an N*T matrix, and its column vectors are composed of the m-th column vectors from the downlink channel matrices of time units x to x+T-1. The column vectors of the m-th fifth sub-matrix correspond one-to-one with those of the m-th sixth sub-matrix. A column vector in the m-th fifth sub-matrix is ​​a sparsed representation of the corresponding column vector in the m-th sixth sub-matrix. The number of non-zero elements in each column vector of the m-th fifth sub-matrix is ​​equal to S1, where S1 is an integer greater than or equal to 1 and less than N. This possible implementation provides a method for determining the third matrix.

[0020] In one possible implementation, the M′ frequency domain units are determined by a second matrix that satisfies the following conditions: the column correlation of the matrix obtained by multiplying the second matrix and the first matrix is ​​minimized; the second matrix is ​​a row decimation matrix; each row of the second matrix contains only one non-zero element; and the positions of the non-zero elements in different rows are different. This possible implementation provides a method for determining the second matrix.

[0021] In one possible implementation, the first matrix satisfies the following condition: the F-norm of the matrix difference between the product of the first matrix and the eighth matrix and the conjugate transpose of the ninth matrix is ​​minimized. The ninth matrix is ​​determined based on channel information samples and the fourth matrix. The m-th column vector of the ninth matrix is ​​composed of T column vectors arranged from top to bottom. The s-th column vector among the T column vectors is the product of the m-th column of the s-th downlink channel matrix and the m-th fourth sub-matrix. The s-th downlink channel matrix is ​​the downlink channel matrix for the x+s-1 time units from the x-th to the x+T-1-th time units, where s is an integer greater than or equal to 1 and less than or equal to T. The column vectors in the eighth matrix correspond one-to-one with the column vectors in the conjugate transpose of the ninth matrix. A column vector in the eighth matrix is ​​a sparsed representation of the corresponding column vector in the conjugate transpose of the ninth matrix. The number of non-zero elements in each column vector of the eighth matrix is ​​equal to S², where S² is an integer greater than or equal to 1 and less than M. This possible implementation provides a method for determining the first matrix.

[0022] Secondly, a signal transmission method is provided, comprising: a terminal receiving a measurement pilot signal on time-frequency resources for transmitting measurement pilot signals, and determining feedback information, wherein the feedback information is used to indicate information about the measurement pilot signal received by the terminal; wherein the measurement pilot signal received by the terminal includes N′*M′*R elements, each element representing the measurement pilot signal received by the terminal at one of the N′ ports for transmitting measurement pilot signals, one of the M′ frequency domain units for transmitting measurement pilot signals, and one of the R ports for receiving measurement pilot signals, wherein R, N′, and M′ are all integers greater than or equal to 1; and the terminal sending the feedback information to the access network device. In the method provided by the second aspect, after measuring the measurement pilot signal, the terminal does not perform channel estimation and compression feedback, but directly feeds back the feedback information indicating information about the measurement pilot signal received by the terminal in the x+T time unit to the access network device, thereby reducing the overhead of uplink feedback.

[0023] In one possible implementation, the feedback information is specifically used to indicate the amplitude and phase information of each element in the measurement pilot signal received by the terminal.

[0024] In one possible implementation, the feedback information is specifically used to indicate the absolute amplitude and absolute phase of each element in the measurement pilot signal received by the terminal.

[0025] In one possible implementation, the measurement pilot signal received by the terminal includes multiple sets of elements, each set containing more than one element, and the feedback information is specifically used to indicate the amplitude and phase information of each set of elements.

[0026] In one possible implementation, the feedback information is specifically used to indicate the absolute amplitude and absolute phase of each group of elements.

[0027] In one possible implementation, the method further includes: the terminal receiving first indication information and second indication information from the access network device, wherein the first indication information is used to indicate the location information of the M′ frequency domain units for transmitting the measurement pilot, and the second indication information is used to indicate the value of N′; the terminal receiving the measurement pilot signal on the time-frequency resources for transmitting the measurement pilot and determining feedback information, including: the terminal receiving the measurement pilot signal on N′ ports of each of the M′ frequency domain units and determining the feedback information. In this possible implementation, the terminal can determine the frequency domain units for transmitting the measurement pilot and the number of ports for transmitting the measurement pilot.

[0028] Thirdly, a signal transmission method is provided, comprising: a terminal acquiring channel information samples, the channel information samples including a first downlink channel matrix; the terminal determining a first matrix, a second matrix, and a first coefficient matrix based on the channel information samples, wherein the first coefficient matrix = first downlink channel matrix * second matrix, or the first coefficient matrix = second downlink channel matrix * second matrix, the second downlink channel matrix being determined by the first downlink channel matrix; when the first coefficient matrix = first downlink channel matrix * second matrix, the first downlink channel matrix = second coefficient matrix * first matrix, the number of columns of the first coefficient matrix being less than the number of columns of the first downlink channel matrix; when the first coefficient matrix = second downlink channel matrix * second matrix, the second downlink channel matrix = third coefficient matrix * first matrix, the number of columns of the first coefficient matrix being less than the number of columns of the second downlink channel matrix; the terminal reporting matrix information to an access network device, the matrix information including relevant information of the first coefficient matrix, the first matrix, and the second matrix; or, the matrix information including relevant information of the first coefficient matrix and a tenth matrix, the tenth matrix being determined based on the first matrix and the second matrix. The method provided in this third aspect allows the terminal to compress the first downlink channel matrix (frequency domain compression or spatial domain compression), thereby reducing uplink feedback overhead.

[0029] In one possible implementation, the second coefficient matrix satisfies at least one of the following conditions: 1) the number of non-zero elements in the second coefficient matrix is ​​less than the number of non-zero elements in the first downlink channel matrix; 2) if the ratio of the sum of the energy of the P1 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​greater than a first threshold, and the ratio of the sum of the energy of the P1-1 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​less than the first threshold, then the ratio of the sum of the energy of the P2 largest amplitude elements in the second coefficient matrix to the energy of the second coefficient matrix is ​​greater than the first threshold, and the ratio of the sum of the energy of the P2-1 largest amplitude elements in the second coefficient matrix to the energy of the second coefficient matrix is ​​less than the first threshold, where P2 < P1, P1 and P2 are both integers greater than or equal to 1, and the first threshold is an integer greater than 0 and less than or equal to 1. When the second coefficient matrix satisfies at least one of conditions 1) and 2), it indicates that the second coefficient matrix is ​​a sparse representation of the first downlink channel matrix, thus providing the possibility for downlink channel matrix compression.

[0030] In one possible implementation, the third coefficient matrix satisfies at least one of the following conditions: 1) the number of non-zero elements in the third coefficient matrix is ​​less than the number of non-zero elements in the second downlink channel matrix; 2) if the ratio of the sum of the energies of the P3 largest amplitude elements in the second downlink channel matrix to the energy of the second downlink channel matrix is ​​greater than a second threshold, and the ratio of the sum of the energies of the P3-1 largest amplitude elements in the second downlink channel matrix to the energy of the second downlink channel matrix is ​​less than the second threshold, then the ratio of the sum of the energies of the P4 largest amplitude elements in the third coefficient matrix to the energy of the third coefficient matrix is ​​greater than the second threshold, and the ratio of the sum of the energies of the P4-1 largest amplitude elements in the third coefficient matrix to the energy of the third coefficient matrix is ​​less than the second threshold, where P4 < P3, P3 and P4 are both integers greater than or equal to 1, and the second threshold is an integer greater than 0 and less than or equal to 1. When the third coefficient matrix satisfies at least one of conditions 1) and 2), it indicates that the third coefficient matrix is ​​a sparse representation of the second downlink channel matrix, thus providing the possibility for downlink channel matrix compression.

[0031] In one possible implementation, the first matrix and / or the second matrix satisfy at least one of the following conditions: 1) the magnitude of at least one element in the first matrix and / or the second matrix is ​​different from the magnitude of at least one other element; 2) at least one row or at least one column in the first matrix and / or the second matrix is ​​a non-geometric sequence.

[0032] In one possible implementation, the matrix information further includes information related to the third and fourth matrices, or it may include information related to an eleventh matrix, which is determined based on the third and fourth matrices. The method further includes: the terminal determining the third and fourth matrices based on channel information samples, wherein the second downlink channel matrix = the fourth matrix * the first downlink channel matrix, the first downlink channel matrix = the third matrix * the fourth coefficient matrix, and the number of rows in the second downlink channel matrix is ​​less than the number of rows in the first downlink channel matrix. In this possible implementation, the terminal can further compress the first downlink channel matrix, thereby reducing uplink feedback overhead.

[0033] In one possible implementation, the fourth coefficient matrix satisfies at least one of the following conditions: 1) the number of non-zero elements in the fourth coefficient matrix is ​​less than the number of non-zero elements in the first downlink channel matrix; 2) when the ratio of the sum of the energy of the P5 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​greater than a third threshold, and the ratio of the sum of the energy of the P5-1 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​less than the third threshold, the ratio of the sum of the energy of the P6 largest amplitude elements in the fourth coefficient matrix to the energy of the fourth coefficient matrix is ​​greater than the third threshold, and the ratio of the sum of the energy of the P6-1 largest amplitude elements in the fourth coefficient matrix to the energy of the fourth coefficient matrix is ​​less than the third threshold, where P6 < P5, P5 and P6 are both integers greater than or equal to 1, and the third threshold is an integer greater than 0 and less than or equal to 1. When the fourth coefficient matrix satisfies at least one of conditions 1) and 2), it indicates that the fourth coefficient matrix is ​​a sparse representation of the first downlink channel matrix, thus providing the possibility for downlink channel matrix compression.

[0034] In one possible implementation, the third matrix and / or the fourth matrix satisfy at least one of the following conditions: 1) the magnitude of at least one element in the third matrix and / or the fourth matrix is ​​different from the magnitude of at least one other element; 2) at least one row or at least one column in the third matrix and / or the fourth matrix is ​​a non-geometric sequence.

[0035] In one possible implementation, the channel information samples include T N*M first downlink channel matrices, where T, N, and M are all integers greater than or equal to 1. The third matrix satisfies the following condition: the F-norm of the difference between the product of the third and fifth matrices and the sixth matrix is ​​minimized. The sixth matrix is ​​an N*(M*T) matrix, and its column vectors are composed of all column vectors from the T first downlink channel matrices. The column vectors in the fifth matrix are composed of all column vectors from the T fourth coefficient matrices. A column vector in the fifth matrix is ​​a sparsed representation of the corresponding column vector in the sixth matrix. The number of non-zero elements in each column vector of the fifth matrix is ​​equal to S1, where S1 is an integer greater than or equal to 1 and less than N. This possible implementation provides a method for determining the third matrix.

[0036] In one possible implementation, the fourth matrix satisfies the following condition: the column correlation of the seventh matrix obtained by multiplying the fourth matrix and the third matrix is ​​minimized. This possible implementation provides a method for determining the fourth matrix.

[0037] In one possible implementation, the first matrix satisfies the following condition: the F-norm of the matrix difference between the product of the first matrix and the eighth matrix and the conjugate transpose of the ninth matrix is ​​minimized; wherein, the ninth matrix is ​​determined based on the channel information samples and the fourth matrix; the m-th column vector of the ninth matrix is ​​composed of T column vectors arranged from top to bottom, the s-th column vector among the T column vectors is the product of the m-th column of the s-th first downlink channel matrix and the fourth matrix, the s-th first downlink channel matrix is ​​the first downlink channel matrix of the x+s-1 time unit from the x-th time unit to the x+T-1-th time unit, m is an integer greater than or equal to 1 and less than or equal to M, and s is an integer greater than or equal to 1 and less than or equal to T; wherein, the column vectors in the eighth matrix correspond one-to-one with the column vectors in the conjugate transpose of the ninth matrix, and one column vector in the eighth matrix is ​​a sparsed representation of the corresponding column vector in the conjugate transpose of the ninth matrix, and the number of non-zero elements in each column vector in the eighth matrix is ​​equal to S2, where S2 is an integer greater than or equal to 1 and less than M. This possible implementation provides a method for determining the first matrix.

[0038] In one possible implementation, the channel information samples include T N*M first downlink channel matrices, where T, N, and M are all integers greater than or equal to 1. The third matrix includes M third sub-matrices, where the m-th third sub-matrix satisfies the following condition: the F-norm of the product of the m-th third sub-matrix and the m-th fifth sub-matrix is ​​minimized by the matrix difference of the m-th sixth sub-matrix, where m is an integer greater than or equal to 1 and less than or equal to M. The m-th sixth sub-matrix is ​​an N*T matrix, and its column vectors are composed of the m-th column vectors from the T first downlink channel matrices. The column vectors in the m-th fifth sub-matrix are composed of the m-th column vectors from the T fourth coefficient matrices. One column vector in the m-th fifth sub-matrix is ​​a sparsed representation of the corresponding column vector in the m-th sixth sub-matrix. The number of non-zero elements in each column vector of the m-th fifth sub-matrix is ​​equal to S1, where S1 is an integer greater than or equal to 1 and less than N. This possible implementation provides a method for determining the third matrix.

[0039] In one possible implementation, the fourth matrix comprises M fourth sub-matrices, where the m-th fourth sub-matrix satisfies the condition that the column correlation of the m-th seventh sub-matrix obtained by multiplying the m-th fourth sub-matrix by the m-th third sub-matrix is ​​minimized. This possible implementation provides a method for determining the fourth matrix.

[0040] In one possible implementation, the first matrix satisfies the following condition: the F-norm of the matrix difference between the product of the first matrix and the eighth matrix and the conjugate transpose of the ninth matrix is ​​minimized. The ninth matrix is ​​determined based on channel information samples and the fourth matrix. The m-th column vector of the ninth matrix is ​​composed of T column vectors arranged from top to bottom. The s-th column vector among the T column vectors is the product of the m-th column of the s-th first downlink channel matrix and the m-th fourth sub-matrix. The s-th first downlink channel matrix is ​​the first downlink channel matrix for the x+s-1 time units from the x-th to the x+T-1-th time units, where s is an integer greater than or equal to 1 and less than or equal to T. The column vectors in the eighth matrix correspond one-to-one with the column vectors in the conjugate transpose of the ninth matrix. A column vector in the eighth matrix is ​​a sparsed representation of the corresponding column vector in the conjugate transpose of the ninth matrix. The number of non-zero elements in each column vector of the eighth matrix is ​​equal to S², where S² is an integer greater than or equal to 1 and less than M. This possible implementation provides a method for determining the first matrix.

[0041] In one possible implementation, the second matrix satisfies the following condition: the column correlation of the matrix obtained by multiplying the second matrix and the first matrix is ​​minimized. This possible implementation provides a method for determining the second matrix.

[0042] In one possible implementation, T represents the total number of time units included in the channel information sample; N represents the product of the number of ports configured by the access network device to transmit measurement pilots and the number of ports configured by the terminal to receive measurement pilot signals; M represents the number of frequency domain units configured by the access network device to transmit measurement pilots, or N represents the number of frequency domain units configured by the access network device to transmit measurement pilots and M represents the product of the number of ports configured by the access network device to transmit measurement pilots and the number of ports configured by the terminal to receive measurement pilot signals; wherein, the measurement pilots are used to measure channel state information (CSI).

[0043] In one possible implementation, the reporting period of the first coefficient matrix is ​​shorter than that of the first and second matrices, respectively; or, the reporting period of the first coefficient matrix is ​​shorter than that of the tenth matrix. In this possible implementation, the first and second matrices (or the tenth matrix) do not need to be reported frequently, thereby reducing uplink feedback overhead.

[0044] In one possible implementation, the reporting period of the first coefficient matrix is ​​shorter than that of the first, second, third, and fourth matrices, respectively; or, the reporting period of the first coefficient matrix is ​​shorter than that of the tenth and eleventh matrices. In this possible implementation, the third and fourth (or eleventh) matrices do not need to be reported frequently, thereby reducing uplink feedback overhead.

[0045] Fourthly, a channel matrix acquisition method is provided, comprising: an access network device receiving matrix information from a terminal; wherein the matrix information includes a first coefficient matrix, related information of a first matrix and a second matrix; or, the matrix information includes related information of a first coefficient matrix and a tenth matrix, the tenth matrix being determined based on the first matrix and the second matrix; the first coefficient matrix = a first downlink channel matrix * a second matrix, or the first coefficient matrix = a second downlink channel matrix * a second matrix, the second downlink channel matrix being determined by the first downlink channel matrix; when the first coefficient matrix = a first downlink channel matrix * a second matrix, the first downlink channel matrix = a second coefficient matrix * a first matrix, the number of columns of the first coefficient matrix being less than the number of columns of the first downlink channel matrix; when the first coefficient matrix = a second downlink channel matrix * a second matrix, the second downlink channel matrix = a third coefficient matrix * a first matrix, the number of columns of the first coefficient matrix being less than the number of columns of the second downlink channel matrix; the access network device acquiring the first downlink channel matrix based on the matrix information. The method provided in the fourth aspect allows the terminal to compress the first downlink channel matrix (frequency domain compression or spatial domain compression), thereby reducing uplink feedback overhead.

[0046] In one possible implementation, the second coefficient matrix satisfies at least one of the following conditions: 1) the number of non-zero elements in the second coefficient matrix is ​​less than the number of non-zero elements in the first downlink channel matrix; 2) when the ratio of the sum of the energy of the P1 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​greater than a first threshold, and the ratio of the sum of the energy of the P1-1 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​less than the first threshold, the ratio of the sum of the energy of the P2 largest amplitude elements in the second coefficient matrix to the energy of the second coefficient matrix is ​​greater than the first threshold, and the ratio of the sum of the energy of the P2-1 largest amplitude elements in the second coefficient matrix to the energy of the second coefficient matrix is ​​less than the first threshold, wherein P2 < P1, P1 and P2 are both integers greater than or equal to 1, and the first threshold is an integer greater than 0 and less than or equal to 1.

[0047] In one possible implementation, the third coefficient matrix satisfies at least one of the following conditions: 1) the number of non-zero elements in the third coefficient matrix is ​​less than the number of non-zero elements in the second downlink channel matrix; 2) when the ratio of the sum of the energy of the P3 largest amplitude elements in the second downlink channel matrix to the energy of the second downlink channel matrix is ​​greater than a second threshold, and the ratio of the sum of the energy of the P3-1 largest amplitude elements in the second downlink channel matrix to the energy of the second downlink channel matrix is ​​less than a second threshold, the ratio of the sum of the energy of the P4 largest amplitude elements in the third coefficient matrix to the energy of the third coefficient matrix is ​​greater than a second threshold, and the ratio of the sum of the energy of the P4-1 largest amplitude elements in the third coefficient matrix to the energy of the third coefficient matrix is ​​less than a second threshold, where P4 < P3, P3 and P4 are both integers greater than or equal to 1, and the second threshold is an integer greater than 0 and less than or equal to 1.

[0048] In one possible implementation, the first matrix and / or the second matrix satisfy at least one of the following conditions: 1) the magnitude of at least one element in the first matrix and / or the second matrix is ​​different from the magnitude of at least one other element; 2) at least one row or at least one column in the first matrix and / or the second matrix is ​​a non-geometric sequence.

[0049] In one possible implementation, the reporting period of the first coefficient matrix is ​​shorter than that of the first and second matrices, respectively; or, the reporting period of the first coefficient matrix is ​​shorter than that of the tenth matrix. In this possible implementation, the first and second matrices (or the tenth matrix) do not need to be reported frequently, thereby reducing uplink feedback overhead.

[0050] Fifthly, a channel matrix acquisition method is provided, comprising: an access network device receiving matrix information from a terminal; wherein the matrix information includes a first coefficient matrix, a first matrix, a second matrix, a third matrix, and a fourth matrix; or, the matrix information includes relevant information of a first coefficient matrix, a tenth matrix, and an eleventh matrix, wherein the tenth matrix is ​​determined based on the first and second matrices, and the eleventh matrix is ​​determined based on the third and fourth matrices; the first coefficient matrix = a second downlink channel matrix * a second matrix, the second downlink channel matrix being determined by the first downlink channel matrix; the second downlink channel matrix = a third coefficient matrix * a first matrix, the number of columns in the first coefficient matrix being less than the number of columns in the second downlink channel matrix; the second downlink channel matrix = a fourth matrix * a first downlink channel matrix, the first downlink channel matrix = a third matrix * a fourth coefficient matrix, the number of rows in the second downlink channel matrix being less than the number of rows in the first downlink channel matrix; and the access network device acquiring the first downlink channel matrix based on the matrix information. The method provided in the fifth aspect allows the terminal to compress the first downlink channel matrix (frequency domain compression or spatial domain compression), thereby reducing uplink feedback overhead.

[0051] In one possible implementation, the third coefficient matrix satisfies at least one of the following conditions: 1) the number of non-zero elements in the third coefficient matrix is ​​less than the number of non-zero elements in the second downlink channel matrix; 2) when the ratio of the sum of the energy of the P3 largest amplitude elements in the second downlink channel matrix to the energy of the second downlink channel matrix is ​​greater than a second threshold, and the ratio of the sum of the energy of the P3-1 largest amplitude elements in the second downlink channel matrix to the energy of the second downlink channel matrix is ​​less than a second threshold, the ratio of the sum of the energy of the P4 largest amplitude elements in the third coefficient matrix to the energy of the third coefficient matrix is ​​greater than a second threshold, and the ratio of the sum of the energy of the P4-1 largest amplitude elements in the third coefficient matrix to the energy of the third coefficient matrix is ​​less than a second threshold, where P4 < P3, P3 and P4 are both integers greater than or equal to 1, and the second threshold is an integer greater than 0 and less than or equal to 1.

[0052] In one possible implementation, the fourth coefficient matrix satisfies at least one of the following conditions: 1) the number of non-zero elements in the fourth coefficient matrix is ​​less than the number of non-zero elements in the first downlink channel matrix; 2) when the ratio of the sum of the energy of the P5 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​greater than a third threshold, and the ratio of the sum of the energy of the P5-1 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​less than a third threshold, the ratio of the sum of the energy of the P6 largest amplitude elements in the fourth coefficient matrix to the energy of the fourth coefficient matrix is ​​greater than a third threshold, and the ratio of the sum of the energy of the P6-1 largest amplitude elements in the fourth coefficient matrix to the energy of the fourth coefficient matrix is ​​less than a third threshold, where P6 < P5, P5 and P6 are both integers greater than or equal to 1, and the third threshold is an integer greater than 0 and less than or equal to 1.

[0053] In one possible implementation, the first matrix and / or the second matrix satisfy at least one of the following conditions: 1) the magnitude of at least one element in the first matrix and / or the second matrix is ​​different from the magnitude of at least one other element; 2) at least one row or at least one column in the first matrix and / or the second matrix is ​​a non-geometric sequence.

[0054] In one possible implementation, the third matrix and / or the fourth matrix satisfy at least one of the following conditions: 1) the magnitude of at least one element in the third matrix and / or the fourth matrix is ​​different from the magnitude of at least one other element; 2) at least one row or at least one column in the third matrix and / or the fourth matrix is ​​a non-geometric sequence.

[0055] In one possible implementation, the reporting period of the first coefficient matrix is ​​shorter than that of the first, second, third, and fourth matrices, respectively; or, the reporting period of the first coefficient matrix is ​​shorter than that of the tenth and eleventh matrices. In this possible implementation, the first, second, third, and fourth matrices (or the tenth and eleventh matrices) do not need to report frequently, thereby reducing uplink feedback overhead.

[0056] In a sixth aspect, a signal transmitting apparatus is provided, comprising: a communication unit and a processing unit; the processing unit is configured to acquire channel information samples, the channel information samples including channel information from the x-th time unit to the x+T-1-th time unit, where x and T are both integers greater than or equal to 1; the processing unit is further configured to determine M′ frequency domain units from M frequency domain units in the x+T-th time unit based on the channel information samples, the M′ frequency domain units being used to transmit measurement pilots, the measurement pilots being used to measure CSI, where M′ and M are both integers greater than or equal to 1, and M′ < M; the communication unit is configured to transmit the measurement pilots to a terminal in the M′ frequency domain units.

[0057] In one possible implementation, the communication unit is further configured to send first indication information to the terminal, the first indication information being used to indicate the position information of the M′ frequency domain units.

[0058] In one possible implementation, the communication unit is specifically configured to send measurement pilots of N′ ports to the terminal in each of the M′ frequency domain units according to the channel information sample, and to send second indication information to the terminal, the second indication information being used to indicate the value of N′, where N′ is an integer greater than or equal to 1.

[0059] In one possible implementation, the communication unit is further configured to receive feedback information from the terminal, the feedback information indicating information about the measurement pilot signal received by the terminal in the (x+T)th time unit; the processing unit is further configured to determine the downlink channel matrix of the (x+T)th time unit based on the feedback information.

[0060] In one possible implementation, the processing unit is further configured to use the channel information of p time units as channel information samples to calculate the downlink channel matrix of the (x+T+1)th time unit, where p is an integer greater than or equal to 1 and less than or equal to x+T.

[0061] In one possible implementation, the measurement pilot transmitted on each of the M′ frequency domain units is a measurement pilot pre-coded by a fourth matrix; wherein the fourth matrix satisfies the following condition: the column correlation of the seventh matrix obtained by multiplying the fourth matrix and the third matrix is ​​minimal.

[0062] In one possible implementation, the downlink channel matrix of a time unit represents the channel information of that time unit. The downlink channel matrix of each time unit is an N*M matrix, where N is an integer greater than or equal to 1. The third matrix satisfies the following condition: the F-norm of the matrix difference between the product of the third matrix and the fifth matrix and the sixth matrix is ​​minimized. The sixth matrix is ​​an N*(M*T) matrix, and the column vectors in the sixth matrix are composed of all the column vectors in the downlink channel matrices from the x-th time unit to the x+T-1-th time unit. The column vectors in the fifth matrix correspond one-to-one with the column vectors in the sixth matrix. One column vector in the fifth matrix is ​​a sparse representation of the corresponding column vector in the sixth matrix. The number of non-zero elements in each column vector of the fifth matrix is ​​equal to S1, where S1 is an integer greater than or equal to 1 and less than N.

[0063] In one possible implementation, the M′ frequency domain units are determined by a second matrix that satisfies the following conditions: the column correlation of the matrix obtained by multiplying the second matrix with the first matrix is ​​minimized; the second matrix is ​​a row decimation matrix; each row of the second matrix has only one non-zero element; and the non-zero elements in different rows are in different positions.

[0064] In one possible implementation, the first matrix satisfies the following condition: the F-norm of the matrix difference between the product of the first matrix and the eighth matrix and the conjugate transpose of the ninth matrix is ​​minimized; wherein, the ninth matrix is ​​determined based on the channel information samples and the fourth matrix; the m-th column vector of the ninth matrix is ​​composed of T column vectors arranged from top to bottom, the s-th column vector of the T column vectors is the product of the m-th column of the s-th downlink channel matrix and the fourth matrix, and the s-th downlink channel matrix is ​​the matrix from the x-th time unit to the x-th time unit. The downlink channel matrix of the x+s-1th time unit in the +T-1th time unit, where m is an integer greater than or equal to 1 and less than or equal to M, and s is an integer greater than or equal to 1 and less than or equal to T; wherein, the column vectors in the eighth matrix correspond one-to-one with the column vectors in the conjugate transpose of the ninth matrix, and one column vector in the eighth matrix is ​​a sparsed representation of the corresponding column vector in the conjugate transpose of the ninth matrix, and the number of non-zero elements in each column vector in the eighth matrix is ​​equal to S2, where S2 is an integer greater than or equal to 1 and less than M.

[0065] In one possible implementation, the measurement pilot transmitted on the m′-th frequency domain unit among the M′ frequency domain units is a measurement pilot pre-coded by the m′-th fourth submatrix among the M′ fourth submatrixes in the fourth matrix. The fourth matrix includes M fourth submatrixes corresponding one-to-one with the M frequency domain units, and the M′ fourth submatrixes are the fourth submatrixes corresponding to the M′ frequency domain units. The m-th fourth submatrix among the M fourth submatrixes satisfies the following condition: the column correlation of the m-th seventh submatrix obtained by multiplying the m-th fourth submatrix with the m-th third submatrix in the third matrix is ​​minimal. The third matrix includes M third submatrixes. m is an integer greater than or equal to 1 and less than or equal to M, and m′ is an integer greater than or equal to 1 and less than or equal to M′.

[0066] In one possible implementation, the downlink channel matrix of a time unit represents the channel information of that time unit. The downlink channel matrix of each time unit is an N*M matrix, where N is an integer greater than or equal to 1. The m-th third sub-matrix among the M third sub-matrixes satisfies the following condition: the F-norm of the product of the m-th third sub-matrix and the m-th fifth sub-matrix and the matrix difference of the m-th sixth sub-matrix is ​​minimized. The m-th sixth sub-matrix is ​​an N*T matrix, and the column vectors in the m-th sixth sub-matrix are composed of the m-th column vectors in the downlink channel matrices from the x-th time unit to the x+T-1-th time unit. The column vectors in the m-th fifth sub-matrix correspond one-to-one with the column vectors in the m-th sixth sub-matrix. One column vector in the m-th fifth sub-matrix is ​​a sparse representation of the corresponding column vector in the m-th sixth sub-matrix. The number of non-zero elements in each column vector of the m-th fifth sub-matrix is ​​equal to S1, where S1 is an integer greater than or equal to 1 and less than N.

[0067] In one possible implementation, the M′ frequency domain units are determined by a second matrix that satisfies the following conditions: the column correlation of the matrix obtained by multiplying the second matrix with the first matrix is ​​minimized; the second matrix is ​​a row decimation matrix; each row of the second matrix has only one non-zero element; and the non-zero elements in different rows are in different positions.

[0068] In one possible implementation, the first matrix satisfies the following condition: the F-norm of the matrix difference between the product of the first matrix and the eighth matrix and the conjugate transpose of the ninth matrix is ​​minimized; wherein, the ninth matrix is ​​determined based on the channel information sample and the fourth matrix, the m-th column vector of the ninth matrix is ​​composed of T column vectors arranged from top to bottom, the s-th column vector among the T column vectors is the product of the m-th column of the s-th downlink channel matrix and the m-th fourth sub-matrix, the s-th downlink channel matrix is ​​the downlink channel matrix of the x-th time unit to the x+t-1-th time unit, s is an integer greater than or equal to 1 and less than or equal to T; wherein, the column vectors in the eighth matrix correspond one-to-one with the column vectors in the conjugate transpose of the ninth matrix, one column vector in the eighth matrix is ​​a sparsed representation of the corresponding column vector in the conjugate transpose of the ninth matrix, and the number of non-zero elements in each column vector in the eighth matrix is ​​equal to S2, S2 is an integer greater than or equal to 1 and less than M.

[0069] In a seventh aspect, a signal transmitting apparatus is provided, comprising: a communication unit and a processing unit; the processing unit is configured to receive a measurement pilot signal on a time-frequency resource for transmitting a measurement pilot signal via the communication unit, and determine feedback information, the feedback information being used to indicate information about the measurement pilot signal received by the apparatus; wherein the measurement pilot signal received by the apparatus comprises N′*M′*R elements, each element representing the measurement pilot signal received by the apparatus at one of the N′ ports for transmitting the measurement pilot signal, one of the M′ frequency domain units for transmitting the measurement pilot signal, and one of the R ports for receiving the measurement pilot signal, wherein R, N′, and M′ are all integers greater than or equal to 1; the processing unit is further configured to send the feedback information to an access network device via the communication unit.

[0070] In one possible implementation, the feedback information is specifically used to indicate the amplitude and phase information of each element in the measurement pilot signal received by the device.

[0071] In one possible implementation, the feedback information is specifically used to indicate the absolute amplitude and absolute phase of each element in the measurement pilot signal received by the device.

[0072] In one possible implementation, the measurement pilot signal received by the device includes multiple sets of elements, each set containing more than one element, and the feedback information is specifically used to indicate the amplitude and phase information of each set of elements.

[0073] In one possible implementation, the feedback information is specifically used to indicate the absolute amplitude and absolute phase of each group of elements.

[0074] In one possible implementation, the processing unit is further configured to receive first indication information and second indication information from the access network device via the communication unit, wherein the first indication information is used to indicate the location information of the M′ frequency domain units that transmit the measurement pilot, and the second indication information is used to indicate the value of N′; specifically, the processing unit is configured to receive the measurement pilot signal on the N′ ports of each of the M′ frequency domain units and determine the feedback information.

[0075] Eighthly, a signal transmitting apparatus is provided, comprising: a communication unit and a processing unit; the processing unit is configured to acquire channel information samples, the channel information samples including a first downlink channel matrix; the processing unit is further configured to determine a first matrix, a second matrix, and a first coefficient matrix based on the channel information samples, wherein the first coefficient matrix = the first downlink channel matrix * the second matrix, or, the first coefficient matrix = the second downlink channel matrix * the second matrix, the second downlink channel matrix being determined by the first downlink channel matrix; when the first coefficient matrix = the first downlink channel matrix * the second matrix, the first downlink channel matrix = the second coefficient matrix * the first matrix, the number of columns of the first coefficient matrix being less than the number of columns of the first downlink channel matrix; when the first coefficient matrix = the second downlink channel matrix * the second matrix, the second downlink channel matrix = a third coefficient matrix * the first matrix, the number of columns of the first coefficient matrix being less than the number of columns of the second downlink channel matrix; the communication unit is configured to report matrix information to an access network device, the matrix information including relevant information of the first coefficient matrix, the first matrix, and the second matrix; or, the matrix information including relevant information of the first coefficient matrix and a tenth matrix, the tenth matrix being determined based on the first matrix and the second matrix.

[0076] In one possible implementation, the second coefficient matrix satisfies at least one of the following conditions: 1) the number of non-zero elements in the second coefficient matrix is ​​less than the number of non-zero elements in the first downlink channel matrix; 2) when the ratio of the sum of the energy of the P1 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​greater than a first threshold, and the ratio of the sum of the energy of the P1-1 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​less than the first threshold, the ratio of the sum of the energy of the P2 largest amplitude elements in the second coefficient matrix to the energy of the second coefficient matrix is ​​greater than the first threshold, and the ratio of the sum of the energy of the P2-1 largest amplitude elements in the second coefficient matrix to the energy of the second coefficient matrix is ​​less than the first threshold, wherein P2 < P1, P1 and P2 are both integers greater than or equal to 1, and the first threshold is an integer greater than 0 and less than or equal to 1.

[0077] In one possible implementation, the third coefficient matrix satisfies at least one of the following conditions: 1) the number of non-zero elements in the third coefficient matrix is ​​less than the number of non-zero elements in the second downlink channel matrix; 2) when the ratio of the sum of the energies of the P3 largest amplitude elements in the second downlink channel matrix to the energy of the second downlink channel matrix is ​​greater than a second threshold, and the ratio of the sum of the energies of the P3-1 largest amplitude elements in the second downlink channel matrix to the energy of the second downlink channel matrix is ​​less than the second threshold, the ratio of the sum of the energies of the P4 largest amplitude elements in the third coefficient matrix to the energy of the third coefficient matrix is ​​greater than the second threshold, and the ratio of the sum of the energies of the P4-1 largest amplitude elements in the third coefficient matrix to the energy of the third coefficient matrix is ​​less than the second threshold, wherein P4 < P3, P3 and P4 are both integers greater than or equal to 1, and the second threshold is an integer greater than 0 and less than or equal to 1.

[0078] In one possible implementation, the first matrix and / or the second matrix satisfies at least one of the following conditions: 1) the magnitude of at least one element in the first matrix and / or the second matrix is ​​different from the magnitude of at least one other element; 2) at least one row or at least one column in the first matrix and / or the second matrix is ​​a non-geometric sequence.

[0079] In one possible implementation, the matrix information further includes relevant information about the third and fourth matrices, or the matrix information further includes relevant information about the eleventh matrix, which is determined based on the third and fourth matrices. The processing unit is further configured to determine the third and fourth matrices based on the channel information samples, wherein the second downlink channel matrix = the fourth matrix * the first downlink channel matrix, the first downlink channel matrix = the third matrix * the fourth coefficient matrix, and the number of rows in the second downlink channel matrix is ​​less than the number of rows in the first downlink channel matrix.

[0080] In one possible implementation, the fourth coefficient matrix satisfies at least one of the following conditions: 1) the number of non-zero elements in the fourth coefficient matrix is ​​less than the number of non-zero elements in the first downlink channel matrix; 2) when the ratio of the sum of the energy of the P5 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​greater than a third threshold, and the ratio of the sum of the energy of the P5-1 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​less than the third threshold, the ratio of the sum of the energy of the P6 largest amplitude elements in the fourth coefficient matrix to the energy of the fourth coefficient matrix is ​​greater than the third threshold, and the ratio of the sum of the energy of the P6-1 largest amplitude elements in the fourth coefficient matrix to the energy of the fourth coefficient matrix is ​​less than the third threshold, wherein P6 < P5, P5 and P6 are both integers greater than or equal to 1, and the third threshold is an integer greater than 0 and less than or equal to 1.

[0081] In one possible implementation, the third matrix and / or the fourth matrix satisfies at least one of the following conditions: 1) the magnitude of at least one element in the third matrix and / or the fourth matrix is ​​different from the magnitude of at least one other element; 2) at least one row or at least one column in the third matrix and / or the fourth matrix is ​​a non-geometric sequence.

[0082] In one possible implementation, the channel information sample includes T N*M first downlink channel matrices, where T, N, and M are all integers greater than or equal to 1. The third matrix satisfies the following condition: the F-norm of the matrix difference between the product of the third matrix and the fifth matrix and the sixth matrix is ​​minimized. The sixth matrix is ​​an N*(M*T) matrix, and its column vectors are composed of all column vectors from the T first downlink channel matrices. The column vectors in the fifth matrix are composed of all column vectors from the T fourth coefficient matrices. One column vector in the fifth matrix is ​​a sparse representation of the corresponding column vector in the sixth matrix. The number of non-zero elements in each column vector of the fifth matrix is ​​equal to S1, where S1 is an integer greater than or equal to 1 and less than N.

[0083] In one possible implementation, the fourth matrix satisfies the following condition: the column correlation of the seventh matrix obtained by multiplying the fourth matrix and the third matrix is ​​minimal.

[0084] In one possible implementation, the first matrix satisfies the following condition: the F-norm of the matrix difference between the product of the first matrix and the eighth matrix and the conjugate transpose of the ninth matrix is ​​minimized; wherein, the ninth matrix is ​​determined based on the channel information samples and the fourth matrix; the m-th column vector of the ninth matrix is ​​composed of T column vectors arranged from top to bottom, the s-th column vector of the T column vectors is the product of the m-th column of the s-th first downlink channel matrix and the fourth matrix, and the s-th first downlink channel matrix is ​​the product of the x-th time unit to the s-th time unit. The first downlink channel matrix of the x+s-1th time unit in the x+T-1th time unit, where m is an integer greater than or equal to 1 and less than or equal to M, and s is an integer greater than or equal to 1 and less than or equal to T; wherein, the column vectors in the eighth matrix correspond one-to-one with the column vectors in the conjugate transpose of the ninth matrix, and one column vector in the eighth matrix is ​​a sparsed representation of the corresponding column vector in the conjugate transpose of the ninth matrix, and the number of non-zero elements in each column vector in the eighth matrix is ​​equal to S2, where S2 is an integer greater than or equal to 1 and less than M.

[0085] In one possible implementation, the channel information sample includes T N*M first downlink channel matrices, where T, N, and M are all integers greater than or equal to 1. The third matrix includes M third sub-matrices, and the m-th third sub-matrix among the M third sub-matrices satisfies the following condition: the F-norm of the matrix difference between the product of the m-th third sub-matrix and the m-th fifth sub-matrix and the m-th sixth sub-matrix is ​​minimized, where m is an integer greater than or equal to 1 and less than or equal to M. The m-th sixth sub-matrix is ​​an N*T matrix, and the column vectors in the m-th sixth sub-matrix are composed of the m-th column vectors from the T first downlink channel matrices. The column vectors in the m-th fifth sub-matrix are composed of the m-th column vectors from the T fourth coefficient matrices. One column vector in the m-th fifth sub-matrix is ​​a sparse representation of the corresponding column vector in the m-th sixth sub-matrix. The number of non-zero elements in each column vector of the m-th fifth sub-matrix is ​​equal to S1, where S1 is an integer greater than or equal to 1 and less than N.

[0086] In one possible implementation, the fourth matrix comprises M fourth sub-matrices, wherein the m-th fourth sub-matrix satisfies the following condition: the column correlation of the m-th seventh sub-matrix obtained by multiplying the m-th fourth sub-matrix with the m-th third sub-matrix is ​​minimized.

[0087] In one possible implementation, the first matrix satisfies the following condition: the F-norm of the matrix difference between the product of the first matrix and the eighth matrix and the conjugate transpose of the ninth matrix is ​​minimized; wherein, the ninth matrix is ​​determined based on the channel information sample and the fourth matrix, the m-th column vector of the ninth matrix is ​​composed of T column vectors arranged from top to bottom, the s-th column vector among the T column vectors is the product of the m-th column of the s-th first downlink channel matrix and the m-th fourth sub-matrix, the s-th first downlink channel matrix is ​​the first downlink channel matrix of the x-th time unit to the x+t-1-th time unit, s is an integer greater than or equal to 1 and less than or equal to T; wherein, the column vectors in the eighth matrix correspond one-to-one with the column vectors in the conjugate transpose of the ninth matrix, one column vector in the eighth matrix is ​​a sparse representation of the corresponding column vector in the conjugate transpose of the ninth matrix, and the number of non-zero elements in each column vector in the eighth matrix is ​​equal to S2, S2 is an integer greater than or equal to 1 and less than M.

[0088] In one possible implementation, the second matrix satisfies the following condition: the column correlation of the matrix obtained by multiplying the second matrix and the first matrix is ​​minimized.

[0089] In one possible implementation, T represents the total number of time units included in the channel information sample; N represents the product of the number of ports configured by the access network device to transmit measurement pilots and the number of ports configured by the device to receive measurement pilot signals; M represents the number of frequency domain units configured by the access network device to transmit the measurement pilots, or, N represents the number of frequency domain units configured by the access network device to transmit the measurement pilots, and M represents the product of the number of ports configured by the access network device to transmit the measurement pilots and the number of ports configured by the device to receive measurement pilot signals; wherein, the measurement pilots are used to measure channel state information (CSI).

[0090] In one possible implementation, the reporting period of the first coefficient matrix is ​​less than the reporting periods of the first matrix and the second matrix, respectively; or, the reporting period of the first coefficient matrix is ​​less than the reporting period of the tenth matrix.

[0091] In one possible implementation, the reporting period of the first coefficient matrix is ​​less than the reporting periods of the first matrix, the second matrix, the third matrix, and the fourth matrix, respectively; or, the reporting period of the first coefficient matrix is ​​less than the reporting periods of the tenth matrix and the eleventh matrix.

[0092] A ninth aspect provides a channel matrix acquisition apparatus, comprising: a communication unit and a processing unit; the communication unit being configured to receive matrix information from a terminal; wherein the matrix information includes information related to a first coefficient matrix, a first matrix, and a second matrix; or, the matrix information includes information related to a first coefficient matrix and a tenth matrix, the tenth matrix being determined based on the first matrix and the second matrix; the first coefficient matrix = a first downlink channel matrix * the second matrix, or the first coefficient matrix = a second downlink channel matrix * the second matrix, the second downlink channel matrix being determined by the first downlink channel matrix; in the case where the first coefficient matrix = the first downlink channel matrix * the second matrix, the first downlink channel matrix = the second coefficient matrix * the first matrix, the number of columns of the first coefficient matrix being less than the number of columns of the first downlink channel matrix; in the case where the first coefficient matrix = the second downlink channel matrix * the second matrix, the second downlink channel matrix = a third coefficient matrix * the first matrix, the number of columns of the first coefficient matrix being less than the number of columns of the second downlink channel matrix; the processing unit being configured to acquire the first downlink channel matrix based on the matrix information.

[0093] In one possible implementation, the second coefficient matrix satisfies at least one of the following conditions: 1) the number of non-zero elements in the second coefficient matrix is ​​less than the number of non-zero elements in the first downlink channel matrix; 2) when the ratio of the sum of the energy of the P1 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​greater than a first threshold, and the ratio of the sum of the energy of the P1-1 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​less than the first threshold, the ratio of the sum of the energy of the P2 largest amplitude elements in the second coefficient matrix to the energy of the second coefficient matrix is ​​greater than the first threshold, and the ratio of the sum of the energy of the P2-1 largest amplitude elements in the second coefficient matrix to the energy of the second coefficient matrix is ​​less than the first threshold, wherein P2 < P1, P1 and P2 are both integers greater than or equal to 1, and the first threshold is an integer greater than 0 and less than or equal to 1.

[0094] In one possible implementation, the third coefficient matrix satisfies at least one of the following conditions: 1) the number of non-zero elements in the third coefficient matrix is ​​less than the number of non-zero elements in the second downlink channel matrix; 2) when the ratio of the sum of the energies of the P3 largest amplitude elements in the second downlink channel matrix to the energy of the second downlink channel matrix is ​​greater than a second threshold, and the ratio of the sum of the energies of the P3-1 largest amplitude elements in the second downlink channel matrix to the energy of the second downlink channel matrix is ​​less than the second threshold, the ratio of the sum of the energies of the P4 largest amplitude elements in the third coefficient matrix to the energy of the third coefficient matrix is ​​greater than the second threshold, and the ratio of the sum of the energies of the P4-1 largest amplitude elements in the third coefficient matrix to the energy of the third coefficient matrix is ​​less than the second threshold, wherein P4 < P3, P3 and P4 are both integers greater than or equal to 1, and the second threshold is an integer greater than 0 and less than or equal to 1.

[0095] In one possible implementation, the first matrix and / or the second matrix satisfies at least one of the following conditions: 1) the magnitude of at least one element in the first matrix and / or the second matrix is ​​different from the magnitude of at least one other element; 2) at least one row or at least one column in the first matrix and / or the second matrix is ​​a non-geometric sequence.

[0096] In one possible implementation, the reporting period of the first coefficient matrix is ​​less than the reporting periods of the first matrix and the second matrix, respectively; or, the reporting period of the first coefficient matrix is ​​less than the reporting period of the tenth matrix.

[0097] A tenth aspect provides a channel matrix acquisition device, comprising: a communication unit and a processing unit; the communication unit being configured to receive matrix information from a terminal; wherein the matrix information includes a first coefficient matrix, a first matrix, a second matrix, a third matrix, and a fourth matrix; or, the matrix information includes relevant information of a first coefficient matrix, a tenth matrix, and an eleventh matrix, wherein the tenth matrix is ​​determined based on the first matrix and the second matrix, and the eleventh matrix is ​​determined based on the third matrix and the fourth matrix; the first coefficient matrix = a second downlink channel matrix * the second matrix, the second downlink channel matrix being determined by the first downlink channel matrix; the second downlink channel matrix = a third coefficient matrix * the first matrix, the number of columns of the first coefficient matrix being less than the number of columns of the second downlink channel matrix; the second downlink channel matrix = the fourth matrix * the first downlink channel matrix, the first downlink channel matrix = the third matrix * the fourth coefficient matrix, the number of rows of the second downlink channel matrix being less than the number of rows of the first downlink channel matrix; the processing unit being configured to acquire the first downlink channel matrix based on the matrix information.

[0098] In one possible implementation, the third coefficient matrix satisfies at least one of the following conditions: 1) the number of non-zero elements in the third coefficient matrix is ​​less than the number of non-zero elements in the second downlink channel matrix; 2) when the ratio of the sum of the energies of the P3 largest amplitude elements in the second downlink channel matrix to the energy of the second downlink channel matrix is ​​greater than a second threshold, and the ratio of the sum of the energies of the P3-1 largest amplitude elements in the second downlink channel matrix to the energy of the second downlink channel matrix is ​​less than the second threshold, the ratio of the sum of the energies of the P4 largest amplitude elements in the third coefficient matrix to the energy of the third coefficient matrix is ​​greater than the second threshold, and the ratio of the sum of the energies of the P4-1 largest amplitude elements in the third coefficient matrix to the energy of the third coefficient matrix is ​​less than the second threshold, wherein P4 < P3, P3 and P4 are both integers greater than or equal to 1, and the second threshold is an integer greater than 0 and less than or equal to 1.

[0099] In one possible implementation, the fourth coefficient matrix satisfies at least one of the following conditions: 1) the number of non-zero elements in the fourth coefficient matrix is ​​less than the number of non-zero elements in the first downlink channel matrix; 2) when the ratio of the sum of the energy of the P5 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​greater than a third threshold, and the ratio of the sum of the energy of the P5-1 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​less than the third threshold, the ratio of the sum of the energy of the P6 largest amplitude elements in the fourth coefficient matrix to the energy of the fourth coefficient matrix is ​​greater than the third threshold, and the ratio of the sum of the energy of the P6-1 largest amplitude elements in the fourth coefficient matrix to the energy of the fourth coefficient matrix is ​​less than the third threshold, wherein P6 < P5, P5 and P6 are both integers greater than or equal to 1, and the third threshold is an integer greater than 0 and less than or equal to 1.

[0100] In one possible implementation, the first matrix and / or the second matrix satisfies at least one of the following conditions: 1) the magnitude of at least one element in the first matrix and / or the second matrix is ​​different from the magnitude of at least one other element; 2) at least one row or at least one column in the first matrix and / or the second matrix is ​​a non-geometric sequence.

[0101] In one possible implementation, the third matrix and / or the fourth matrix satisfies at least one of the following conditions: 1) the magnitude of at least one element in the third matrix and / or the fourth matrix is ​​different from the magnitude of at least one other element; 2) at least one row or at least one column in the third matrix and / or the fourth matrix is ​​a non-geometric sequence.

[0102] In one possible implementation, the reporting period of the first coefficient matrix is ​​less than the reporting periods of the first matrix, the second matrix, the third matrix, and the fourth matrix, respectively; or, the reporting period of the first coefficient matrix is ​​less than the reporting periods of the tenth matrix and the eleventh matrix.

[0103] Eleventhly, a signal transmitting device is provided, comprising: a processor coupled to a memory; optionally, it further comprising at least one communication interface and a communication bus; the memory is used to store computer-executable instructions, and the processor, memory, and at least one communication interface are connected via the communication bus; the processor executes the computer-executable instructions stored in the memory to enable the signal transmitting device to implement any one of the methods provided in any of the first to third aspects. The device may exist in the form of a chip product.

[0104] In a twelfth aspect, a channel matrix acquisition device is provided, comprising: a processor coupled to a memory; optionally, it further comprising at least one communication interface and a communication bus; the memory is used to store computer-executable instructions, and the processor, memory, and at least one communication interface are connected via the communication bus. The processor executes the computer-executable instructions stored in the memory to enable the channel matrix acquisition device to implement any one of the methods provided in the fourth or fifth aspect. This device may exist in the form of a chip product.

[0105] In a thirteenth aspect, a communication system is provided, comprising: a signal transmitting device provided in the sixth and seventh aspects; or, a signal transmitting device provided in the eighth aspect and a channel matrix acquisition device provided in the ninth aspect; or, a signal transmitting device provided in the eighth aspect and a channel matrix acquisition device provided in the tenth aspect.

[0106] In a fourteenth aspect, a computer-readable storage medium is provided, including instructions that, when executed on a computer, cause the computer to perform any one of the methods provided in any one of the first to fifth aspects.

[0107] In a fifteenth aspect, a computer program product containing instructions is provided, which, when run on a computer, causes the computer to perform any one of the methods provided in any one of the first to fifth aspects.

[0108] In a sixteenth aspect, a chip is provided, comprising: a processor and an interface, the processor being coupled to a memory via the interface, wherein when the processor executes a computer program or instructions in the memory, any one of the methods provided in any one of the first to fifth aspects is executed.

[0109] For the beneficial effects of the corresponding devices in the above aspects, please refer to the beneficial effects of the corresponding methods, which will not be repeated here. It should be noted that various possible implementations of any of the above aspects can be combined, provided that the solutions do not contradict each other. Attached Figure Description

[0110] Figure 1 This is a schematic diagram of a communication scenario;

[0111] Figure 2 This is a schematic diagram illustrating how the number of CSI-RS changes as the number of ports increases;

[0112] Figure 3 This is a schematic diagram of a scheme to reduce uplink feedback;

[0113] Figure 4 and Figure 5 These are schematic diagrams of a communication scenario provided in the embodiments of this application;

[0114] Figure 6 A flowchart illustrating a signal transmission method provided in an embodiment of this application;

[0115] Figure 7 A schematic diagram of the distribution of CSI-RS on M RBs is provided for an embodiment of this application;

[0116] Figures 8 to 12 Each of the following is a flowchart of a signal transmission method provided in an embodiment of this application;

[0117] Figure 13 A schematic diagram illustrating a spatial sparsity representation provided in an embodiment of this application;

[0118] Figure 14 A schematic diagram illustrating spatial compression provided in an embodiment of this application;

[0119] Figure 15 A schematic diagram of the distribution of CSI-RS after spatial and frequency domain compression, provided for an embodiment of this application;

[0120] Figure 16 A schematic diagram illustrating spatial domain recovery and frequency domain recovery provided in an embodiment of this application;

[0121] Figure 17 A schematic diagram illustrating a frequency domain sparsity representation provided in an embodiment of this application;

[0122] Figure 18 A schematic diagram illustrating frequency domain compression provided in an embodiment of this application;

[0123] Figure 19 A schematic diagram of a simulation result provided for an embodiment of this application;

[0124] Figures 20 to 22 Each of the following is a flowchart of a signal transmission method provided in an embodiment of this application;

[0125] Figure 23 This is a schematic diagram of the composition of a communication device provided in an embodiment of this application;

[0126] Figure 24 This is a schematic diagram of the hardware structure of a communication device provided in an embodiment of this application. Detailed Implementation

[0127] In the description of this application, unless otherwise stated, " / " means "or," for example, A / B can mean A or B. The "and / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone. Furthermore, "at least one" means one or more, and "multiple" means two or more. The terms "first," "second," etc., do not limit the quantity or order of execution, and "first," "second," etc., do not necessarily imply differences.

[0128] It should be noted that, in this application, the terms "exemplary" or "for example" are used to indicate that something is being described as an example, illustration, or illustration. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or design solutions. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0129] The network elements involved in the embodiments of this application include access network equipment and terminals in a communication network, as detailed in the following examples. Figure 1 .

[0130] The communication systems in this application include, but are not limited to, long-term evolution (LTE) systems, 5G systems, NR systems, wireless local area networks (WLAN) systems, and future evolution systems or multiple converged communication systems. For example, the methods provided in this application can be specifically applied to evolved-universal terrestrial radio access network (E-UTRAN) and next-generation radio access network (NG-RAN) systems.

[0131] In this application embodiment, the access network device is a network-side entity used to transmit signals, receive signals, or both. The access network device can be a device deployed in a radio access network (RAN) to provide wireless communication functions for terminals. For example, it can be a transmission reception point (TRP), a base station, various types of control nodes, a road side unit (RSU), etc. The base station can be various types of macro base stations, micro base stations (also called small stations), relay stations, access points (APs), etc. For example, the base station can be an evolved NodeB (eNB or eNodeB), a next-generation node base station (gNB), a next-generation eNB (ng-eNB), a relay node (RN), an integrated access and backhaul (IAB) node, etc. In systems employing different radio access technologies (RAT), the names of devices with base station functions may differ. For example, in an LTE system, it can be called an eNB or eNodeB, and in a 5G or NR system, it can be called a gNB. This application does not limit the specific name of the base station. A control node can connect to multiple base stations and configure resources for multiple terminals covered by multiple base stations. For example, the control node can be a network controller or a radio controller (e.g., a radio controller in a cloud radio access network (CRAN) scenario). Access network equipment can also be access network equipment in a future evolved public land mobile network (PLMN), etc.

[0132] The terminal in this application embodiment can be a user-side entity used to receive signals, or transmit signals, or both receive and transmit signals. The terminal is used to provide users with one or more of voice services and data connectivity services. The terminal can also be referred to as user equipment (UE), terminal equipment, access terminal, user unit, user station, mobile station, remote station, remote terminal, mobile device, user terminal, wireless communication equipment, user agent, or user device. The terminal can be a vehicle-to-everything (V2X) device, such as a smart car, digital car, unmanned car, driverless car, pilotless car, or automobile, self-driving car, or autonomous car, pure electric vehicle (EV), hybrid electric vehicle (HEV), range-extended electric vehicle (REEV), plug-in hybrid electric vehicle (PHEV), or new energy vehicle, etc. Terminals can also be device-to-device (D2D) devices, such as electricity meters and water meters. Terminals can also be mobile stations (MS), subscriber units, drones, Internet of Things (IoT) devices, stations (ST) in WLANs, cellular phones, smartphones, cordless phones, wireless data cards, tablets, session initiation protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistant (PDA) devices, laptop computers, machine-type communication (MTC) terminals, handheld devices with wireless communication capabilities, computing devices or other processing devices connected to a wireless modem, in-vehicle devices, and wearable devices (also known as wearable smart devices).The terminal can also be a terminal in next-generation communication systems, such as a terminal in a 5G system, a terminal in a future PLMN, or a terminal in an NR system.

[0133] In the aforementioned background techniques, besides the inability to reduce the CSI-RS overhead of downlink space-frequency two-dimensional channel matrix estimation, the methods for reducing uplink feedback overhead in the background techniques all use space-frequency bases composed of fixed DFT vectors, which typically make it difficult to represent the downlink space-frequency two-dimensional channel matrix in the sparsest way. Therefore, there is still room for further reduction in uplink feedback overhead.

[0134] To address these issues, this application provides a signal transmission method. The following examples, namely Embodiment 1 and Embodiment 2, illustrate the signal transmission method provided in this application.

[0135] In Example 1, see Figure 4 The access network device reduces the measurement pilot overhead for downlink channel matrix estimation by sending dimensionality-reduced (frequency domain dimensionality reduction and / or spatial domain dimensionality reduction) measurement pilots to the terminal. Additionally, the terminal can directly feed back feedback information to the access network device, indicating the measurement pilot signals received by the terminal, without performing channel estimation, thereby reducing uplink feedback overhead. The measurement pilots in this embodiment are used to measure CSI. For example, the measurement pilots can be CSI-RS or other pilots used for measuring CSI; this application is not limited to these. The downlink channel matrix in this embodiment can be a downlink space-frequency two-dimensional channel matrix or other downlink channel matrices; this application is not limited to these.

[0136] In Example 2, see Figure 5 Access network equipment may not reduce the dimensionality of the measurement pilots, but the terminal can compress the estimated downlink channel matrix (frequency domain compression and / or spatial domain compression) and feed back the compressed downlink channel matrix and the various matrices used to compress the downlink channel matrix to the access network equipment, thereby reducing the overhead of uplink feedback.

[0137] In this application's embodiments, the time unit is a collection of multiple consecutive orthogonal frequency division multiplexing (OFDM) symbols. For example, a time unit can be a minislot, a slot, a subframe, a transmission time interval (TTI), etc. In NR systems, for a normal cyclic prefix (CP), one slot contains 14 OFDM symbols. For extended CP, one slot contains 12 OFDM symbols. A time unit can also be called a time-domain unit, time-domain granularity, etc.

[0138] In this application embodiment, the frequency domain unit can be the frequency domain width of one or more RBs. For example, the frequency domain unit can include the frequency domain width of x RBs. x can be any positive integer. For example, x can be 1, 2, 4, 8, 16, etc. The frequency domain unit can also be one or more subcarriers. For example, the frequency domain unit can include y subcarriers. y can be any positive integer. For example, y can be 1, 12, 60, 120, etc. The frequency domain unit can also be a predefined subband, a band, a bandwidth part (BWP), a component carrier (CC), etc.

[0139] Example 1

[0140] In specific implementation, Implementation Example 1 can be divided into the following scenarios: Scenario 1 (access network device only performs frequency domain compression on the measurement pilot), Scenario 2 (access network device performs both frequency domain compression and spatial domain compression on the measurement pilot), and Scenario 3 (access network device only performs spatial domain compression on the measurement pilot). The signal transmission methods provided by Implementation Example 1 under these three scenarios will be described in detail below.

[0141] Scenario 1: The access network equipment only performs frequency domain compression on the measurement pilot.

[0142] In scenario 1, see Figure 6 The signal transmission method provided in Embodiment 1 includes:

[0143] 601. Access network equipment acquires channel information samples. The channel information samples include channel information from the x-th time unit to the x+T-1-th time unit, where x and T are both integers greater than or equal to 1.

[0144] It should be noted that the multiple time units corresponding to the channel information in the channel information sample can be continuous or discontinuous, and this application does not impose any restrictions. However, this application uses continuous as an example to illustrate the method provided in the embodiments of this application. The channel information sample may also include channel information from other time units, rather than the channel information from the xth time unit to the x+T-1th time unit. This application does not impose any restrictions on this either.

[0145] For example, channel information may include, but is not limited to, any one or more of the following: uplink channel matrix, downlink channel matrix, CSI, channel delay, channel multipath angle information, uplink channel covariance matrix, and downlink channel covariance matrix. CSI includes, but is not limited to, any one or more of the following: precoding matrix indicator (PMI) and rank indication (RI).

[0146] In the specific implementation of step 601, taking the downlink channel matrix of the actual measurement as an example, if the downlink channel matrices of T time units (i.e., from the x-th time unit to the (x+T-1)-th time unit) are respectively denoted as: H1, H2, ..., H T For each frequency domain unit of the s-th time unit (where s is an integer greater than or equal to 1 and less than or equal to T) within T time units, the access network device uses multiple REs (e.g., N different REs, where N is the number of ports of the access network device, and N ports can be a portion of the access network device's port resources) to send measurement pilots to distinguish different ports of the access network device. For example, taking the frequency domain unit as RB, see [link to documentation]. Figure 7 , Figure 7 The example shown uses 24 different REs to transmit measurement pilots on each RB in the s-th time unit of T time units to distinguish 24 different ports of the access network equipment. The terminal measures the measurement pilots on each RB to obtain the downlink channel matrix on each RB. If there are M RBs in each time unit, the terminal concatenates the downlink channel matrices of the M RBs to obtain H. s Afterwards, the terminal can use the methods described in the background technology for uplink feedback and H... s The recovery. Among them, H s The m-th column (m is an integer greater than or equal to 1 and less than or equal to M) corresponds to the m-th RB in the s-th time unit of T time units, H s The nth row (n is an integer greater than or equal to 1 and less than or equal to N) corresponds to the nth port of the access network device in the sth time unit. M is the number of RBs considered in each time unit. The terminal in the embodiments of this application is a single-port or multi-port terminal.

[0147] The process by which access network devices acquire channel information samples can be referred to as initialization.

[0148] 602. The access network equipment determines M′ frequency domain units from the M frequency domain units in the x+T time unit based on the channel information sample. The M′ frequency domain units are used to transmit measurement pilots, where M′ and M are both integers greater than or equal to 1, and M′ < M.

[0149] For example, M frequency domain units are all the frequency domain units in the time unit.

[0150] Optionally, the M′ frequency domain units are determined by the second matrix. The second matrix is ​​used to perform frequency domain compression on the second downlink channel matrix. The second downlink channel matrix is ​​the downlink channel matrix for the (x+T)th time unit. It should be noted that the access network device does not need to actually determine the second downlink channel matrix; therefore, the second downlink channel matrix can be considered as the downlink channel matrix for the (x+T)th time unit predicted or estimated by the access network device.

[0151] The second matrix can be determined based on the first matrix, which is used to perform a frequency-domain sparsified representation of the second downlink channel matrix in the (x+T)th time unit. The first matrix can be determined based on channel information samples.

[0152] 603. The access network equipment sends measurement pilot signals to the terminal on M′ frequency domain units.

[0153] Figure 6 The method shown allows the access network device to determine M′ frequency domain units from the M frequency domain units in the x+T time unit to transmit measurement pilots, thereby achieving frequency domain dimensionality reduction of the measurement pilots. Compared with the existing technology that uses M frequency domain units to transmit measurement pilots, this method can reduce the number of frequency domain units used to transmit measurement pilots, reduce the overhead of transmitting measurement pilots, and thus reduce the measurement pilot overhead of downlink channel matrix estimation.

[0154] In scenario 1, optional, see [link / reference] Figure 8 The method also includes:

[0155] 604. The terminal receives the measurement pilot signal and determines the feedback information.

[0156] Specifically, the terminal receives the measurement pilot signal on the time-frequency resources where the access network equipment sends the measurement pilot.

[0157] The feedback information is used to indicate the measurement pilot signal information received by the terminal. Specifically, for the x+T time unit, it is used to indicate the measurement pilot signal information received by the terminal in the x+T time unit.

[0158] The measurement pilot signal received by the terminal includes N*M′*R elements. Each element represents the measurement pilot signal received at one of the N ports that transmit measurement pilot signals, one of the M′ frequency domain units that transmit measurement pilot signals, and one of the R ports that receive measurement pilot signals. R, N, and M′ are all integers greater than or equal to 1.

[0159] Understandably, N represents the number of ports of the access network device that transmit measurement pilot signals, or the number of ports that transmit measurement pilot signals; M′ represents the number of frequency domain units occupied by the measurement pilot signals; and R represents the number of ports of the terminal that receive measurement pilot signals, or the number of ports that receive measurement pilot signals.

[0160] In one scenario, the feedback information is specifically used to indicate the amplitude and phase information of each element in the measurement pilot signal received by the terminal. For example, the feedback information is specifically used to indicate the absolute amplitude and absolute phase of each element in the measurement pilot signal received by the terminal.

[0161] In another scenario, the measurement pilot signal received by the terminal comprises multiple sets of elements, each set containing more than one element. The feedback information specifically indicates the amplitude and phase information of each set of elements. For example, the feedback information specifically indicates the absolute amplitude and absolute phase of each set of elements.

[0162] 605. The terminal sends feedback information to the access network device. Correspondingly, the access network device receives the feedback information from the terminal.

[0163] 606. The access network equipment determines the downlink channel matrix for the (x+T)th time unit based on the feedback information.

[0164] Optionally, in the specific implementation of step 606, the access network device determines the third downlink channel matrix in the (x+T)th time unit based on the feedback information, the first matrix, and the second matrix. The third downlink channel matrix is ​​the actually measured downlink channel matrix.

[0165] In specific implementation, step 606 involves the access network device performing frequency domain recovery based on feedback information, the first matrix, and the second matrix to obtain the third downlink channel matrix.

[0166] In this alternative method, after the terminal completes the measurement pilot, it does not perform channel estimation and compression feedback. Instead, it directly feeds back the feedback information, which indicates the information of the measurement pilot signal received by the terminal in the x+T time unit, to the access network equipment, thereby reducing the uplink feedback overhead.

[0167] In scenario 1, optional, see [link / reference] Figure 8 Before step 603, the method further includes:

[0168] 602A. The access network device sends first indication information to the terminal. The first indication information indicates the location information of M′ frequency domain units. Correspondingly, the terminal receives the first indication information from the access network device, and the terminal can determine the respective frequency domain units from which the access network device sends the measurement pilots based on the first indication information.

[0169] In this case, step 604 specifically includes: the terminal receiving the measurement pilot signal in the M′ frequency domain unit at the x+T time unit and determining the feedback information.

[0170] Optionally, the first indication information is carried in radio resource control (RRC) signaling, or in medium access control (MAC) control element (MAC CE) signaling, or in downlink control information (DCI).

[0171] Optionally, the access network device sends the first indication information at the x+T time unit. It should be noted that the location information of the M′ frequency domain units can be indicated by the access network device through the first indication information, or it can be agreed upon by both parties according to the communication protocol.

[0172] For example, the first indication information can indicate the indices of M′ frequency domain units. For instance, if M′ = 4, and the indices indicated by the first indication information are 0, 2, 4, and 8, it means that the access network device transmits measurement pilots on frequency domain units 0, 2, 4, and 8. Here, frequency domain unit i refers to the frequency domain unit with index i, where i is an integer greater than or equal to 0.

[0173] In scenario 1, optional, see [link / reference] Figure 8 The method also includes:

[0174] 607. The access network device uses the channel information of p time units as channel information samples to calculate the downlink channel matrix of the (x+T+1)th time unit. Here, p is an integer greater than or equal to 1 and less than or equal to x+T.

[0175] For example, the p time units can be the p time units preceding the (x+T+1)th time unit and the p time units closest to the (x+T+1)th time unit. Of course, they can also be other time units, such as the xth time unit to the (x+T-1)th time unit, and this application does not impose any restrictions.

[0176] The specific implementation of step 607 is similar to the process of determining the third downlink channel matrix in the (x+T)th time unit, and will not be described in detail here.

[0177] Scenario 2: Access network equipment performs frequency domain compression and spatial domain compression on the measurement pilot.

[0178] In scenario 2, see Figure 9 The signal transmission method provided in Embodiment 1 includes:

[0179] 901. Access network equipment acquires channel information samples. The channel information samples include channel information from the x-th time unit to the (x+T-1)-th time unit, where x and T are both integers greater than or equal to 1.

[0180] The relevant description of step 901 can be found in step 601 above, and will not be repeated here.

[0181] 902. The access network equipment determines M′ frequency domain units from the M frequency domain units in the x+T time unit based on the channel information sample. These M′ frequency domain units are used to transmit measurement pilot signals, where M′ and M are both integers greater than or equal to 1, and M′ < M.

[0182] For example, M frequency domain units are all the frequency domain units in the time unit.

[0183] Optionally, the M′ frequency domain units are determined by a second matrix. This second matrix is ​​used to perform frequency domain compression on the first downlink channel matrix. The first downlink channel matrix is ​​obtained by spatially compressing the second downlink channel matrix, which is the downlink channel matrix of the (x+T)th time unit. In this case, the method for spatially compressing the second downlink channel matrix is ​​not limited in this application.

[0184] It should be noted that the access network equipment does not need to actually determine the second downlink channel matrix. Therefore, the second downlink channel matrix can be considered as the downlink channel matrix predicted or estimated by the access network equipment in the (x+T)th time unit.

[0185] The second matrix can be determined based on the first matrix, which is used to perform a frequency-domain sparse representation of the first downlink channel matrix in the (x+T)th time unit. The first matrix can also be determined based on channel information samples. For example, the access network device determines the third matrix based on the channel information samples, then determines the fourth matrix based on the third matrix, and finally determines the first matrix based on the channel information samples and the fourth matrix. The third matrix is ​​used to perform a spatial-domain sparse representation of the second downlink channel matrix. The fourth matrix is ​​used to perform spatial compression on the second downlink channel matrix; the first downlink channel matrix is ​​obtained after spatial compression using the fourth matrix.

[0186] By using the first matrix determined through this example, and then using the second matrix determined thereafter, to determine M′ frequency domain units in the M frequency domain units at the x+T time unit, the number of REs for the transmitted measurement pilot in each of the M′ frequency domain units can be reduced, thus achieving spatial dimensionality reduction of the measurement pilot and further reducing the overhead of transmitting the measurement pilot, thereby further reducing the measurement pilot overhead of downlink channel matrix estimation.

[0187] 903. The access network equipment sends N′ port measurement pilots to the terminal in each of the M′ frequency domain units based on the channel information sample.

[0188] Where N′ is an integer greater than or equal to 1, and N′ < N. Optionally, the measurement pilot transmitted in each of the M′ frequency domain units is the measurement pilot pre-coded by the fourth matrix.

[0189] Figure 9 The method shown allows the access network device to determine M′ frequency domain units from M frequency domain units in the (x+T)th time unit, and transmit N′ port measurement pilots in each of the M′ frequency domain units, achieving frequency domain and spatial domain dimensionality reduction of the measurement pilots. By reducing frequency domain dimensionality, compared to existing technologies that use M frequency domain units to transmit measurement pilots, the number of frequency domain units for transmitting measurement pilots can be reduced, lowering the overhead of transmitting measurement pilots and thus reducing the measurement pilot overhead for downlink channel matrix estimation. By reducing spatial domain dimensionality, the number of REs for transmitting measurement pilots in each of the M′ frequency domain units is further reduced, further lowering the overhead of transmitting measurement pilots and thus further reducing the measurement pilot overhead for downlink channel matrix estimation.

[0190] In scenario 2, optional, see [link / reference] Figure 10 The method also includes:

[0191] 904. The terminal receives the measurement pilot signal and determines the feedback information.

[0192] The feedback information is used to indicate the measurement pilot signal information received by the terminal. Specifically, for the x+T time unit, it is used to indicate the measurement pilot signal information received by the terminal in the x+T time unit.

[0193] The measurement pilot signal received by the terminal includes N′*M′*R elements. Each element represents the measurement pilot signal received by the terminal at one of the N′ ports that transmit measurement pilot signals, one of the M′ frequency domain units that transmit measurement pilot signals, and one of the R ports that receive measurement pilot signals. R, N′, and M′ are all integers greater than or equal to 1.

[0194] Understandably, N′ represents the number of ports of the access network device that transmit measurement pilot signals, or the number of ports that transmit measurement pilot signals; M′ represents the number of frequency domain units occupied by the measurement pilot signals; and R represents the number of ports of the terminal that receive measurement pilot signals, or the number of ports that receive measurement pilot signals.

[0195] In one scenario, the feedback information is specifically used to indicate the amplitude and phase information of each element in the measurement pilot signal received by the terminal. For example, the feedback information is specifically used to indicate the absolute amplitude and absolute phase of each element in the measurement pilot signal received by the terminal.

[0196] In another scenario, the measurement pilot signal received by the terminal comprises multiple sets of elements, each set containing more than one element. The feedback information specifically indicates the amplitude and phase information of each set of elements. For example, the feedback information specifically indicates the absolute amplitude and absolute phase of each set of elements.

[0197] 905. The terminal sends feedback information to the access network device. Correspondingly, the access network device receives the feedback information from the terminal.

[0198] 906. The access network equipment determines the downlink channel matrix for the x+Tth time unit based on the feedback information.

[0199] Optionally, in the specific implementation of step 906, the access network device determines the third downlink channel matrix in the (x+T)th time unit based on the feedback information, the first matrix, the second matrix, the third matrix, and the fourth matrix. The third downlink channel matrix is ​​the actually measured downlink channel matrix. Specifically, the access network device can perform frequency domain recovery based on the feedback information, the first matrix, and the second matrix to obtain the spatially compressed third downlink channel matrix, and then perform spatial domain recovery based on the third matrix and the fourth matrix to obtain the third downlink channel matrix.

[0200] In this alternative method, after measuring the measurement pilot, the terminal does not perform channel estimation, but instead directly feeds back the feedback information of the x+T time unit, which indicates the information of the measurement pilot signal received by the terminal in the x+T time unit, to the access network device, thereby reducing the uplink feedback overhead.

[0201] In scenario 2, optional, see [link / reference] Figure 10 Before step 903, the method further includes:

[0202] 902A. The access network device sends first indication information to the terminal. The first indication information indicates M′ frequency domain elements. Correspondingly, the terminal receives the first indication information from the access network device, and the terminal can determine the respective frequency domain elements from which the access network device sends the measurement pilot based on the first indication information. A description related to step 902A can be found in step 602A above.

[0203] 902B. The access network device sends a second indication information to the terminal. This second indication information indicates the value of N′. Correspondingly, the terminal receives the second indication information from the access network device and can determine the number of ports from which the access network device sends the measurement pilot based on this information.

[0204] In this case, step 904 specifically includes: the terminal receiving the measurement pilot signal on the N′ ports of each frequency domain unit in the M′ frequency domain units, and determining the feedback information.

[0205] For a description of the first instruction information, please refer to the above text, and it will not be repeated here.

[0206] Optionally, the second indication information is carried in RRC signaling, or MAC CE signaling, or DCI.

[0207] Optionally, the access network device sends the second indication information at the x+T time unit. It should be noted that the value of N′ can be indicated by the access network device through the second indication information, or it can be predetermined by both parties according to the communication protocol.

[0208] The first instruction information and the second instruction information can be carried in the same message or in different messages; this application does not impose any restrictions.

[0209] In scenario 2, optional, see [link / reference] Figure 10 The method also includes:

[0210] 907. The access network device uses the channel information from p time units as channel information samples to calculate the downlink channel matrix for the (x+T+1)th time unit. Here, p is an integer greater than or equal to 1 and less than or equal to x+T. A description related to step 907 can be found in step 607 above.

[0211] Scenario 3: Access network equipment only performs spatial compression on the measurement pilot.

[0212] In scenario 3, see Figure 11 The signal transmission method provided in Embodiment 1 includes:

[0213] 1101. Access network equipment acquires channel information samples. The channel information samples include channel information from the x-th time unit to the x+T-1-th time unit, where x and T are both integers greater than or equal to 1.

[0214] The relevant description of step 1101 can be found in step 601 above, and will not be repeated here.

[0215] 1102. The access network device sends N′ port measurement pilots to the terminal in each of the M frequency domain units in the x+T time unit based on the channel information sample.

[0216] In a specific implementation, step 1102 may include: the access network device determining a third matrix based on the channel information samples, determining a fourth matrix based on the third matrix, using the fourth matrix to precode the measurement pilot in each of the M frequency domain units at the (x+T)th time unit, and transmitting the precoded measurement pilot in each of the M frequency domain units. The third matrix is ​​used to perform spatially sparse representation of the second downlink channel matrix. The fourth matrix is ​​used to perform spatial compression on the second downlink channel matrix. The second downlink channel matrix is ​​the downlink channel matrix at the (x+T)th time unit.

[0217] It should be noted that the access network equipment does not need to actually determine the second downlink channel matrix. Therefore, the second downlink channel matrix can be considered as the downlink channel matrix predicted or estimated by the access network equipment in the (x+T)th time unit.

[0218] Optionally, the measurement pilot transmitted in each of the M frequency domain units is the measurement pilot that has been pre-coded by the fourth matrix.

[0219] Figure 11 The method shown allows the access network device to achieve spatial dimensionality reduction of the measurement pilot, which reduces the number of REs for transmitting the measurement pilot in each of the M frequency domain units, thereby reducing the overhead of transmitting the measurement pilot and thus reducing the measurement pilot overhead for downlink channel matrix estimation.

[0220] In scenario 3, optional, see [link / reference] Figure 12 The method also includes:

[0221] 1103. The terminal receives the measurement pilot signal on the time-frequency resources for sending the measurement pilot and determines the feedback information.

[0222] The feedback information is used to indicate the measurement pilot signal information received by the terminal. Specifically, for the x+T time unit, it is used to indicate the measurement pilot signal information received by the terminal in the x+T time unit.

[0223] The measurement pilot signal received by the terminal includes N′*M*R elements. Each element represents the measurement pilot signal received by the terminal at one of the N′ ports that transmit measurement pilot signals, one of the M frequency domain units that transmit measurement pilot signals, and one of the R ports that receive measurement pilot signals. R, N′, and M are all integers greater than or equal to 1.

[0224] In one scenario, the feedback information is specifically used to indicate the amplitude and phase information of each element in the measurement pilot signal received by the terminal. For example, the feedback information is specifically used to indicate the absolute amplitude and absolute phase of each element in the measurement pilot signal received by the terminal.

[0225] In another scenario, the measurement pilot signal received by the terminal comprises multiple sets of elements, each set containing more than one element. The feedback information specifically indicates the amplitude and phase information of each set of elements. For example, the feedback information specifically indicates the absolute amplitude and absolute phase of each set of elements.

[0226] 1104. The terminal sends feedback information to the access network device. Correspondingly, the access network device receives the feedback information from the terminal.

[0227] 1105. The access network equipment determines the downlink channel matrix for the (x+T)th time unit based on the feedback information.

[0228] In specific implementation, step 1105 involves the access network device determining the third downlink channel matrix for the (x+T)th time unit based on the feedback information, the third matrix, and the fourth matrix. The third downlink channel matrix is ​​the actually measured downlink channel matrix.

[0229] In this alternative method, after measuring the measurement pilot, the terminal does not perform channel estimation, but instead directly feeds back the feedback information, which indicates the information of the measurement pilot signal received by the terminal in the x+T time unit, to the access network equipment, thereby reducing the overhead of uplink feedback.

[0230] Optionally, step 1105 may include the following in a specific implementation: the access network device performs spatial recovery based on the feedback information, the third matrix, and the fourth matrix to obtain the third downlink channel matrix.

[0231] In scenario 3, optional, see [link / reference] Figure 12 Before step 1102, the method further includes:

[0232] 1101A. The access network device sends a second indication information to the terminal. This second indication information indicates the value of N′. Correspondingly, the terminal receives the second indication information from the access network device and can determine the number of ports from which the access network device sends the measurement pilot signal based on this information.

[0233] In this case, step 1103 specifically includes: the terminal receiving the measurement pilot signal on the N′ ports of each of the M frequency domain units and determining the feedback information.

[0234] In scenario 3, optional, see [link / reference] Figure 12 The method also includes:

[0235] 1106. The access network device uses the channel information of p time units as channel information samples to calculate the downlink channel matrix of the (x+T+1)th time unit. Here, p is an integer greater than or equal to 1 and less than or equal to x+T. A description related to step 1106 can be found in step 607 above.

[0236] In the above three scenarios, the access network device can update each matrix periodically, or update each matrix when the channel quality is poor, or update each matrix under other circumstances. This application does not impose any restrictions.

[0237] The following example, using scenario 2, illustrates the determination method for each matrix. For ease of description, the following example uses the downlink channel matrix of a time unit to represent the channel information of that time unit, and each time unit's downlink channel matrix is ​​an N*M matrix (N is an integer greater than or equal to 1) to illustrate the determination method for each matrix in the above method. Specifically, it is illustrated in case 1 (determining each matrix for all frequency domain units in the time unit) and case 2 (determining each matrix individually for each frequency domain unit in the time unit).

[0238] Case 1: Determine each matrix for all frequency domain units in the time unit.

[0239] In case 1, optionally, the third matrix satisfies the following condition: the F-norm of the difference between the product of the third and fifth matrices and the sixth matrix is ​​minimized.

[0240] The sixth matrix is ​​an N*(M*T) matrix. The column vectors in the sixth matrix are composed of all the column vectors in the downlink channel matrix from the x-th time unit to the x+T-1-th time unit. The column vectors in the fifth matrix correspond one-to-one with the column vectors in the sixth matrix. A column vector in the fifth matrix is ​​a sparse representation of the corresponding column vector in the sixth matrix. The number of non-zero elements in each column vector in the fifth matrix is ​​equal to S1, where S1 is an integer greater than or equal to 1 and less than N.

[0241] When determining the third matrix, multiple candidate third matrices can be determined based on the fifth matrix. The third matrix required in this application is the one among the multiple candidate third matrices that minimizes the aforementioned F-norm.

[0242] Let the sixth matrix be denoted as but h m，s For H s The m-th column.

[0243] Let the fifth matrix be denoted as X, then X = [x 1，1 x 2，1 , ..., xM，1 x 1，2 x 2，2 , ..., x M，2 x 1，3 , ..., x M，T ], x m，s for h m，s The corresponding sparse vector X satisfies Let x represent any value of s and m. m，s The number of non-zero elements in S1 is equal to S1. S1 can be predefined, preset, protocol-specified, or determined through negotiation between the access network equipment and the terminal; this application does not impose any restrictions on it.

[0244] Let the third matrix be Ψ D ,but

[0245] in, express The F-norm, Ψ is Ψ D A feasible solution. The algorithm for solving Equation 11 includes, but is not limited to, the classic K-singular value decomposition (KSVD) algorithm. Ψ can be obtained by solving Equation 1.1. D And X. As can be seen from Formula 11, Ψ D Its main function is to Each column h m，s Perform sparsity representation, i.e., h m，s =Ψ D x m，s .

[0246] For example, assume S1 = 2, see [link to relevant documentation]. Figure 13 Regarding h m，s It can be done through Ψ D Sparsification to x m，s .

[0247] In case 1, optionally, the fourth matrix satisfies the following condition: the matrix obtained by multiplying the fourth matrix and the third matrix (which can be denoted as the seventh matrix) has the minimum column correlation.

[0248] The column correlation of a matrix refers to the maximum value of the modulus of the inner product of any two columns in the matrix.

[0249] When determining the fourth matrix, multiple fourth matrices can be obtained for the third matrix. The fourth matrix that minimizes the column correlation of the seventh matrix is ​​the fourth matrix required in this application.

[0250] Let the fourth matrix be Φ. D Then the seventh matrix is ​​Φ D *ΨD .

[0251] For example, based on the irrelevance properties of compressed sensing theory, Φ D When the following formula 1.2 is satisfied, the column correlation of the seventh matrix can be minimized.

[0252]

[0253] Where I is the identity matrix, express The F-norm, Φ is Φ D One feasible solution. For Ψ D The conjugate transpose of Φ H Let be the conjugate transpose of Φ.

[0254] The algorithms for solving Equation 1.2 include, but are not limited to, the classic principal component elimination algorithm. Φ D Its main function is to Each column h m，s Compression yields an N′*1 dimensional observation vector y (N′<<N, where “<<” indicates “much smaller”). m，s y m，s =Φ D h m，s =Φ D Ψ D x m，s .

[0255] For example, see Figure 14 , through Φ D You can use the 8 lines of h m，s Compress it to 4 lines.

[0256] Formula 1.2 guarantees that the observed vector y m，s Able to accurately recover sparse vector x m，s Thus, the accurate h is obtained. m，s =Ψ D x m，s .

[0257] In case 1, optionally, the first matrix satisfies the following condition: the F-norm of the matrix difference between the product of the first matrix and the eighth matrix and the conjugate transpose of the ninth matrix is ​​minimized.

[0258] The ninth matrix is ​​determined based on the channel information samples and the fourth matrix. The m-th column vector of the ninth matrix is ​​composed of T column vectors arranged from top to bottom. The s-th column vector among the T column vectors is the product of the m-th column of the s-th downlink channel matrix and the fourth matrix. The s-th downlink channel matrix is ​​the downlink channel matrix of the x+s-1 time unit from the x-th time unit to the x+T-1-th time unit. m is an integer greater than or equal to 1 and less than or equal to M, and s is an integer greater than or equal to 1 and less than or equal to T.

[0259] In this matrix, the column vectors in the eighth matrix correspond one-to-one with the column vectors in the conjugate transpose of the ninth matrix. A column vector in the eighth matrix is ​​a sparse representation of the corresponding column vector in the conjugate transpose of the ninth matrix. The number of non-zero elements in each column vector in the eighth matrix is ​​equal to S2, where S2 is an integer greater than or equal to 1 and less than M.

[0260] In this matrix, the column vectors in the eighth matrix correspond one-to-one with the column vectors in the conjugate transpose of the ninth matrix. A column vector in the eighth matrix is ​​a sparse representation of the corresponding column vector in the conjugate transpose of the ninth matrix. The number of non-zero elements in each column vector of the eighth matrix is ​​equal to S2, where S2 is an integer less than M.

[0261] When determining the first matrix, multiple candidate first matrices can be determined based on the eighth matrix. The first matrix required in this application is the matrix that minimizes the aforementioned F-norm among the multiple candidate first matrices.

[0262] Let Y be the conjugate transpose of the ninth matrix. It is understandable that Y is an M*N′T dimensional matrix, and each row of Y corresponds to a frequency domain unit.

[0263] Let the eighth matrix be denoted as W, then W = [w1, w2, ..., w N′T ], where w k Let W be a sparse representation of the k-th column vector in Y, where k is an integer greater than or equal to 1 and less than or equal to N′T. W satisfies... For any value of k, w represents k The number of non-zero elements in S2 is equal to the number of non-zero elements in S2.

[0264] Let the first matrix be Ψ f , then Ψ f When the following formula 1.3 is satisfied, Ψ can be made f The F-norm of the product of W and the matrix difference of Y is minimized.

[0265]

[0266] in, Let F be the norm of “Y-ΨW”, where Ψ is Ψ f A feasible solution. The solution algorithm for Equation 1.3 includes, but is not limited to, the classic KSVD algorithm. Ψ f Its main function is to sparsely represent each column of Y.

[0267] In case 1, optionally, the second matrix satisfies the following conditions: the column correlation of the matrix obtained by multiplying the second matrix and the first matrix is ​​minimal; the second matrix is ​​a row extraction matrix; each row of the second matrix has only one non-zero element; and the non-zero elements in different rows are in different positions.

[0268] When determining the second matrix, multiple row extraction matrices can be obtained from the first matrix. The multiple row extraction matrices and the first matrix are multiplied to obtain multiple matrices. The row extraction matrix that minimizes the column correlation of the multiple matrices is the second matrix required by this application.

[0269] Let the second matrix be Φ f ,but Φ f satisfy Φ is limited f For a target Ψ f The row extraction matrix, specifically for any value of i, The number of non-zero elements in Φ is 1, and the rank of Φ is M′. It refers to Φ f The i-th row, Φ f The non-zero elements in each row are in different positions, i is an integer greater than or equal to 1 and less than or equal to M′, M′ << M. I is the identity matrix. Indicates "1- The F-norm, Φ is Φ f One feasible solution.

[0270] Formula 1.4 means that Φ f Ψ f The column correlation should be minimized to ensure that the sparse vector can still be recovered from the compressed vector.

[0271] It is worth noting that Formula 1.4 is a combination problem, which is difficult to solve efficiently. Therefore, this application proposes a greedy search algorithm to obtain a near-optimal solution with low complexity. Specifically, firstly, Ψ is extracted... f Let the first row of a given row be the row extraction set U = {p1}. Then, for Ψ... fThe remaining rows are searched, and the row that minimizes the cost function in Formula 1.4 is selected. Let's say it's the second row, then U is updated to U = {p1, p2}. This process is repeated M′-1 times to obtain the complete row extraction set U = {p1, p2, ..., p...}. M′ It should be noted that the position of each element in U is not restricted in this application, p m′ It refers to Φ f The non-zero elements in the m′-th row of the array are given by U, where m′ is an integer greater than or equal to 1 and less than or equal to M′. The near-optimal Φ can be obtained based on U. f For example, if M = 5, M′ = 3, U = {p1 = 4, p2 = 3, p3 = 1}, then Φ f The first row, fourth column contains non-zero elements; the second row, third column contains non-zero elements; and the third row, first column contains non-zero elements. Assume the non-zero element is 1.

[0272] In scenario 1, step 902, in its specific implementation, involves the access network device determining the value based on Φ. f (Or equivalently, M′) Select M′ frequency domain units in the x+T time unit to send measurement pilot signals, achieving frequency domain dimensionality reduction, and subsequently inform the terminal of M′ through the first indication information. Specifically, the M′ frequency domain units are determined based on the positions of the non-zero elements in the second matrix.

[0273] For example, taking the frequency domain unit as RB, if See Figure 15 The access network equipment can select 3 RBs from the 5 RBs to send measurement pilots. These 3 RBs are RB1, RB3 and RB4, thereby reducing the overhead of sending measurement pilots.

[0274] In case 1, optionally, step 903 may include the following in its implementation:

[0275] 21) The access network equipment uses the fourth matrix to precode the measurement pilots on each of the M′ frequency domain units.

[0276] 22) The access network equipment transmits precoded measurement pilots in each of the M′ frequency domain units.

[0277] For example, see Figure 15 By precoding the measurement pilots on RB1, RB3, and RB4, the number of REs transmitting the measurement pilots on RB1, RB3, and RB4 can be reduced, thereby further reducing the overhead of transmitting the measurement pilots. It is understood that the space-frequency downsampling measurement pilots transmitted by the access network equipment have a new measurement pilot pattern.

[0278] In step 21), it is assumed that the precoding vector used by the access network device for the n′-th measurement pilot transmitted in each of the M′ frequency domain units is f. D，n′ , and f D，n′ For Φ D The n′ row. Then, after all N′*M′ measurement pilots on the extracted M′ frequency domain units are transmitted, the N′*M′ dimensional measurement signal obtained by the terminal can be expressed as:

[0279] Z x+T =Φ f [Φ D h 1，x+T , Φ D h 2，x+T , …, Φ D h M，x+T ] H +N=Φ f V x+T +N (Formula 15).

[0280] Among them, h m，x+T Let N represent the channel vector in the m-th frequency domain unit at the x+T-th time unit, and N be the noise matrix.

[0281] In this case, in the specific implementation of step 905, the terminal may not perform channel estimation, but instead use Z... x+T Directly quantified ( (That is, feedback information), and then feed it back to the access network equipment.

[0282] Of course, step 905 can also be implemented in other ways. For example, the terminal can find a value related to Z from the codebook. x+T The closest codeword (i.e., the one with the smallest F-norm difference between the two) is selected, and then the sequence number of this codeword is fed back to the access network device. For example, the terminal can feed back Z to the access network device. x+T The amplitude and phase of one element (e.g., the element with the largest amplitude and phase), and the relative amplitude and relative phase of other elements relative to that element, are not specifically limited in this application.

[0283] Optionally, step 906, in its specific implementation, includes: the access network device receiving... Then, the N*M dimensional space-frequency two-dimensional downlink channel matrix H can be obtained through frequency domain recovery and spatial domain recovery. x+T .like Figure 16 As shown, V x+T Similar to Y, if the characteristics of the downlink channel exhibit slow-varying properties in the time domain, then V x+T It should still be possible to pass through Ψ fSparse representation is then performed. Therefore, classic sparse signal reconstruction algorithms such as Orthogonal Matching Pursuit (OMP) can be used, based on... Restore V x+T =[Φ D h 1，x+T , Φ D h 2，x+T , …, Φ D h M，x+T ] H Then, also based on the slow-varying characteristics of the downlink channel in the time domain, h m，x+T It should still be possible to pass through Ψ D Sparse representation is then performed. Therefore, we can use classic sparse signal reconstruction algorithms such as OMP, based on Φ. D h m，x+T Restore h m，x+T Finally, the high-dimensional downlink channel matrix H can be obtained. x+T =[h 1，x+T h 2，x+T , ..., h M，x+T ].

[0284] Case 2: Determine each matrix separately for each frequency domain unit in the time unit.

[0285] In case 2, optionally, the third matrix includes M third sub-matrices, and the m-th third sub-matrix among the M third sub-matrices satisfies the following condition: the F-norm of the matrix difference between the product of the m-th third sub-matrix and the m-th fifth sub-matrix and the m-th sixth sub-matrix is ​​minimized, where m is an integer greater than or equal to 1 and less than or equal to M.

[0286] The m-th sixth submatrix is ​​an N*T matrix. The column vectors in the m-th sixth submatrix are composed of the m-th column vectors in the downlink channel matrix from the x-th time unit to the x+T-1-th time unit. The column vectors in the m-th fifth submatrix correspond one-to-one with the column vectors in the m-th sixth submatrix. A column vector in the m-th fifth submatrix is ​​a sparse representation of the corresponding column vector in the m-th sixth submatrix. The number of non-zero elements in each column vector of the m-th fifth submatrix is ​​equal to S1, where S1 is an integer greater than or equal to 1 and less than N.

[0287] When determining the m-th third submatrix, multiple candidate m-th third submatrixes can be determined based on the m-th fifth submatrix. The matrix among the multiple candidate m-th third submatrixes that minimizes the aforementioned F-norm is the m-th third submatrix required in this application.

[0288] Let the m-th sixth submatrix be denoted as but hm，s For H s The m-th column.

[0289] Let the m-th fifth submatrix be denoted as X. m Then X m =[x m，1 x m，2 , ..., x m，T ], x m，s for h m，s The corresponding sparse vector, X m satisfy For any value of s, x represents m，s The number of non-zero elements in S1 is equal to S1. S1 can be predefined, preset, protocol-specified, or determined through negotiation between the access network equipment and the terminal; this application does not impose any restrictions on it.

[0290] Let the m-th third submatrix be denoted as Ψ. m ,but

[0291] in, express The F-norm, Ψ is Ψ m A feasible solution. The solution algorithm for Equation 2.1 includes, but is not limited to, the classic KSVD algorithm. Ψ can be obtained by solving Equation 2.1. m and X m As can be seen from Formula 2.1, Ψ m Its main function is to Each column h m，s Perform sparsity representation, i.e., h m，s =Ψ m x m，s .

[0292] For example, assume S1 = 2, see [link to relevant documentation]. Figure 17 Regarding h m，s It can be done through Ψ m Sparsification to x m，s .

[0293] In case 2, optionally, the fourth matrix includes M fourth sub-matrices. The m-th fourth sub-matrix among the M fourth sub-matrices satisfies the following condition: the column correlation of the m-th seventh sub-matrix obtained by multiplying the m-th fourth sub-matrix with the m-th third sub-matrix in the third matrix is ​​minimal. m′ is an integer greater than or equal to 1 and less than or equal to M′. Wherein, the fourth matrix includes M fourth sub-matrices corresponding one-to-one with the M frequency domain units. In this case, the measurement pilot transmitted on the m′-th frequency domain unit among the M′ frequency domain units is the measurement pilot after precoding by the m′-th fourth sub-matrix among the M′ fourth sub-matrices in the fourth matrix. The M′ fourth sub-matrices are the fourth sub-matrices corresponding to the M′ frequency domain units.

[0294] The column correlation of a matrix refers to the maximum value of the modulus of the inner product of any two columns in the matrix.

[0295] Let the m-th fourth submatrix be Φ. m Then the m-th seventh submatrix is ​​Φ m *Ψ m .

[0296] For example, based on the irrelevance properties of compressed sensing theory, Φ m When the following formula 22 is satisfied, the column correlation of the m-th seventh submatrix can be minimized.

[0297]

[0298] Where I is the identity matrix, express The F-norm, Φ is Φ m One feasible solution. For Ψ m The conjugate transpose of Φ H Let be the conjugate transpose of Φ.

[0299] The algorithms for solving Equation 2.2 include, but are not limited to, the classic principal component elimination algorithm. Φ m Its main function is to Each column h m，s Compression yields an N′*1 dimensional observation vector y (N′<<N, where “<<” indicates “much smaller”). m，s y m，s =Φ m h m，s =Φ m Ψ m x m，s .

[0300] For example, see Figure 18 , through Φ m You can use the 8 lines of h m，sCompress it to 4 lines.

[0301] Formula 22 guarantees that the observed vector y m，s Able to accurately recover sparse vector x m，s Thus, the accurate h is obtained. m，s =Ψ m x m，s .

[0302] In case 2, optionally, the first matrix satisfies the following condition: the F-norm of the matrix difference between the product of the first matrix and the eighth matrix and the conjugate transpose of the ninth matrix is ​​minimized.

[0303] The ninth matrix is ​​determined based on the channel information sample and the fourth matrix. The m-th column vector of the ninth matrix is ​​composed of T column vectors arranged from top to bottom. The s-th column vector among the T column vectors is the product of the m-th column of the s-th downlink channel matrix and the m-th fourth sub-matrix. The s-th downlink channel matrix is ​​the downlink channel matrix of the x+s-1 time unit from the x-th time unit to the x+T-1-th time unit, where s is an integer greater than or equal to 1 and less than or equal to T.

[0304] In this matrix, the column vectors in the eighth matrix correspond one-to-one with the column vectors in the conjugate transpose of the ninth matrix. A column vector in the eighth matrix is ​​a sparse representation of the corresponding column vector in the conjugate transpose of the ninth matrix. The number of non-zero elements in each column vector in the eighth matrix is ​​equal to S2, where S2 is an integer greater than or equal to 1 and less than M.

[0305] When determining the first matrix, multiple candidate first matrices can be determined based on the eighth matrix. The first matrix required in this application is the matrix that minimizes the aforementioned F-norm among the multiple candidate first matrices.

[0306] Let Y be the conjugate transpose of the ninth matrix. It is understandable that Y is an M*N′T dimensional matrix, and each row of Y corresponds to a frequency domain unit.

[0307] Let the eighth matrix be denoted as W, then W = [w1, w2, ..., w N′T ], where w k Let W be a sparse representation of the k-th column vector in Y, where k is an integer greater than or equal to 1 and less than or equal to N′T. W satisfies... For any value of k, w represents k The number of non-zero elements in S2 is equal to the number of non-zero elements in S2.

[0308] Let the first matrix be Ψ f , then Ψ fWhen the following formula 2.3 is satisfied, Ψ can be made f The F-norm of the product of W and the matrix difference of Y is minimized.

[0309]

[0310] in, Let F be the norm of “Y-ΨW”, where Ψ is Ψ f A feasible solution. The algorithm for solving Equation 23 includes, but is not limited to, the classic KSVD algorithm. Ψ f Its main function is to sparsely represent each column of Y.

[0311] In case 2, optionally, the second matrix satisfies the following conditions: the column correlation of the matrix obtained by multiplying the second matrix and the first matrix is ​​minimal; the second matrix is ​​a row extraction matrix; each row of the second matrix has only one non-zero element; and the non-zero elements in different rows are in different positions.

[0312] The method for determining the second matrix in case 2 is the same as in case 1, as detailed above, and will not be repeated here.

[0313] In case 2, the specific implementation of step 902 is the same as in case 1.

[0314] In case 2, optionally, step 903 may include the following in its implementation:

[0315] 31) The access network equipment uses the m′-th fourth submatrix (hereinafter denoted as Φ) of the M′ fourth submatrixes in the fourth matrix. m ′), pre-encode the measurement pilot on the m′-th frequency domain unit in M′ frequency domain units to achieve spatial dimensionality reduction, and the M′ fourth sub-matrix is ​​the fourth sub-matrix corresponding to the M′ frequency domain units.

[0316] 32) The access network equipment transmits precoded measurement pilots in each of the M′ frequency domain units.

[0317] For example, see Figure 15 By precoding the measurement pilots on RB1, RB3, and RB4, the number of REs transmitting the measurement pilots on RB1, RB3, and RB4 can be reduced, thereby further reducing the overhead of transmitting the measurement pilots. It is understood that the space-frequency downsampling measurement pilots transmitted by the access network equipment have a new measurement pilot pattern.

[0318] In step 31), it is assumed that the precoding vector used by the access network device for the n′-th measurement pilot transmitted in the m′-th frequency domain unit out of M′ frequency domain units is f. m′，n′ , and f m′，n′ For Φm′ The n′ row. Then, after all N′*M′ measurement pilots on the extracted M′ frequency domain units are transmitted, the N′*M′ dimensional measurement signal obtained by the terminal can be expressed as:

[0319] Z x+T =Φ f [Φ1h 1，x+T ,Φ2h 2，x+T , …, Φ M h M，x+T ] H +N=Φ f V x+T +N (Formula 25).

[0320] Among them, h m，x+T Let N represent the channel vector in the m-th frequency domain unit at the x+T-th time unit, and N be the noise matrix.

[0321] In this case, in the specific implementation of step 905, the terminal does not perform channel estimation, but instead uses Z... x+T Directly quantified ( (That is, feedback information), and then feed it back to the access network equipment.

[0322] Of course, step 905 can also be implemented in other ways. For example, the terminal can find a value related to Z from the codebook. x+T The closest codeword (i.e., the one with the smallest F-norm difference between the two) is selected, and then the sequence number of this codeword is fed back to the access network device. For example, the terminal can feed back Z to the access network device. x+T The amplitude and phase of one element (e.g., the element with the largest amplitude and phase), and the relative amplitude and phase of other elements relative to that element, are not specifically limited in this application. Optionally, step 906, in a specific implementation, includes: the access network device receiving... Then, the N×M dimensional space-frequency two-dimensional downlink channel matrix H can be obtained through frequency domain recovery and spatial domain recovery. x+T .like Figure 16 As shown, V x+T Similar to Y, if the characteristics of the downlink channel exhibit slow-varying properties in the time domain, then V x+T It should still be possible to pass through Ψ f Sparse representation is then performed. Therefore, classic sparse signal reconstruction algorithms such as OMP can be used, based on... Restore V x+T =[Φ1h 1，x+T ,Φ2h 2，x+T , …, Φ M h M，x+T ] H Then, also based on the slow-varying characteristics of the downlink channel in the time domain, hm，x+T It should still be possible to pass through Ψ m Sparse representation is then performed. Therefore, we can use classic sparse signal reconstruction algorithms such as OMP, based on Φ. m h m，x+T Restore h m，x+T Finally, the high-dimensional downlink channel matrix H can be obtained. x+T =[h 1，x+T h 2，x+T , ..., h M，x+T ].

[0323] When compressed sensing technology is used to implement the method described in this application, the first to fourth matrices can also be referred to as: frequency domain sensing matrix, frequency domain compression matrix, spatial domain sensing matrix, and spatial domain compression matrix, respectively. Furthermore, the access network device can also send indication information to the terminal, thereby instructing the terminal whether to use the conventional method for downlink channel matrix estimation or the method provided in this application. The terminal can then provide corresponding uplink feedback based on this indication information.

[0324] For example, taking scenario 2 as an example, with N=64, M=128, T=10, N′=4, M′=24, if the channel model adopts clustered delay line-B (CDL-B), see [link to relevant documentation]. Figure 19 Using normalized mean square error (NMSE) as a performance metric, the method provided in this application is used to acquire the downlink channel matrix for 100 time slots. Under the same signal-to-noise ratio (SNR) (in dB), the accuracy of the method provided in this application (shown by the dashed line) is higher than that of the prior art (shown by the dotted line). In other words, under the same measurement pilot overhead and uplink feedback overhead, the method provided in this application can significantly improve the estimation accuracy. Conversely, at the same accuracy, the overhead of downlink channel matrix acquisition (including measurement pilot overhead and uplink feedback overhead) can be significantly reduced.

[0325] For scenario 1, there is no need to distinguish between different frequency domain units. The access network device can use formula 13 to calculate the first matrix. The difference is that when determining Y, Φ is not required. D ,Right now Access network devices can use Formula 14 to calculate the second matrix.

[0326] For scenario 1, the specific implementation of step 602 is similar to that of step 902, the specific implementation of step 605 is similar to that of step 905, and the specific implementation of step 606 is similar to that of step 906, except that spatial domain restoration is not performed, so it will not be described in detail here.

[0327] For scenario 2, in another implementation, the access network device can first determine the first and second matrices, and then determine the third and fourth matrices based on the second matrix. Correspondingly, during recovery, spatial domain recovery can be performed first, followed by frequency domain recovery. In this case, it is not necessary to distinguish between different frequency domain units, and the access network device can use formula 1.3 to calculate the first matrix. The difference is that when determining Y, Φ is not required. D ,Right now Access network devices can use Formula 1.4 to calculate the second matrix; access network devices can use Formula 11 to determine the third matrix, the difference being that... Among them, h n，s H represents s The nth row; the access network device can use formula 1.2 to determine the fourth matrix.

[0328] For scenario 3, the method for determining the third and fourth matrices under case 1 above is also applicable to scenario 3. The method for determining the third and fourth matrices under case 2 above, as well as steps 31) and 32), is also applicable to scenario 3. Furthermore, in cases 1 and 2, the specific implementation of step 1104 is similar to step 905. In the specific implementation of step 1105, compared to step 906, frequency domain recovery is not performed, and will not be elaborated further.

[0329] The main idea of ​​Implementation Example 1 is to utilize a sparse learning algorithm to first represent the downlink channel matrix of the (x+T)th time unit more sparsely in the spatial and / or frequency domains, thereby reducing the measurement pilot overhead required by the access network device. Then, after measuring the measurement pilot, the terminal does not perform channel estimation but directly feeds back feedback information indicating the measurement pilot signal received by the terminal in the (x+T)th time unit to the access network device. The access network device receives the feedback information from the terminal containing the measurement pilot carrying channel information and, using a compressed sensing algorithm, recovers the actual downlink channel matrix of the (x+T)th time unit through a small number of measurements.

[0330] Example 2

[0331] In specific implementation, Example 2 can be divided into the following scenarios: Scenario 1 (the terminal performs only one type of compression (e.g., frequency domain compression or spatial domain compression) on the downlink channel matrix actually measured in one or more time units) and Scenario 2 (the terminal performs two types of compression (e.g., frequency domain compression and spatial domain compression) on the downlink channel matrix actually measured in one or more time units). The signal transmission method provided by Example 2 under these two scenarios will be described in detail below.

[0332] Scenario 1: The terminal performs a compression (e.g., frequency domain compression or spatial domain compression) only on the downlink channel matrix of the actual measurements for one or more time units.

[0333] In scenario 1, see Figure 20 The signal transmission method provided in Embodiment 2 includes:

[0334] 2001. The terminal acquires channel information samples, which include the first downlink channel matrix.

[0335] Optionally, the first downlink channel matrix can be a downlink channel matrix, the covariance matrix of the downlink channel matrix, or the characteristic matrix of the downlink channel matrix.

[0336] For example, the channel information sample includes T first downlink channel matrices, which are the first downlink channel matrices from the x-th time unit to the x+T-1-th time unit, where x and T are both integers greater than or equal to 1.

[0337] It should be noted that the multiple time units corresponding to the T first downlink channel matrices can be continuous or discontinuous, and this application does not impose any restrictions. However, this application uses continuous as an example to illustrate the method provided in the embodiments of this application. The channel information sample may also include the first downlink channel matrices of other time units, rather than the first downlink channel matrices of the xth time unit to the x+T-1th time unit. This application does not impose any restrictions.

[0338] If the first downlink channel matrix for T time units (i.e., from the x-th time unit to the (x+T-1)-th time unit) is denoted as H1, H2, ..., H... T The process for determining each first downlink channel matrix can be found in the relevant description in step 601 above, and will not be repeated here.

[0339] The process of the terminal acquiring channel information samples can be called initialization.

[0340] 2002. The terminal determines the first matrix, the second matrix, and the first coefficient matrix based on the channel information sample.

[0341] In a first possible implementation, the first coefficient matrix equals the first downlink channel matrix multiplied by the second matrix. The first downlink channel matrix equals the second coefficient matrix multiplied by the first matrix, and the number of columns in the first coefficient matrix is ​​less than the number of columns in the first downlink channel matrix. Each second coefficient matrix corresponds to one first downlink channel matrix, and the row vectors in the second coefficient matrix correspond one-to-one with the row vectors in the corresponding first downlink channel matrix. A row vector in the second coefficient matrix is ​​a sparsed representation of the corresponding row vector in the first downlink channel matrix.

[0342] In the second possible implementation, the first coefficient matrix = the second matrix * the first downlink channel matrix. The first downlink channel matrix = the first matrix * the second coefficient matrix, where the number of rows in the first coefficient matrix is ​​less than the number of rows in the first downlink channel matrix. Each second coefficient matrix corresponds to one first downlink channel matrix, and the column vectors in the second coefficient matrix correspond one-to-one with the column vectors in the corresponding first downlink channel matrix. A column vector in the second coefficient matrix is ​​a sparsed representation of the corresponding column vector in the first downlink channel matrix.

[0343] When determining the first coefficient matrix, the first matrix and the second coefficient matrix can be determined first based on the first downlink channel matrix, the second matrix can be determined based on the first matrix, and finally the first coefficient matrix can be determined based on the second matrix and the first downlink channel matrix.

[0344] Optionally, the second coefficient matrix satisfies at least one of the following conditions:

[0345] 1) The number of non-zero elements in the second coefficient matrix is ​​less than the number of non-zero elements in the first downlink channel matrix;

[0346] 2) If the ratio of the sum of the energy of the P1 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​greater than the first threshold, and the ratio of the sum of the energy of the P1-1 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​less than the first threshold, then the ratio of the sum of the energy of the P2 largest amplitude elements in the second coefficient matrix to the energy of the second coefficient matrix is ​​greater than the first threshold, and the ratio of the sum of the energy of the P2-1 largest amplitude elements in the second coefficient matrix to the energy of the second coefficient matrix is ​​less than the first threshold, where P2 < P1, P1 and P2 are both integers greater than or equal to 1, and the first threshold is an integer greater than 0 and less than or equal to 1.

[0347] When the second coefficient matrix satisfies at least one of the conditions 1) and 2), it indicates that the second coefficient matrix is ​​a sparsed representation of the first downlink channel matrix.

[0348] The first threshold can be preset, predefined, specified by protocol, or determined through negotiation between the access network equipment and the terminal; this application does not impose any restrictions. For example, the first threshold can be 95%, 90%, 85%, 80%, etc.

[0349] 2003. The terminal reports the matrix information to the access network device. Correspondingly, the access network device receives the matrix information from the terminal. The matrix information includes the first coefficient matrix and related information about the first and second matrices; or, the matrix information includes related information about the first coefficient matrix and the tenth matrix, where the tenth matrix is ​​determined based on the first and second matrices.

[0350] For example, the tenth matrix can be the product of the first and second matrices.

[0351] The relevant information of a matrix can be the matrix itself, or the magnitude, phase, and position of each element in the matrix within the candidate matrix set.

[0352] 2004. The access network equipment obtains the first downlink channel matrix based on the matrix information.

[0353] Figure 20 The method shown allows the terminal to compress the first downlink channel matrix (frequency domain compression or spatial domain compression), thereby reducing the overhead of uplink feedback.

[0354] Optionally, the first matrix and / or the second matrix satisfy at least one of the following conditions:

[0355] 1) The magnitude of at least one element in the first matrix and / or the second matrix is ​​different from the magnitude of at least one other element;

[0356] 2) At least one row or at least one column in the first matrix and / or the second matrix is ​​a non-geometric sequence.

[0357] Optionally, the reporting period of the first coefficient matrix is ​​less than the reporting periods of the first matrix and the second matrix, respectively; or, the reporting period of the first coefficient matrix is ​​less than the reporting period of the tenth matrix.

[0358] It should be noted that the terminal can update each matrix periodically, or update each matrix when the channel quality is poor, or update each matrix under other circumstances. This application does not impose any restrictions.

[0359] Scenario 2: The terminal performs two compressions (e.g., frequency domain compression and spatial domain compression) on the downlink channel matrix of actual measurements for one or more time units.

[0360] In scenario 2, such as Figure 21 As shown, the method also includes:

[0361] 2201. The terminal acquires channel information samples, which include the first downlink channel matrix.

[0362] The relevant description of step 2201 can be found in step 2001 above, and will not be repeated here.

[0363] 2202. The terminal determines the first matrix, the second matrix, and the first coefficient matrix based on the channel information sample.

[0364] In a first possible implementation, the first coefficient matrix equals the second downlink channel matrix multiplied by the second matrix, where the second downlink channel matrix is ​​determined by the first downlink channel matrix. The second downlink channel matrix equals the third coefficient matrix multiplied by the first matrix, where the number of columns in the first coefficient matrix is ​​less than the number of columns in the second downlink channel matrix. Each third coefficient matrix corresponds to one second downlink channel matrix, and the row vectors in the third coefficient matrix correspond one-to-one with the row vectors in the corresponding second downlink channel matrix. A row vector in the third coefficient matrix is ​​a sparsed representation of the corresponding row vector in the second downlink channel matrix.

[0365] In a second possible implementation, the first coefficient matrix equals the second matrix multiplied by the second downlink channel matrix, where the second downlink channel matrix is ​​determined by the first downlink channel matrix. The second downlink channel matrix equals the first matrix multiplied by the third coefficient matrix, where the number of rows in the first coefficient matrix is ​​less than the number of rows in the second downlink channel matrix. Each third coefficient matrix corresponds to one second downlink channel matrix, and the column vectors in the third coefficient matrix correspond one-to-one with the column vectors in the corresponding second downlink channel matrix. Each column vector in the third coefficient matrix is ​​a sparsed representation of the corresponding column vector in the second downlink channel matrix.

[0366] Optionally, the third coefficient matrix satisfies at least one of the following conditions:

[0367] 1) The number of non-zero elements in the third coefficient matrix is ​​less than the number of non-zero elements in the second downlink channel matrix;

[0368] 2) If the ratio of the sum of the energies of the P3 largest amplitude elements in the second downlink channel matrix to the energy of the second downlink channel matrix is ​​greater than the second threshold, and the ratio of the sum of the energies of the P3-1 largest amplitude elements in the second downlink channel matrix to the energy of the second downlink channel matrix is ​​less than the second threshold, then the ratio of the sum of the energies of the P4 largest amplitude elements in the third coefficient matrix to the energy of the third coefficient matrix is ​​greater than the second threshold, and the ratio of the sum of the energies of the P4-1 largest amplitude elements in the third coefficient matrix to the energy of the third coefficient matrix is ​​less than the second threshold, where P4 < P3, P3 and P4 are both integers greater than or equal to 1, and the second threshold is an integer greater than 0 and less than or equal to 1.

[0369] When the third coefficient matrix satisfies at least one of the conditions 1) and 2), it indicates that the third coefficient matrix is ​​a sparsed representation of the second downlink channel matrix.

[0370] The second threshold can be preset, predefined, specified by protocol, or determined through negotiation between the access network equipment and the terminal; this application does not impose any restrictions. For example, the second threshold can be 95%, 90%, 85%, 80%, etc.

[0371] When determining the first coefficient matrix, the first and third coefficient matrices can be determined first based on the second downlink channel matrix, the second matrix can be determined based on the first matrix, and finally the first coefficient matrix can be determined based on the second matrix and the second downlink channel matrix.

[0372] 2203. The terminal reports the matrix information to the access network device. Correspondingly, the access network device receives the matrix information from the terminal.

[0373] The matrix information includes the first coefficient matrix and related information of the first and second matrices; or, the matrix information includes related information of the first coefficient matrix and the tenth matrix, wherein the tenth matrix is ​​determined based on the first and second matrices. The access network device obtains the first downlink channel matrix based on the matrix information.

[0374] 2204. The access network equipment obtains the first downlink channel matrix based on the matrix information.

[0375] Figure 21 The method shown allows the terminal to compress the second downlink channel matrix. Since the second downlink channel matrix is ​​determined based on the first downlink channel matrix, the terminal can compress the first downlink channel matrix, thereby reducing the overhead of uplink feedback.

[0376] Optionally, the first matrix and / or the second matrix satisfy at least one of the following conditions:

[0377] 1) The magnitude of at least one element in the first matrix and / or the second matrix is ​​different from the magnitude of at least one other element;

[0378] 2) At least one row or at least one column in the first matrix and / or the second matrix is ​​a non-geometric sequence.

[0379] Optionally, the reporting period of the first coefficient matrix is ​​less than the reporting periods of the first matrix and the second matrix, respectively; or, the reporting period of the first coefficient matrix is ​​less than the reporting period of the tenth matrix.

[0380] Optional, see Figure 22 The matrix information also includes information about the third and fourth matrices, or the matrix information also includes information about the eleventh matrix, which is determined based on the third and fourth matrices. The method further includes:

[0381] 2202A. The terminal determines the third and fourth matrices based on the channel information samples. Specifically, the second downlink channel matrix = the fourth matrix * the first downlink channel matrix, the first downlink channel matrix = the third matrix * the fourth coefficient matrix, and the number of rows in the second downlink channel matrix is ​​less than the number of rows in the first downlink channel matrix.

[0382] In this case, the terminal can determine the first matrix based on the channel information sample and the fourth matrix.

[0383] For example, the eleventh matrix can be the product of the third and fourth matrices.

[0384] Optionally, the fourth coefficient matrix satisfies at least one of the following conditions:

[0385] 1) The number of non-zero elements in the fourth coefficient matrix is ​​less than the number of non-zero elements in the first downlink channel matrix;

[0386] 2) If the ratio of the sum of the energy of the P5 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​greater than the third threshold, and the ratio of the sum of the energy of the P5-1 largest amplitude elements in the first downlink channel matrix to the energy of the first downlink channel matrix is ​​less than the third threshold, then the ratio of the sum of the energy of the P6 largest amplitude elements in the fourth coefficient matrix to the energy of the fourth coefficient matrix is ​​greater than the third threshold, and the ratio of the sum of the energy of the P6-1 largest amplitude elements in the fourth coefficient matrix to the energy of the fourth coefficient matrix is ​​less than the third threshold, where P6 < P5, P5 and P6 are both integers greater than or equal to 1, and the third threshold is an integer greater than 0 and less than or equal to 1.

[0387] When the fourth coefficient matrix satisfies at least one of the conditions 1) and 2), it indicates that the fourth coefficient matrix is ​​a sparse representation of the first downlink channel matrix.

[0388] The third threshold can be preset, predefined, stipulated by protocol, or determined through negotiation between the access network equipment and the terminal; this application does not impose any restrictions. For example, the third threshold can be 95%, 90%, 85%, 80%, etc.

[0389] Optionally, the third and / or fourth matrices satisfy at least one of the following conditions:

[0390] 1) The magnitude of at least one element in the third and / or fourth matrix differs from the magnitude of at least one other element;

[0391] 2) At least one row or at least one column in the third and / or fourth matrix is ​​a non-geometric sequence.

[0392] Optionally, the reporting period of the first coefficient matrix is ​​shorter than the reporting periods of the first matrix, the second matrix, the third matrix, and the fourth matrix, respectively; or, the reporting period of the first coefficient matrix is ​​shorter than the reporting periods of the tenth matrix and the eleventh matrix.

[0393] It should be noted that the terminal can update each matrix periodically, or update each matrix when the channel quality is poor, or update each matrix under other circumstances. This application does not impose any restrictions.

[0394] Similar to Embodiment 1, for ease of description, the following example uses a channel information sample comprising T N*M first downlink channel matrices (T, N, and M are all integers greater than or equal to 1) to illustrate the method for determining each matrix in the above method. Specifically, it is illustrated in Case 1 (determining each matrix for all frequency domain units in the time unit) and Case 2 (determining each matrix individually for each frequency domain unit in the time unit).

[0395] Case 1: Determine each matrix for all frequency domain units in the time unit.

[0396] In case 1, optionally, the third matrix satisfies the following condition: the F-norm of the difference between the product of the third and fifth matrices and the sixth matrix is ​​minimized.

[0397] The sixth matrix is ​​an N*(M*T) matrix. The column vectors in the sixth matrix are composed of all the column vectors in the T first downlink channel matrices. The column vectors in the fifth matrix are composed of all the column vectors in the T fourth coefficient matrices. A column vector in the fifth matrix is ​​a sparse representation of the corresponding column vector in the sixth matrix. The number of non-zero elements in each column vector in the fifth matrix is ​​equal to S1, where S1 is an integer greater than or equal to 1 and less than N.

[0398] In case 1, optionally, the fourth matrix satisfies the following condition: the matrix obtained by multiplying the fourth matrix and the third matrix (which can be denoted as the seventh matrix) has the minimum column correlation.

[0399] In case 1, optionally, the first matrix satisfies the following condition: the F-norm of the matrix difference between the product of the first matrix and the eighth matrix and the conjugate transpose of the ninth matrix is ​​minimized.

[0400] The ninth matrix is ​​determined based on the channel information sample and the fourth matrix. The m-th column vector of the ninth matrix is ​​composed of T column vectors arranged from top to bottom. The s-th column vector among the T column vectors is the product of the m-th column of the s-th first downlink channel matrix and the fourth matrix. The s-th first downlink channel matrix is ​​the s-th first downlink channel matrix among the T first downlink channel matrices. m is an integer greater than or equal to 1 and less than or equal to M, and s is an integer greater than or equal to 1 and less than or equal to T.

[0401] In this matrix, the column vectors in the eighth matrix correspond one-to-one with the column vectors in the conjugate transpose of the ninth matrix. A column vector in the eighth matrix is ​​a sparse representation of the corresponding column vector in the conjugate transpose of the ninth matrix. The number of non-zero elements in each column vector in the eighth matrix is ​​equal to S2, where S2 is an integer greater than or equal to 1 and less than M.

[0402] Among them, the m-th column vector in the eighth matrix can be composed of T column vectors arranged from top to bottom. The s-th column vector among these T column vectors is the m-th column in the s-th third coefficient matrix, and the s-th third coefficient matrix is ​​the third coefficient matrix corresponding to the x+s-1 time unit.

[0403] In case 1, optionally, the second matrix satisfies the following condition: the column correlation of the matrix obtained by multiplying the second matrix and the first matrix is ​​minimized.

[0404] The specific implementation of the method for determining the third, fourth, and first matrices under Case 1 of Embodiment 2 can be found in Case 1 of Embodiment 1 above. The difference between the method for determining the second matrix under Case 1 of Embodiment 2 and Case 1 of Embodiment 1 is that, in Embodiment 2, the fourth matrix may not be a row extraction matrix.

[0405] Case 2: Determine each matrix separately for each frequency domain unit in the time unit.

[0406] In case 2, optionally, the third matrix includes M third sub-matrices, and the m-th third sub-matrix among the M third sub-matrices satisfies the following condition: the F-norm of the matrix difference between the product of the m-th third sub-matrix and the m-th fifth sub-matrix and the m-th sixth sub-matrix is ​​minimized, where m is an integer greater than or equal to 1 and less than or equal to M.

[0407] The m-th sixth submatrix is ​​an N*T matrix. The column vectors in the m-th sixth submatrix are composed of the m-th column vectors in the T first downlink channel matrices. The column vectors in the m-th fifth submatrix are composed of the m-th column vectors in the T fourth coefficient matrices. A column vector in the m-th fifth submatrix is ​​a sparse representation of the corresponding column vector in the m-th sixth submatrix. The number of non-zero elements in each column vector of the m-th fifth submatrix is ​​equal to S1, where S1 is an integer greater than or equal to 1 and less than N.

[0408] In case 2, optionally, the fourth matrix includes M fourth sub-matrices, and the m-th fourth sub-matrix among the M fourth sub-matrices satisfies the following condition: the m-th seventh sub-matrix obtained by multiplying the m-th fourth sub-matrix with the m-th third sub-matrix has the minimum column correlation.

[0409] In case 2, optionally, the first matrix satisfies the following condition: the F-norm of the matrix difference between the product of the first matrix and the eighth matrix and the conjugate transpose of the ninth matrix is ​​minimized.

[0410] The ninth matrix is ​​determined based on the channel information sample and the fourth matrix. The m-th column vector of the ninth matrix is ​​composed of T column vectors arranged from top to bottom. The s-th column vector among the T column vectors is the product of the m-th column of the s-th first downlink channel matrix and the m-th fourth sub-matrix. The s-th first downlink channel matrix is ​​the s-th first downlink channel matrix among the T first downlink channel matrices, where s is an integer greater than or equal to 1 and less than or equal to T.

[0411] In this matrix, the column vectors in the eighth matrix correspond one-to-one with the column vectors in the conjugate transpose of the ninth matrix. A column vector in the eighth matrix is ​​a sparse representation of the corresponding column vector in the conjugate transpose of the ninth matrix. The number of non-zero elements in each column vector in the eighth matrix is ​​equal to S2, where S2 is an integer greater than or equal to 1 and less than M.

[0412] In this matrix, the m-th column vector is composed of T column vectors arranged from top to bottom. The s-th column vector among these T column vectors is the m-th column of the s-th third coefficient matrix, and the s-th third coefficient matrix is ​​the third coefficient matrix corresponding to the x+s-1 time unit.

[0413] In case 2, optionally, the second matrix satisfies the following condition: the column correlation of the matrix obtained by multiplying the second matrix and the first matrix is ​​minimized.

[0414] The specific implementation of the method for determining the third, fourth, and first matrices in Case 2 of Embodiment 2 can be found in Case 2 of Embodiment 1 above. The method for determining the second matrix in Case 2 of Embodiment 2 is the same as in Case 1 of Embodiment 2, and can be found above, so it will not be repeated here.

[0415] Optionally, in one case, T represents the total number of time units included in the channel information sample; N represents the product of the number of ports configured by the access network device to transmit measurement pilots and the number of ports configured by the terminal to receive measurement pilot signals; and M represents the number of frequency domain units configured by the access network device to transmit measurement pilots. In this case, the terminal performs spatial domain compression on the first downlink channel matrix first, and then frequency domain compression. In another case, N represents the number of frequency domain units configured by the access network device to transmit measurement pilots, and M represents the product of the number of ports configured by the access network device to transmit measurement pilots and the number of ports configured by the terminal to receive measurement pilot signals. In this case, the terminal performs frequency domain compression on the first downlink channel matrix first, and then spatial domain compression.

[0416] For scenarios 1 and 2, when the access network device restores T first downlink channel matrices, it can restore each first downlink channel matrix. The restoration method is similar to that in Implementation Example 1, and can be found above for details, which will not be repeated here.

[0417] For scenario 1, if the terminal only performs spatial compression on the first downlink channel matrix, then in case 1, the first and second matrices can be calculated using formulas 1.1 and 1.2 respectively. In this case, Ψ D Let Φ be the first matrix. D This is the second matrix. In case 2, the first and second matrices can be calculated using formulas 21 and 22 respectively. In this case, Ψ m Let Φ be the m-th first submatrix in the first matrix. m It is the m-th second submatrix in the second matrix.

[0418] For scenario 1, if the terminal only performs frequency domain compression on the first downlink channel matrix, in this case, it is not necessary to distinguish between different frequency domain units. The terminal can use formula 1.3 to calculate the first matrix. The difference is that when determining Y, Φ is not required. D ,Right now The terminal can use Formula 14 to calculate the second matrix.

[0419] For scenario 2, when the terminal performs frequency domain compression followed by spatial domain compression on the first downlink channel matrix, it is not necessary to distinguish between different frequency domain units. The terminal can use formula 1.3 to calculate the third matrix. The difference is that when determining Y, Φ is not required. D ,Right now At this time, Ψ f This is the third matrix; the terminal can calculate the fourth matrix using formula 1.4, in which case Φ f This is the fourth matrix; the terminal can calculate the first matrix using formula 1.1, the difference being that... Among them, h n，s H represents s In the nth row, at this time, Ψ D This is the first matrix; the terminal can calculate the second matrix using formula 1.2, where Φ D This is the second matrix.

[0420] When compressed sensing technology is used to implement the method described in this application, if the terminal first performs spatial domain compression on the first downlink channel matrix and then performs frequency domain compression, then the first to fourth matrices can also be referred to as: frequency domain sensing matrix, frequency domain compression matrix, spatial domain sensing matrix, and spatial domain compression matrix, respectively. Furthermore, the access network device can also send indication information to the terminal, thereby instructing the terminal whether to use the conventional method for uplink feedback or the method provided in this application for uplink feedback. The terminal can then perform the corresponding uplink feedback based on this indication information.

[0421] The method provided in Embodiment 2 obtains a more accurate first downlink channel matrix because the terminal does not use historical time unit first downlink channel matrix information when determining each matrix. Furthermore, the compression process of the first downlink channel matrix provided in this embodiment allows for a more sparse representation of the first downlink channel matrix compared to existing technologies. Therefore, the terminal feeds back less information to the access network equipment, thus reducing uplink feedback overhead compared to existing technologies.

[0422] Based on the above embodiments one and two, in another embodiment, "minimum F-norm" can also be replaced with "F-norm is less than or equal to the threshold" or "F-norm is within a numerical range". The threshold and numerical range can be preset, predefined, protocol-specified, or determined by negotiation between the access network equipment and the terminal. This application does not impose any restrictions.

[0423] The signal transmission method provided in the above embodiments of this application can also be applied to large-scale MIM0 uplink channel estimation to reduce the overhead of the sounding reference signal (SRS), and can also be applied to uplink and downlink data channel estimation to reduce the overhead of the demodulation reference signal (DMRS). It can also be applied to scenarios where the channel phase deviation caused by the crystal oscillator is estimated to reduce the overhead of the phase tracking reference signal (PT-RS). The implementation principle is similar and will not be described again.

[0424] The above mainly describes the solutions of the embodiments of this application from the perspective of interaction between various network elements. It is understood that each network element, such as access network equipment and terminals, includes corresponding hardware structures and / or software modules to perform the above functions in order to achieve them. Those skilled in the art should readily recognize that, based on the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed by hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0425] This application embodiment can divide access network devices and terminals into functional units according to the above method examples. For example, each function can be divided into a separate functional unit, or two or more functions can be integrated into one processing unit. The integrated unit can be implemented in hardware or as a software functional unit. It should be noted that the unit division in this application embodiment is illustrative and only represents one logical functional division; other division methods may be used in actual implementation.

[0426] When using integrated units Figure 23 A communication device 240 according to the above embodiments is shown. The communication device 240 may include a processing unit 2401 and a communication unit 2402. Optionally, the communication device 240 may also include a storage unit 2403. Figure 23 The structural diagram shown can be used to illustrate the structure of the signal transmitting device and the channel matrix acquisition device involved in the above embodiments, and can specifically be used to illustrate the structure of the access network equipment and the terminal.

[0427] when Figure 23 The structural diagram shown is used to illustrate the structure of the access network device involved in the above embodiments. The processing unit 2401 is used to control and manage the actions of the access network device. For example, the processing unit 2401 is used to support the access network device in performing... Figure 6 601 to 603 in the middle, Figure 8 The numbers 601, 602, 602A, 603, 605 to 607 are listed below. Figure 9 901 to 903 in the middle, Figure 10 The numbers 901, 902, 902A, 902B, 903, 905 to 907 are included. Figure 11 1101 and 1102 in the middle, Figure 12 1101, 1101A, 1102, 1104 to 1106, Figure 20 2003 and 2004 in the middle, Figure 21 2203 and 2204 in the middle, Figure 22 The actions performed by the access network device in processes 2203, 2204, and / or other processes described in the embodiments of this application. The processing unit 2401 can communicate with other network entities via the communication unit 2402, for example, with... Figure 6 The diagram shows communication between terminals. Storage unit 2403 is used to store program code and data from the access network device.

[0428] when Figure 23 When the schematic diagram shown is used to illustrate the structure of the access network device involved in the above embodiments, the communication device 240 can be a device or a chip within a device.

[0429] when Figure 23 The schematic diagram shown illustrates the structure of the terminal involved in the above embodiments. The processing unit 2401 is used to control and manage the actions of the terminal. For example, the processing unit 2401 is used to support the terminal in performing... Figure 6 603 in Figure 8 602A, 603 to 605, Figure 9 903 in the middle, Figure 10 902A, 902B, 903 to 905, Figure 11 1102 in the middle, Figure 12 1101A, 1102 to 1104, Figure 20 From 2001 to 2003, Figure 21 2201 to 2203 in the middle, Figure 22 The actions performed by the terminal in processes 2201, 2202, 2202A, 2203, and / or other processes described in the embodiments of this application. The processing unit 2401 can communicate with other network entities via the communication unit 2402, for example, with... Figure 6 The diagram shows communication between access network devices. Storage unit 2403 is used to store the terminal's program code and data.

[0430] when Figure 23 When the schematic diagram shown is used to illustrate the structure of the terminal involved in the above embodiments, the communication device 240 can be a device or a chip within a device.

[0431] Figure 23 If the integrated units in the process are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, in essence, or the parts that contribute to the prior art, or all or part of the technical solutions, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods of the various embodiments of this application. Storage media for storing computer software products include various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0432] Figure 23 The units in the process can also be called modules; for example, a processing unit can be called a processing module.

[0433] Figure 24 This is a schematic diagram of the hardware structure of the communication device 250 provided in an embodiment of this application. The communication device 250 includes one or more processors 2501 and a communication interface 2503.

[0434] Optionally, the communication device 250 further includes a memory 2504, which is coupled to the processor 2501. The memory 2504 may include ROM and RAM, and provides operation instructions and data to the processor 2501. A portion of the memory 2504 may also include non-volatile random access memory (NVRAM).

[0435] In this embodiment of the application, the communication device 250 performs corresponding operations by calling the operation instructions stored in the memory 2504 (the operation instructions may be stored in the operating system).

[0436] The processor 2501 can also be called a central processing unit (CPU).

[0437] The processor 2501, communication interface 2503, and memory 2504 are coupled together via a communication bus 2502. This communication bus 2502 may include, in addition to a data bus, a power bus, a control bus, and a status signal bus. However, for clarity, in... Figure 24 The general labeled all buses as communication bus 2502.

[0438] The methods disclosed in the embodiments of this application can be applied to or implemented by the processor 2501. The processor 2501 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the integrated logic circuit of the hardware in the processor 2501 or by instructions in the form of software. The processor 2501 may be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules may be located in RAM, flash memory, ROM, programmable read-only memory or electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 2504. Processor 2501 reads the information in memory 2504 and completes the steps of the above method in conjunction with its hardware.

[0439] For example, Figure 24 The structural diagram shown can be used to illustrate the structure of the signal transmitting device and the channel matrix acquisition device involved in the above embodiments, and can specifically be used to illustrate the structure of the access network equipment and the terminal.

[0440] when Figure 24 The schematic diagram shown illustrates the structure of the access network device involved in the above embodiments. The processor 2501 is used to control and manage the actions of the access network device. For example, the processor 2501 is used to support the access network device in performing... Figure 6 601 to 603 in the middle, Figure 8 The numbers 601, 602, 602A, 603, 605 to 607 are listed below. Figure 9 901 to 903 in the middle, Figure 10 The numbers 901, 902, 902A, 902B, 903, 905 to 907 are included. Figure 11 1101 and 1102 in the middle, Figure 12 1101, 1101A, 1102, 1104 to 1106, Figure 20 2003 and 2004 in the middle, Figure 212203 and 2204 in the middle, Figure 22 The actions performed by the access network device in processes 2203, 2204, and / or other processes described in the embodiments of this application. The processor 2501 can communicate with other network entities via the communication interface 2503, for example, with... Figure 6 The diagram shows communication between terminals. Memory 2504 is used to store program code and data from the access network device.

[0441] when Figure 24 The schematic diagram shown illustrates the structure of the terminal involved in the above embodiments. The processor 2501 is used to control and manage the actions of the terminal; for example, the processor 2501 is used to support the terminal in executing... Figure 6 603 in Figure 8 602A, 603 to 605, Figure 9 903 in the middle, Figure 10 902A, 902B, 903 to 905, Figure 11 1102 in the middle, Figure 12 1101A, 1102 to 1104, Figure 20 From 2001 to 2003, Figure 21 2201 to 2203 in the middle, Figure 22 The actions performed by the terminal in processes 2201, 2202, 2202A, 2203, and / or other processes described in the embodiments of this application. The processor 2501 can communicate with other network entities via the communication interface 2503, for example, with... Figure 6 The diagram shows communication between access network devices. Memory 2504 is used to store the terminal's program code and data.

[0442] The communication unit or communication interface described above can be an interface circuit or communication interface of the device, used to receive signals from other devices. For example, when the device is implemented as a chip, the communication unit or communication interface is an interface circuit or communication interface used by the chip to receive or send signals from other chips or devices.

[0443] In the above embodiments, the instructions stored in the memory for execution by the processor can be implemented in the form of a computer program product. The computer program product can be pre-written into the memory, or it can be downloaded and installed into the memory as software.

[0444] A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or data center that integrates one or more available media. The available medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk, SSD), etc.

[0445] Optionally, embodiments of this application also provide a computer-readable storage medium storing computer instructions; when the computer-readable storage medium is run on a computer, the computer executes the communication method provided in embodiments of this application.

[0446] This application also provides a computer program product containing computer instructions, which, when run on a computer, enables the computer to execute the methods provided in this application.

[0447] This application also provides a chip, which includes a processor and an interface. The processor is coupled to a memory through the interface. When the processor executes a computer program or instructions in the memory, the method provided in this application is executed.

[0448] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented using software programs, implementation can be, in whole or in part, in the form of a computer program product. This computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, DSL) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or include one or more data storage devices such as servers or data centers that can be integrated with the medium. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., SSD), etc.

[0449] Although this application has been described herein in conjunction with various embodiments, those skilled in the art, by reviewing the accompanying drawings, disclosure, and appended claims, will understand and implement other variations of the disclosed embodiments in carrying out the claimed application. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude multiple components. A single processor or other unit can implement several functions listed in the claims. While different dependent claims may recite certain measures, this does not mean that these measures cannot be combined to produce good results.

[0450] Although this application has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made thereto without departing from the scope of this application. Accordingly, this specification and drawings are merely illustrative descriptions of the application as defined by the appended claims, and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Thus, if such modifications and modifications fall within the scope of the claims and their equivalents, this application is also intended to include such modifications and modifications.

Claims

1. A signal transmission method, characterized in that, Applied to access network equipment, including: Obtain channel information samples, which include channel information from the xth time unit to the x+T-1th time unit, where x and T are both integers greater than or equal to 1; Based on the channel information sample, M' frequency domain units are determined from the M frequency domain units in the x+T time unit. The M' frequency domain units are used to transmit measurement pilots. The measurement pilots are used to measure channel state information (CSI). M' and M are both integers greater than or equal to 1, and M' < M. The measurement pilot signals are sent to the terminal in the M' frequency domain units; The measurement pilot transmitted in each of the M' frequency domain units is the measurement pilot after precoding by the fourth matrix; wherein the fourth matrix satisfies the following condition: the column correlation of the seventh matrix obtained by multiplying the fourth matrix and the third matrix is ​​minimal.

2. The method according to claim 1, characterized in that, The method further includes: Send a first indication message to the terminal, the first indication message being used to indicate the position information of the M' frequency domain units.

3. The method according to claim 1, characterized in that, The step of sending the measurement pilot to the terminal in the M' frequency domain units includes: Based on the channel information sample, N′ port measurement pilots are sent to the terminal in each of the M′ frequency domain units, and second indication information is sent to the terminal. The second indication information is used to indicate the value of N′, where N′ is an integer greater than or equal to 1.

4. The method according to claim 1, characterized in that, The method further includes: The terminal receives feedback information, which is used to indicate the information of the measurement pilot signal received by the terminal at the (x+T)th time unit; The downlink channel matrix for the (x+T)th time unit is determined based on the feedback information.

5. The method according to claim 1, characterized in that, The method further includes: The downlink channel matrix of the (x+T+1)th time unit is calculated using the channel information of p time units as channel information samples, where p is an integer greater than or equal to 1 and less than or equal to x+T.

6. The method according to any one of claims 1-5, characterized in that, The downlink channel matrix of a time unit represents the channel information of that time unit. The downlink channel matrix of each time unit is an N*M matrix, where N is an integer greater than or equal to 1. The third matrix satisfies the following condition: the F-norm of the difference between the product of the third matrix and the fifth matrix and the sixth matrix is ​​minimized. The sixth matrix is ​​an N*(M*T) matrix. The column vectors in the sixth matrix are composed of all the column vectors in the downlink channel matrix from the x-th time unit to the x+T-1-th time unit. The column vectors in the fifth matrix correspond one-to-one with the column vectors in the sixth matrix. A column vector in the fifth matrix is ​​a sparse representation of the corresponding column vector in the sixth matrix. The number of non-zero elements in each column vector of the fifth matrix is ​​equal to S1, where S1 is an integer greater than or equal to 1 and less than N.

7. The method according to any one of claims 1-5, characterized in that, The M' frequency domain units are determined by a second matrix, which satisfies the following conditions: the column correlation of the matrix obtained by multiplying the second matrix with the first matrix is ​​minimized; the second matrix is ​​a row decimation matrix; each row of the second matrix has only one non-zero element; and the non-zero elements in different rows are in different positions.

8. The method according to claim 7, characterized in that, The first matrix satisfies the following condition: the F-norm of the matrix difference between the product of the first matrix and the eighth matrix and the conjugate transpose of the ninth matrix is ​​minimized. The ninth matrix is ​​determined based on the channel information sample and the fourth matrix; the m-th column vector of the ninth matrix is ​​composed of T column vectors arranged from top to bottom, the s-th column vector of the T column vectors is the product of the m-th column of the s-th downlink channel matrix and the fourth matrix, and the s-th downlink channel matrix is ​​the downlink channel matrix of the x+s-1 time units from the x-th time unit to the x+T-1-th time unit, where m is an integer greater than or equal to 1 and less than or equal to M, and s is an integer greater than or equal to 1 and less than or equal to T; In this matrix, the column vectors in the eighth matrix correspond one-to-one with the column vectors in the conjugate transpose of the ninth matrix. Each column vector in the eighth matrix is ​​a sparse representation of the corresponding column vector in the conjugate transpose of the ninth matrix. The number of non-zero elements in each column vector of the eighth matrix is ​​equal to S2, where S2 is an integer greater than or equal to 1 and less than M.

9. The method according to any one of claims 1-5, characterized in that, The measurement pilot transmitted on each of the M' frequency domain units is a measurement pilot pre-coded by the fourth matrix, including: the measurement pilot transmitted on the m'th frequency domain unit of the M' frequency domain units is a measurement pilot pre-coded by the m'th fourth sub-matrix of the M' fourth sub-matrixes in the fourth matrix, the fourth matrix including M fourth sub-matrixes corresponding one-to-one with the M frequency domain units, the M' fourth sub-matrixes being the fourth sub-matrixes corresponding to the M' frequency domain units; wherein, the m'th fourth sub-matrix of the M fourth sub-matrix satisfies the following condition: the column correlation of the m'th seventh sub-matrix obtained by multiplying the m'th fourth sub-matrix with the m'th third sub-matrix in the third matrix is ​​minimal, the third matrix including M third sub-matrixes; m is an integer greater than or equal to 1 and less than or equal to M, m' is an integer greater than or equal to 1 and less than or equal to M'.

10. The method according to claim 9, characterized in that, The downlink channel matrix of a time unit represents the channel information of that time unit. The downlink channel matrix of each time unit is an N*M matrix, where N is an integer greater than or equal to 1. The m-th third sub-matrix among the M third sub-matrixes satisfies the following condition: the F-norm of the matrix difference between the product of the m-th third sub-matrix and the m-th fifth sub-matrix and the m-th sixth sub-matrix is ​​minimized. Wherein, the m-th sixth sub-matrix is ​​an N*T matrix, the column vectors in the m-th sixth sub-matrix are composed of the m-th column vectors in the downlink channel matrix from the x-th time unit to the x+T-1-th time unit, the column vectors in the m-th fifth sub-matrix correspond one-to-one with the column vectors in the m-th sixth sub-matrix, one column vector in the m-th fifth sub-matrix is ​​a sparse representation of the corresponding column vector in the m-th sixth sub-matrix, and the number of non-zero elements in each column vector in the m-th fifth sub-matrix is ​​equal to S1, where S1 is an integer greater than or equal to 1 and less than N.

11. The method according to claim 9, characterized in that, The M' frequency domain units are determined by a second matrix, which satisfies the following conditions: the column correlation of the matrix obtained by multiplying the second matrix with the first matrix is ​​minimized; the second matrix is ​​a row decimation matrix; each row of the second matrix has only one non-zero element; and the non-zero elements in different rows are in different positions.

12. The method according to claim 11, characterized in that, The first matrix satisfies the following condition: the F-norm of the matrix difference between the product of the first matrix and the eighth matrix and the conjugate transpose of the ninth matrix is ​​minimized. The ninth matrix is ​​determined based on the channel information sample and the fourth matrix. The m-th column vector of the ninth matrix is ​​composed of T column vectors arranged from top to bottom. The s-th column vector among the T column vectors is the product of the m-th column of the s-th downlink channel matrix and the m-th fourth sub-matrix. The s-th downlink channel matrix is ​​the downlink channel matrix of the x+s-1 time units from the x-th time unit to the x+T-1 time unit, where s is an integer greater than or equal to 1 and less than or equal to T. In this matrix, the column vectors in the eighth matrix correspond one-to-one with the column vectors in the conjugate transpose of the ninth matrix. Each column vector in the eighth matrix is ​​a sparse representation of the corresponding column vector in the conjugate transpose of the ninth matrix. The number of non-zero elements in each column vector of the eighth matrix is ​​equal to S2, where S2 is an integer greater than or equal to 1 and less than M.

13. A signal transmitting device, characterized in that, include: Functional unit for performing the method as described in any one of claims 1-12; The actions performed by the functional unit are implemented through hardware or through hardware executing corresponding software.

14. A signal transmitting device, characterized in that, include: The processor is coupled to the memory; The memory is used to store computer execution instructions, and the processor executes the computer execution instructions stored in the memory to cause the signal transmitting device to implement the method as described in any one of claims 1-12.

15. A computer-readable storage medium, characterized in that, Includes instructions that, when executed on a computer, cause the computer to perform the method as described in any one of claims 1-12.

16. A computer program product, characterized in that, Includes instructions that, when executed on a computer, cause the computer to perform the method as described in any one of claims 1-12.

17. A chip, characterized in that, include: A processor and an interface, the processor being coupled to a memory via the interface, wherein when the processor executes a computer program or instructions in the memory, the method as described in any one of claims 1-12 is performed.

Images