Windowing-based information geometric channel estimation method, device, equipment, medium and product

Abstract

The invention discloses a windowing-based information geometry channel estimation method, device and equipment, a medium and a product. The method comprises the following steps: initializing an auxiliary manifold average natural parameter, wherein the auxiliary manifold average natural parameter is used for approximately processing a nonlinear term of channel estimation; in each iteration, based on a windowed receiving signal model obtained by introducing a window function into the receiving signal model, calculating an approximate solution of m-projection of the auxiliary manifold to the target manifold by using the window function, and updating an average natural parameter of the auxiliary manifold according to a damping coefficient and the approximate solution of m-projection; after iteration is stopped, target manifold natural parameters are updated according to the auxiliary manifold average natural parameters, and a channel estimation result is determined according to the target manifold natural parameters. A windowing receiving signal model is obtained by introducing a window function into a receiving signal model, information geometric channel estimation is performed based on the windowing receiving signal model, and m-projection is simplified and calculated by using the window function, so that the channel estimation complexity is reduced, and the channel estimation efficiency is improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of coding technology, and in particular to a method, apparatus, device, medium and product for estimating information geometry channel based on windowing. Background Art

[0002] Massive Multiple-Input Multiple-Output (MIMO) technology is one of the key technologies in wireless communications. By deploying a large number (tens to hundreds) of antennas at base stations, it provides huge spatial freedom and can provide services to multiple users simultaneously on the same time-frequency resources, thereby significantly improving the system spectrum and energy efficiency. Orthogonal Frequency Division Multiplexing (OFDM) is a multi-carrier modulation technology with excellent ability to provide high data rate transmission and robustness against frequency selectivity. Massive MIMO-OFDM plays an important role in 5G systems and is receiving increasing attention in future sixth-generation (6G) systems.

[0003] In massive MIMO-OFDM systems, obtaining accurate channel state information is a crucial issue because system performance is highly dependent on the quality of channel estimation. In practical systems, pilot-assisted channel estimation is a common channel estimation method, that is, the transmitter periodically sends a pilot signal, and the receiver performs channel estimation based on the received pilot signal. Channel estimation uses the received pilot signal to obtain the a posteriori information of the channel parameters. However, due to the large dimension of the channel parameters, it is challenging to calculate the a posteriori mean and variance. Traditional high-performance algorithms, such as the minimum mean square error channel estimation algorithm, usually require large-dimensional matrix inversion operations, have high computational complexity, and are less practical when the channel dimension is large. The information geometry channel estimation algorithm proposed in recent years has the advantage of low complexity compared to the minimum mean square error algorithm, and its performance can approach the minimum mean square error algorithm, but its complexity is still high. Summary of the invention

[0004] The present application provides a windowing-based information geometry (IG) channel estimation method, apparatus, device, medium and product to reduce the complexity and computational complexity of channel estimation and improve the efficiency of channel estimation.

[0005] In a first aspect, an embodiment of the present application provides a windowing-based information geometry channel estimation method, comprising:

[0006] Initializing an auxiliary manifold average natural parameter, wherein the auxiliary manifold average natural parameter is the mean of the natural parameters of all auxiliary manifolds, and the auxiliary manifold average natural parameter is used to approximate the nonlinear term corresponding to the channel estimation;

[0007] In each iteration, based on the windowed received signal model, the introduced window function is used to calculate the approximate solution of the m-projection of the auxiliary manifold in this iteration to the target manifold in this iteration, and the average natural parameter of the auxiliary manifold is updated according to the damping coefficient and the approximate solution of the m-projection; wherein the windowed received signal model is obtained by introducing the window function into the received signal model; the m-projection is a distribution on the target manifold, and the distribution satisfies the minimum Kulbeck-Leibler divergence to the auxiliary manifold;

[0008] After the iteration stops, the target manifold natural parameters are updated according to the auxiliary manifold average natural parameters, and the channel estimation result is determined according to the target manifold natural parameters.

[0009] In a second aspect, an embodiment of the present application further provides an information geometry channel estimation device based on windowing, comprising:

[0010] An initialization module, used for initializing an auxiliary manifold average natural parameter, wherein the auxiliary manifold average natural parameter is the mean of the natural parameters of all auxiliary manifolds, and the auxiliary manifold average natural parameter is used for approximating the nonlinear term corresponding to the channel estimation;

[0011] An iteration module is used to calculate, in each iteration, an approximate solution of the m-projection of the auxiliary manifold in this iteration to the target manifold in this iteration using the introduced window function based on the windowed received signal model, and update the auxiliary manifold average natural parameter according to the damping coefficient and the approximate solution of the m-projection; wherein the windowed received signal model is obtained by introducing the window function into the received signal model; the m-projection is a distribution on the target manifold, and the distribution satisfies the minimum Kulbeck-Leibler divergence to the auxiliary manifold;

[0012] The channel estimation module is used to update the target manifold natural parameters according to the auxiliary manifold average natural parameters after the iteration stops, and determine the channel estimation result according to the target manifold natural parameters.

[0013] In a third aspect, an embodiment of the present application provides an electronic device, including:

[0014] one or more processors;

[0015] A storage device for storing one or more programs;

[0016] When the one or more programs are executed by the one or more processors, the one or more processors implement the windowing-based information geometry channel estimation method as described in the first aspect.

[0017] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the windowed information geometry channel estimation method as described in the first aspect.

[0018] In a fifth aspect, an embodiment of the present application further provides a computer program product, comprising a computer program and / or instructions, which, when executed by a processor, implements the windowed information geometry channel estimation method as described in any of the above embodiments.

[0019] The embodiment of the present application provides a window-based information geometry channel estimation method, device, equipment, medium and product, the method comprising: initializing the auxiliary manifold average natural parameter, wherein the auxiliary manifold average natural parameter is the mean of the natural parameters of all auxiliary manifolds, and the auxiliary manifold average natural parameter is used to approximate N-1 nonlinear terms; in each iteration, based on the windowed received signal model, using the introduced window function to calculate the approximate solution of the m-projection of the auxiliary manifold in this iteration to the target manifold in this iteration, and updating the auxiliary manifold average natural parameter according to the damping coefficient and the approximate solution of the m-projection; wherein the windowed received signal model is obtained by introducing the window function into the received signal model; the m-projection is a distribution on the target manifold, and is a distribution on the target manifold, which satisfies the minimum Kulbeck-Leibler divergence to the auxiliary manifold; after the iteration stops, updating the target manifold natural parameter according to the auxiliary manifold average natural parameter, and determining the channel estimation result according to the target manifold natural parameter. By introducing the window function into the received signal model, a windowed received signal model is obtained, and information geometry channel estimation is performed based on the windowed received signal model. The window function is used to simplify the calculation of m-projection, thereby reducing the complexity of channel estimation and improving the efficiency of channel estimation. By means of windowing, the proposed windowed information geometry channel estimation algorithm can greatly reduce the complexity of channel estimation and has high practical value. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale.

[0021] Figure 1 A flowchart of a windowed information geometry channel estimation method provided in an embodiment of the present application;

[0022] Figure 2 A schematic diagram of the structure of a windowed information geometry channel estimation device provided in an embodiment of the present application;

[0023] Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0024] The present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present application, rather than to limit the present application. It should also be noted that, for ease of description, only the parts related to the present application, rather than all structures, are shown in the accompanying drawings.

[0025] It should be mentioned before discussing the exemplary embodiments in more detail that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow charts describe the steps as sequential processes, many of the steps therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of the steps can be rearranged. The process can be terminated when its operation is completed, but can also have additional steps not included in the accompanying drawings. The process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0026] It should be noted that the concepts of "first", "second", etc. mentioned in the embodiments of the present application are only used to distinguish different devices, modules, units or other objects, and are not used to limit the order or interdependence of the functions performed by these devices, modules, units or other objects.

[0027] In addition, the embodiments in the present application and the features in the embodiments may be combined with each other if there is no conflict.

[0028] The acquisition, storage, use, and processing of data in the technical solution of this application comply with the relevant provisions of national laws and regulations.

[0029] It should be noted that in the embodiments of the present application, certain software, components, models and other existing solutions in the industry may be mentioned, and they should be regarded as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the relevant content of the solution.

[0030] Figure 1A flowchart of a windowed information geometry channel estimation method is provided for an embodiment of the present application. This embodiment can be applied to the case of channel estimation. Specifically, the windowed information geometry channel estimation method can be performed by a windowed information geometry channel estimation device, which can be implemented by software and / or hardware and integrated in an electronic device. Electronic devices include but are not limited to computers, smart phones, servers, and other devices with image processing functions.

[0031] like Figure 1 As shown, the method specifically comprises the following steps:

[0032] S110. Initialize an average natural parameter of the auxiliary manifold, wherein the average natural parameter of the auxiliary manifold is the mean of the natural parameters of all auxiliary manifolds, and the average natural parameter of the auxiliary manifold is used to approximate the nonlinear term corresponding to the channel estimation.

[0033] In this embodiment, the auxiliary manifold average natural parameter can be used to approximate the nonlinearity in the probability density function of the channel estimation. The target manifold natural parameter can be understood as the natural parameter of the target manifold, including the first-order natural parameter and the second-order natural parameter, and the auxiliary manifold average natural parameter can be understood as the mean of the natural parameters of all auxiliary manifolds, including the auxiliary manifold average first-order natural parameter and the auxiliary manifold average second-order natural parameter.

[0034] S120. In each iteration, based on the windowed received signal model, the introduced window function is used to calculate the approximate solution of the m-projection of the auxiliary manifold in this iteration to the target manifold in this iteration, and the average natural parameters of the auxiliary manifold are updated according to the damping coefficient and the approximate solution of the m-projection; wherein the windowed received signal model is obtained by introducing the window function into the received signal model.

[0035] Specifically, a window function is introduced on the basis of the received signal model, such as a sine window, a rectangular window, a Hanning window, a Hamming window or a Blackman window function, to obtain a windowed received signal model. The window function can be used to truncate or modulate the signal in the time domain, and subsequent calculations are performed on the part of the received signal extracted using the window function to approximate the channel estimation process. The purpose of channel estimation is to calculate the probability density of the posterior distribution corresponding to the windowed received signal model. There are nonlinear terms in its probability density function, which makes the calculation complexity high. In this embodiment, the introduced window function is used to calculate the approximate solution of the m-projection. In this process, the nonlinear terms can be approximated, and the approximate solution is used to update the average natural parameters of the auxiliary manifold and perform multiple iterations, thereby reducing the complexity of the calculation. Among them, the m-projection is a distribution on the target manifold, and the distribution satisfies the minimum Kulbeck-Leibler divergence to the auxiliary manifold.

[0036] S130 , after the iteration stops, updating the target manifold natural parameters according to the auxiliary manifold average natural parameters, and determining the channel estimation result according to the target manifold natural parameters.

[0037] Specifically, the auxiliary manifold average natural parameters are updated in each iteration, while the target manifold natural parameters are updated after the iteration. Channel estimation can be understood as estimating the model parameters of the assumed channel model from the received data.

[0038] An embodiment of the present application provides a windowing-based information geometry channel estimation method, which obtains a windowed received signal model by introducing the window function into the received signal model, and performs information geometry channel estimation based on the windowed received signal model, and uses the window function to simplify the calculation of m-projection, thereby reducing the complexity of channel estimation and improving the efficiency of channel estimation.

[0039] In one embodiment, the windowed received signal model is: Where W is a diagonal matrix constructed according to the window function, y is obtained by vectorizing the received signal, A is a measurement matrix consisting of a delay domain sampling matrix and an angle domain sampling matrix, is Gaussian white noise, and the noise variance is is the windowed mean noise variance, is a mean of the squares of the modulus values ​​of the window function elements, is the noise variance.

[0040] Specifically, in the embodiments of the present application, the channel estimation of the MIMO-OFDM system is mainly described. The system model is as follows: The base station is equipped with N r =N r，v ×N r，h The uniform array of antennas has N antennas in the vertical and horizontal directions. r,v and N r,h The number of users is K, and each user is equipped with a single antenna. For OFDM modulation, the number of subcarriers is N c , the system sampling interval is T s , the cyclic prefix is ​​N g The subcarriers used for uplink channel estimation are Among them, N1 and are the starting and last subcarrier index numbers respectively, N p is the total number of subcarriers used for channel estimation, and i is the subcarrier number.

[0041] For uplink transmission, at the i-th subcarrier, the frequency domain received signal is:

[0042]

[0043] in, To receive the signal, [Y] j,i is the received signal of the i-th subcarrier on the j-th antenna. is a diagonal matrix, is the pilot signal sent, where [X k ] i,i is the pilot signal of user k on the i-th subcarrier. is the spatial frequency domain signal matrix of user k, [G] j,i is the channel between user k and the jth antenna on subcarrier i. is Gaussian white noise, and the noise variance is [Z] j,i is the noise on the i-th subcarrier on the j-th antenna.

[0044] Bringing in the beam domain channel model, the relationship between the beam domain channel and the space-frequency domain channel can be written as:

[0045] G k =VH k F T (2)

[0046] in, is the statistical beam domain channel matrix; is the delay domain sampling matrix, N g is the cyclic prefix length, N c is the total number of subcarriers, F τ is the delay domain oversampling factor. is the angle domain sampling matrix. is the angular domain sampling matrix in the vertical direction, N v =F v N r，v , F v is the oversampling factor in the vertical direction. is the angle domain sampling matrix in the horizontal direction, N h =F h N r，h , F h is the oversampling factor in the horizontal direction.

[0047] Substituting formula (2) into the received signal model yields:

[0048] Y=VH a M+Z (3)

[0049] in,

[0050] The beam domain energy matrix is ​​defined as:

[0051]

[0052] Due to the sparseness of the channel, the values ​​of many elements in the beam domain energy matrix are close to zero. In the algorithm, these elements with small energy can be set to zero.

[0053] Vectorize formula (3) to obtain the received signal model:

[0054] y=Ah+z (5)

[0055] Let h be vec(H a ) is the vector after removing the terms with zero energy, for The matrix after removing the corresponding columns, N = N r N p , M is H a The number of terms with non-zero variance. y and z are vectorized from Y and Z respectively. The prior distribution of h follows a complex Gaussian distribution, i.e. Where D = diag (d) is the prior variance matrix, and d is the vector after vec {Ω} removes the elements with value 0. Its posterior probability distribution p (h | y) = p (h; μ, ∑) is still a complex Gaussian distribution Where μ is the expectation (mean) of the complex Gaussian distribution, and ∑ is the variance of the complex Gaussian distribution.

[0056] Based on formula (5), the window function is introduced to obtain the windowed received signal model:

[0057]

[0058] in is a diagonal matrix and w is the window function. is Gaussian white noise, and the noise variance is ω n =([W] n，n ) * [W] n，n , ω n An average value of, for example, ω n The arithmetic mean of Or ω n The harmonic mean of wait.

[0059] The purpose of channel estimation is to calculate the probability density of the posterior distribution of h, and its probability density function can be written as:

[0060]

[0061] Among them, y n is the nth element of y, w nis the nth element of w. n is the number of nonlinear terms, p s (h s ) is the probability density function of the sth element in h, and the operator O is the inner product of complex variables (defined as x and y are vectors. X and Y are matrices). Define function f(x,X)=[x T ,vec T (X)] T , d h =f(0, -D -1 ), t h =f(h,Ie(hh H )). C, C' and is the normalization factor. n A H The nth column of c n (h) is a nonlinear term, written as

[0062]

[0063] In formula (7), due to the nonlinear term The existence of makes the computational complexity high. The approximate solution of m-projection is calculated using the window function. In this process, the nonlinear terms can be approximated, thereby reducing the computational complexity.

[0064] In one embodiment, the auxiliary manifold average natural parameters include auxiliary manifold average first-order natural parameters and auxiliary manifold average second-order natural parameters; initializing the auxiliary manifold average natural parameters includes: initializing the auxiliary manifold average first-order natural parameters to an arbitrary value; initializing the auxiliary manifold average second-order natural parameters to a diagonal matrix, wherein the diagonal elements in the diagonal matrix are all less than or equal to 0.

[0065] In this embodiment, the average first-order natural parameter of the auxiliary manifold can be initialized to any value Initialize the auxiliary manifold average second-order natural parameters to a diagonal matrix The elements on the diagonal must be real numbers less than or equal to 0.

[0066] Specifically, in order to approximate the nonlinear term, the target manifold is defined:

[0067]

[0068] Where ψ0 = f(θ0, θ0) is the natural parameter of the target manifold, is the first-order natural parameter of the target manifold, Θ0 is an M×M real diagonal matrix, and is the second-order natural parameter of the target manifold. Comparing formula (9) with formula (7), it can be seen that the target manifold uses ψ0ot h Approximately Transform it into the traditional Gaussian distribution probability density function, that is

[0069]

[0070] The relationship between the natural parameters of the target manifold (including first-order and second-order natural parameters) and the mean and variance of the target manifold is:

[0071]

[0072] Right now

[0073] ∑0=(D -1 -Θ0) -1 (12)

[0074]

[0075] Among them, μ0 is the mean of the target manifold, ∑0 is the variance of the target manifold, the superscript “-1” indicates the inverse of the matrix, and D is the prior variance matrix.

[0076] In addition, the computational complexity of directly using the target manifold to approximate the posterior probability is still very high, so an auxiliary manifold is defined:

[0077]

[0078] ψ n =f(θ n , Θ n ) is the average natural parameter of the auxiliary manifold, Θ n is a real diagonal matrix of M×M. The auxiliary manifold average natural parameters include the auxiliary manifold average first-order natural parameters (θ N ) and the auxiliary manifold average second-order natural parameter (Θ N ),Right now

[0079]

[0080] Comparing formula (14) with formula (7), we can see that the auxiliary manifold is represented by ψ n ot h Approximately These n-1 nonlinear terms are transformed into the traditional Gaussian distribution probability density function, that is,

[0081]

[0082] The relationship between the natural parameters of the auxiliary manifold and the mean and variance of the auxiliary manifold is:

[0083]

[0084] Among them, μn is the auxiliary manifold mean, ∑ n is the variance of the auxiliary manifold, and D is the prior variance matrix.

[0085] In addition, in this embodiment, during initialization, there is an equation holds true, t represents the number of iterations, then the formula holds true in subsequent iterations, that is, each round of iteration only needs to update the average natural parameters of the auxiliary manifold, and after the iteration is completed, the natural parameters of the target manifold can be updated according to the corresponding average natural parameters of the auxiliary manifold as follows:

[0086] In one embodiment, based on the windowed received signal model, calculating the m-projection of the auxiliary manifold in the current iteration to the target manifold in the current iteration includes:

[0087] Let the partial derivative of Kullback-Leibler divergence, i.e. KL divergence, be 0, and use the introduced window function for approximate calculation to obtain the approximate solution of the mean and variance of the m-projection.

[0088] According to the approximate solutions of the mean and variance of the m-projection and the relationship between the mean variance and the natural parameters on the target manifold, the approximate solutions of the first-order natural parameters and the second-order natural parameters of the m-projection are calculated again using the window function;

[0089] The m-projection is determined according to an approximate solution of the first-order natural parameter and the second-order natural parameter.

[0090] Specifically, KL divergence can be used to measure the degree of match between two distributions. At the tth iteration, the auxiliary manifold M n Perform m-projection to the target manifold M0 and record the obtained natural parameters as

[0091]

[0092] Denotes the auxiliary manifold M n m-projection to the target manifold M0, which is on the target manifold M0 and M n The KL divergence between the probability distributions of is the smallest, that is,

[0093]

[0094] To simplify the representation, the superscript t is omitted in the following process and added during iteration.

[0095]

[0096] Where det is the determinant of the matrix and tr is the trace of the matrix. Taking the partial derivative of the above equation and setting it equal to zero, we can get

[0097] μ 0n =μ n (twenty two)

[0098]

[0099] Where I represents the identity matrix. Step a of formula (23) is derived from the fact that in actual channel estimation, the matrix A satisfies the property of amplitude 1, that is, |[A] ij |=1. In step b, the formula is simplified by using the window function. Specifically, in step b, ω in the second term is n use Substitution to simplify calculations.

[0100] According to formula (15) and formula (16), using the parameter relationship on the target manifold in formula (12), we can get

[0101]

[0102] Substituting into formula (22) and formula (23) and further calculating, we can get the approximate solutions of the first-order and second-order natural parameters of the m-projection as follows:

[0103]

[0104]

[0105] Steps of formula (26) a It comes from replacing ω in the denominator of the second term n use In the above formula, the calculation is simplified again with the help of the window function. Specifically, the ω in the denominator of the second term is n use The substitution was approximated.

[0106] Based on the approximate solutions of the first-order natural parameters and the second-order natural parameters, the m-projection can be determined as

[0107] In one embodiment, the auxiliary manifold average natural parameter is updated according to the damping coefficient and the approximate solution of the m-projection as the sum of the following two products:

[0108] The first coefficient is the product of the first-order natural parameter or the second-order natural parameter of the m-projection, and the second coefficient is the product of the auxiliary manifold average natural parameter of the previous iteration;

[0109] The first coefficient is The second coefficient is (1-αN), where α is the damping coefficient and N is the spatial-frequency domain channel dimension.

[0110] In this embodiment, the approximate solution of the m-projection updates the parameters on the auxiliary manifold. From formula (14), we can get: Approximately c n (h), and ψ n Yes So we can get:

[0111]

[0112] Where α is a damping factor introduced to increase the stability of the algorithm. According to the definitions of formulas (24) and (25), the update of the average natural parameters of the auxiliary manifold is:

[0113]

[0114] In one embodiment, after the iteration stops, the target manifold natural parameters are updated to the average natural parameters of the auxiliary manifolds in this iteration.

[0115] For the update of the natural parameters of the target manifold, it can be found through calculation that if the initialization is specified, there is an equation If it holds, then this formula holds true in each subsequent iteration. So the natural parameter update of the target manifold is:

[0116]

[0117] In one embodiment, determining a channel estimation result according to a target manifold natural parameter includes:

[0118] Determine an estimate of a beam domain channel according to a natural parameter of a target manifold;

[0119] The space-frequency domain channel is restored according to the estimated value of the beam-domain channel and the relationship between the beam-domain channel and the space-frequency domain channel to obtain a channel estimation result.

[0120] Specifically, convert the natural parameters of the target manifold into mean values:

[0121]

[0122] in, It is the mean value obtained by natural parameter transformation of the target manifold, that is, the estimated value of the beam domain channel h, T m is the number of iterations, calculate After that, the spatial frequency domain channel can be restored using formula (2), thus completing the entire channel estimation process.

[0123] In summary, the overall process of the information geometry channel estimation method based on windowing is as follows:

[0124] 1. Initialization: t = 0, select a suitable damping coefficient α, initialize the average natural parameters of the auxiliary manifold, including any initialization of the first-order natural parameters And the second-order natural parameter Initialize to a diagonal matrix;

[0125] 2. In this iteration, the m-projection is calculated according to equations (25) and (26);

[0126] 3. Update the average natural parameters of the auxiliary manifold according to equations (28) and (29);

[0127] 4. t = t + 1, repeat 2 to 3 times until the number of iterations T m ;

[0128] 5. Update the natural parameters of the target manifold according to equations (30) and (31);

[0129] 6. The target manifold mean is calculated according to formula (32);

[0130] 7. Calculate the spatial-frequency domain channel according to formula (2) to obtain the channel estimation result.

[0131] The information geometry channel estimation method based on windowing in the embodiment of the present application introduces a window function into the received signal model, proposes a windowed received signal model, and derives a windowed information geometry algorithm based on the windowed received signal model. The advantage of the windowed information geometry algorithm is its low complexity. The most complex step of the information geometry algorithm is A in the m-projection calculation. H Av is the term, where v is a vector. This method introduces a window function to transform the above term into the term A in formula (26) H W H WAv, where v = (D -1 -Θ N ) -1 θ N . Based on the selection of a suitable window function, such as the widely used sine window, It can be converted into a sparse matrix, that is, many elements in u are zero, so when multiplying with v, the multiplication of zero elements does not need to be calculated, which greatly reduces the complexity of the algorithm. Performance simulation shows that the complexity of this method is about 20 times lower than that of the original information geometry algorithm.

[0132] Figure 2 A schematic diagram of the structure of a windowing-based information geometry channel estimation device provided in an embodiment of the present application. The windowing-based information geometry channel estimation device provided in this embodiment includes:

[0133] An initialization module 210, used to initialize an auxiliary manifold average natural parameter, wherein the auxiliary manifold average natural parameter is the mean of the natural parameters of all auxiliary manifolds, and the auxiliary manifold average natural parameter is used to approximate the nonlinear term corresponding to the channel estimation;

[0134] Iteration module 220 is used to calculate, in each iteration, based on the windowed received signal model, an approximate solution of the m-projection of the auxiliary manifold in this iteration to the target manifold in this iteration using the introduced window function, and update the auxiliary manifold average natural parameter according to the damping coefficient and the approximate solution of the m-projection; wherein the windowed received signal model is obtained by introducing the window function into the received signal model; the m-projection is a distribution on the target manifold, and the KL divergence between the distribution of the auxiliary manifold and the distribution on the target manifold is the smallest;

[0135] The channel estimation module 230 is used to update the target manifold natural parameters according to the auxiliary manifold average natural parameters after the iteration stops, and determine the channel estimation result according to the target manifold natural parameters.

[0136] Based on the above embodiment, the windowed received signal model is: Where W is a diagonal matrix constructed according to the window function, y is obtained by vectorizing the received signal, A is a measurement matrix consisting of a delay domain sampling matrix and an angle domain sampling matrix, is Gaussian white noise, and the noise variance is is the windowed mean noise variance, is a mean of the squares of the modulus values ​​of the window function elements, is the noise variance.

[0137] Based on the above embodiment, the initialization module 210 is used to:

[0138] The auxiliary manifold average natural parameters include first-order natural parameters and second-order natural parameters;

[0139] The initialization of the auxiliary manifold average natural parameters includes:

[0140] Initializing the first-order natural parameter to an arbitrary value;

[0141] The second-order natural parameters are initialized to a diagonal matrix, and all diagonal elements in the diagonal matrix are less than or equal to 0.

[0142] On the basis of the above embodiment, based on the windowed received signal model, calculating the m-projection of the auxiliary manifold in this iteration to the target manifold in this iteration includes:

[0143] Let the partial derivative of KL divergence be 0, and use the introduced window function for approximate calculation to obtain the approximate solution of the mean and variance of the m-projection;

[0144] According to the approximate solutions of the mean and variance of the m-projection and the relationship between the mean variance and the natural parameters on the target manifold, the window function is used again to calculate the approximate solutions of the first-order natural parameters and the second-order natural parameters of the m-projection;

[0145] The m-projection is determined according to an approximate solution of the first-order natural parameter and the second-order natural parameter.

[0146] On the basis of the above embodiment, the average natural parameter of the auxiliary manifold is updated according to the damping coefficient and the approximate solution of the m-projection as the sum of the following two products:

[0147] The first coefficient is the product of the first-order natural parameter or the second-order natural parameter of the m-projection, and the second coefficient is the product of the auxiliary manifold average natural parameter of the previous iteration;

[0148] The first coefficient is The second coefficient is (1-αN), where α is the damping coefficient and N is the spatial-frequency domain channel dimension.

[0149] Based on the above embodiment, after the iteration stops, the target manifold natural parameters are updated to the average natural parameters of the auxiliary manifold in this iteration.

[0150] Based on the above embodiment, the channel estimation module 230 is used for:

[0151] Determining an estimated value of a beam domain channel according to the target manifold natural parameter;

[0152] The space-frequency domain channel is restored according to the estimated value of the beam-domain channel and the relationship between the beam-domain channel and the space-frequency domain channel to obtain a channel estimation result.

[0153] The information geometry channel estimation device based on windowing provided in the embodiment of the present application can be used to execute the information geometry channel estimation method based on windowing provided in any of the above embodiments, and has corresponding functions and beneficial effects.

[0154] Figure 3A schematic diagram of an electronic device 10 that can be used to implement an embodiment of the present application is shown. The electronic device 10 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device 10 can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, user equipment, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or required herein.

[0155] like Figure 3 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0156] Multiple components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks, wireless networks.

[0157] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 performs the various methods and processes described above.

[0158] In some embodiments, the method of the above embodiment may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to perform any of the above embodiment methods in any other appropriate manner (e.g., by means of firmware).

[0159] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0160] The computer programs for implementing the methods of the present application may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer programs are executed by the processor, the functions / operations specified in the flow charts and / or block diagrams are implemented. The computer programs may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0161] In the context of the present application, a computer readable storage medium may be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, device or equipment. A computer readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer readable storage medium may be a machine readable signal medium. A more specific example of a machine readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0162] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device 10 having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device 10. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0163] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0164] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0165] An embodiment of the present application also provides a computer program product, including a computer program and / or instructions, which, when executed by a processor, implements the windowed information geometry channel estimation method as described in any of the above embodiments.

[0166] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this application can be executed in parallel, sequentially or in different orders, as long as the expected results of the technical solution of this application can be achieved, and this document is not limited here.

[0167] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A windowing-based information geometry channel estimation method, characterized in that: include: Initializing an auxiliary manifold average natural parameter, where the auxiliary manifold average natural parameter is the mean of the natural parameters of all auxiliary manifolds, and the auxiliary manifold average natural parameter is used to approximate a nonlinear term corresponding to channel estimation; In each iteration, based on the windowed received signal model, the approximate solution of the m-projection of the auxiliary manifold in this iteration to the target manifold in this iteration is calculated by using the introduced window function, and the average natural parameter of the auxiliary manifold is updated according to the damping coefficient and the approximate solution of the m-projection; wherein the windowed received signal model is obtained by introducing the window function into the received signal model; the m-projection is a distribution on the target manifold, and the distribution satisfies the minimum Kulbeck-Leibler divergence to the auxiliary manifold; After the iteration stops, the target manifold natural parameters are updated according to the auxiliary manifold average natural parameters, and the channel estimation result is determined according to the target manifold natural parameters.

2. The method according to claim 1, characterized in that The windowed received signal model is: Where W is a diagonal matrix constructed according to the window function, y is obtained by vectorizing the received signal, A is a measurement matrix consisting of a delay domain sampling matrix and an angle domain sampling matrix, is Gaussian white noise, and the noise variance is is the windowed mean noise variance, is a mean of the squares of the modulus values ​​of the window function elements, is the noise variance.

3. The method according to claim 1, characterized in that The auxiliary manifold average natural parameters include auxiliary manifold average first-order natural parameters and auxiliary manifold average second-order natural parameters; The initialization of the auxiliary manifold average natural parameters includes: Initializing the average first-order natural parameter of the auxiliary manifold to an arbitrary value; The average second-order natural parameters of the auxiliary manifold are initialized to a diagonal matrix, and the diagonal elements in the diagonal matrix are all less than or equal to 0.

4. The method according to claim 1, characterized in that: Based on the windowed received signal model, the m-projection of the auxiliary manifold in this iteration to the target manifold in this iteration is calculated, including: Let the partial derivative of Kulbeck-Leibler KL divergence be 0, and use the introduced window function for approximate calculation to obtain the approximate solution of the mean and variance of the m-projection; According to the approximate solutions of the mean and variance of the m-projection and the relationship between the mean variance and the natural parameters on the target manifold, the window function is used again to calculate the approximate solutions of the first-order natural parameters and the second-order natural parameters of the m-projection; The m-projection is determined according to an approximate solution of the first-order natural parameter and the second-order natural parameter.

5. The method according to claim 1, characterized in that The average natural parameter of the auxiliary manifold is updated according to the damping coefficient and the approximate solution of the m-projection as the sum of the following two products: The first coefficient is the product of the first-order natural parameter or the second-order natural parameter of the m-projection, and the second coefficient is the product of the auxiliary manifold average natural parameter of the previous iteration; The first coefficient is The second coefficient is (1-αN), where α is the damping coefficient and N is the spatial-frequency domain channel dimension.

6. The method according to claim 5, characterized in that After the iteration stops, the target manifold natural parameters are updated to the average natural parameters of the auxiliary manifold 7. The method according to claim 1, characterized in that Determining a channel estimation result according to the target manifold natural parameter includes: Determining an estimated value of a beam domain channel according to the target manifold natural parameter; The space-frequency domain channel is restored according to the estimated value of the beam-domain channel and the relationship between the beam-domain channel and the space-frequency domain channel to obtain a channel estimation result.

8. A windowing-based information geometry channel estimation device, characterized in that: include: An initialization module, used for initializing an auxiliary manifold average natural parameter, wherein the auxiliary manifold average natural parameter is the mean of the natural parameters of all auxiliary manifolds, and the auxiliary manifold average natural parameter is used for approximating the nonlinear term corresponding to the channel estimation; An iteration module is used to calculate, in each iteration, an approximate solution of the m-projection of the auxiliary manifold in this iteration to the target manifold in this iteration using the introduced window function based on the windowed received signal model, and update the auxiliary manifold average natural parameter according to the damping coefficient and the approximate solution of the m-projection; wherein the windowed received signal model is obtained by introducing the window function into the received signal model; the m-projection is a distribution on the target manifold, and the distribution satisfies the minimum Kulbeck-Leibler divergence to the auxiliary manifold; The channel estimation module is used to update the target manifold natural parameters according to the auxiliary manifold average natural parameters after the iteration stops, and determine the channel estimation result according to the target manifold natural parameters.

9. An electronic device, characterized in that: include: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the windowing-based information geometry channel estimation method as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the windowing-based information geometry channel estimation method as described in any one of claims 1 to 7 is implemented.

11. A computer program product comprising a computer program and / or instructions, characterized in that: When the computer program and / or instructions are executed by a processor, the windowing-based information geometry channel estimation method as described in any one of claims 1-7 is implemented.

Images