A two-stage cascaded channel estimation method, apparatus, device, medium and product

Abstract

The application relates to the technical field of wireless communication, and in particular provides a two-stage concatenated channel estimation method, device, equipment, medium and product. The method comprises the following steps: performing compressed sensing on an initial system model to obtain a communication system model; calculating a grid index value based on a base station measurement matrix and a sparse measurement matrix included in the communication system model; determining an index interval to which an angle of arrival belongs based on the grid index value; updating the sparse measurement matrix based on the index interval to obtain a sparse representation matrix; and obtaining a compressed concatenated channel matrix based on the sparse representation matrix and a concatenated channel matrix in the communication system model. By determining the index interval to which the angle of arrival belongs, the real angle of arrival can be obtained. Based on the sparsity of the angle domain, the sparse measurement matrix is converted into a sparse representation matrix with a lower dimension. The off-grid error is avoided, the calculation complexity is reduced, and the calculation accuracy and precision of the compressed concatenated channel matrix are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wireless communication, and in particular to a two-stage cascaded channel estimation method, device, equipment, medium and product. BACKGROUND

[0002] In recent years, reconfigurable intelligent surfaces (RIS) have been proposed for use in wireless communication systems due to their low hardware cost and low energy consumption, to enhance the coverage and capacity of wireless communication systems, and RIS is also recognized as one of the key technologies for future sixth-generation wireless communication networks (6G). In order to quantify channel characteristics (such as attenuation and phase) and compensate for transmitted signals to reduce the difference between transmitted signals and received signals, channel estimation is required. Since a RIS-assisted large-scale multiple-input multiple-output (MIMO) antenna system consists of hundreds of antenna elements, and includes two-stage cascaded channels from user equipment (UE) to RIS and from RIS to base station (BS), the difficulty and complexity of channel estimation are dramatically increased.

[0003] Existing two-stage cascaded channel estimation methods propose that cascaded channels share completely common non-zero rows and partially common non-zero columns according to different angles of different user equipment, to achieve channel estimation. However, this method does not take into account the problem of direction mismatch, resulting in off-network errors, which violates the sparsity assumption, not only leading to a dramatic increase in pilot consumption and computational complexity, but also seriously affecting the accuracy of cascaded channel estimation. SUMMARY

[0004] The present application is proposed in view of the above problems, and provides a two-stage cascaded channel estimation method, device, equipment, medium and product.

[0005] According to one aspect of the present application, a two-stage cascaded channel estimation method is provided, comprising:

[0006] Compressing the constructed initial system model to obtain a communication system model, the initial system model including a two-stage cascaded channel of a user-to-RIS side channel and a RIS-side-to-base station side channel;

[0007] Calculating a grid index value based on a base station measurement matrix and a sparse measurement matrix included in the initial system model;

[0008] Determining an index interval to which an angle of arrival belongs based on the grid index value, the angle of arrival being an angle of arrival of a signal received by the base station side;

[0009] Updating the sparse measurement matrix based on the index interval to obtain a sparse representation matrix;

[0010] Based on the sparse representation matrix and a concatenated channel matrix in the communication system model, a compressed concatenated channel matrix is obtained.

[0011] In addition, the two-stage concatenated channel estimation method according to one aspect of the present application further comprises:

[0012] The constructed initial system model is subjected to compressed sensing to obtain a communication system model, comprising:

[0013] Based on a first dictionary matrix and a second dictionary matrix of the discrete grid points generated based on the initial system model, the concatenated channel in the initial system model is updated, the first dictionary matrix being a uniform discrete array of the angle of arrival on the base station side, and the second dictionary matrix being a uniform discrete array of the angle of arrival on the reconfigurable intelligent surface side;

[0014] Based on the updated concatenated channel, the initial system model is rewritten to obtain the communication system model.

[0015] In addition, the two-stage concatenated channel estimation method according to one aspect of the present application further comprises:

[0016] Based on the grid index value, an index interval to which the angle of arrival belongs is determined, comprising:

[0017] Based on the grid index value, a right-side adjacent index value and a left-side adjacent index value of the grid index value are obtained;

[0018] A first channel coefficient corresponding to the left-side adjacent index value and a second channel coefficient corresponding to the right-side adjacent index value are calculated;

[0019] In a case where the first channel coefficient is greater than or equal to the second channel coefficient, the index interval to which the angle of arrival belongs is determined as an interval composed of the left-side adjacent index value and the grid index value;

[0020] In a case where the first channel coefficient is less than the second channel coefficient, the index interval to which the angle of arrival belongs is determined as an interval composed of the grid index value and the right-side adjacent index value.

[0021] In addition, the two-stage concatenated channel estimation method according to one aspect of the present application further comprises:

[0022] Based on the index interval, a sparse measurement matrix in the communication system model is updated to obtain a sparse representation matrix, comprising:

[0023] A sample parameter is selected in the index interval, and based on the sample parameter, a support set of the angle of arrival on the base station side is obtained;

[0024] generating a first dictionary matrix of discrete grid points based on the support set and the initial system model, computing an array vector;

[0025] updating a sparse measurement matrix in the communication system model based on the array vector, obtaining a sparse representation matrix.

[0026] In addition, the two-stage cascaded channel estimation method according to one aspect of the present application further comprises:

[0027] obtaining a compressed cascaded channel matrix based on the sparse representation matrix and a cascaded channel matrix in the communication system model, comprising:

[0028] performing dimension reduction on the cascaded channel matrix in the communication system model based on the sparse representation matrix, obtaining a pattern formula comprising the compressed cascaded channel matrix;

[0029] calculating the compressed cascaded channel matrix based on an expectation maximization algorithm and the pattern formula.

[0030] In addition, the two-stage cascaded channel estimation method according to one aspect of the present application further comprises: calculating the compressed cascaded channel matrix based on an expectation maximization algorithm and the pattern formula, comprising:

[0031] modeling the compressed cascaded channel matrix based on an expectation maximization algorithm and the pattern formula, obtaining a modeling formula of the compressed cascaded channel matrix;

[0032] obtaining a calculation formula of a first posterior parameter based on the modeling formula, the first posterior parameter being a mean value of a random variable of the compressed cascaded channel matrix;

[0033] transforming the calculation formula of the first posterior parameter to move a second posterior parameter in the calculation formula to the other side of an equal sign; and, according to a diagonal estimation criterion, obtaining a calculation formula of a diagonal vector of the second posterior parameter, transforming the calculation formula of the diagonal vector to move the second posterior parameter to the other side of an equal sign in the calculation formula of the diagonal vector, the second posterior parameter being a covariance of the random variable of the compressed cascaded channel matrix;

[0034] calculating the first posterior parameter based on an expectation maximization algorithm, the transformed calculation formula of the first posterior parameter, and the transformed calculation formula of the diagonal vector.

[0035] calculating the compressed cascaded channel matrix based on the first posterior parameter and the pattern formula.

[0036] According to another aspect of the present application, a two-stage cascaded channel estimation device is provided, comprising:

[0037] a compression sensing module configured to perform compression sensing on the constructed initial system model to obtain a communication system model, the initial system model comprising a two-stage concatenated channel of a user-to-reconfigurable intelligent surface side channel and a reconfigurable intelligent surface side-to-base station side channel;

[0038] a first calculation module configured to calculate a grid index value based on a base station measurement matrix and a sparse measurement matrix comprised in the initial system model;

[0039] an interval determination module configured to determine an index interval to which an angle of arrival belongs based on the grid index value, the angle of arrival being an angle of arrival of a signal received at the base station side;

[0040] an updating module configured to update the sparse measurement matrix based on the index interval to obtain a sparse representation matrix;

[0041] a second calculation module configured to obtain a compressed concatenated channel matrix based on the sparse representation matrix and a concatenated channel matrix in the communication system model.

[0042] According to yet another aspect of the present application, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory, the processor executing the computer program to implement the method of the above aspect.

[0043] According to yet another aspect of the present application, a computer readable storage medium is provided, having stored thereon a computer program, the computer program being executed by a processor to implement the method of the above aspect.

[0044] According to yet another aspect of the present application, a computer program product is provided, comprising a computer program, the computer program being executed by a processor to implement the method of the above aspect.

[0045] As will be described in detail below, according to the two-stage concatenated channel estimation method, device, equipment, medium and product of embodiments of the present application, by determining the index interval to which the angle of arrival belongs, the real angle of arrival can be obtained, the non-convex problem is converted into a convex optimization problem, based on the sparsity of the angle domain, the sparse measurement matrix is converted into a sparse representation matrix, the sparse representation matrix only includes a dictionary of the real angle of arrival, the dimension is lower, not only the off-network error is avoided, but also the calculation complexity is reduced, and the calculation precision and accuracy of the compressed concatenated channel matrix are improved.

[0046] It is to be understood that both the foregoing general description and the following detailed description are exemplary and intended to provide further explanation of the subject technology. BRIEF DESCRIPTION OF DRAWINGS

[0047] The above and other objects, features and advantages of the present application will become more apparent from the following detailed description thereof taken in conjunction with the accompanying drawings in which:

[0048] Figure 1 is a flow chart illustrating a two-stage cascaded channel estimation method according to an embodiment of the present application.

[0049] Figure 2 is a flow chart illustrating a calculation of a compressed cascaded channel matrix according to an embodiment of the present application.

[0050] Figure 3 is a flow chart illustrating a two-stage cascaded channel estimation method according to another embodiment of the present application.

[0051] Figure 4 is a structural schematic diagram of a two-stage cascaded channel estimation apparatus according to an embodiment of the present application.

[0052] Figure 5 is a structural schematic diagram of a computer device according to an embodiment of the present application.

[0053] Figure 6 is a schematic diagram of a computer program product according to an embodiment of the present application. DETAILED DESCRIPTION

[0054] In order to make the objects, technical solutions and advantages of the present application more apparent, the following will describe the example embodiments according to the present application in detail with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, and not all the embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein.

[0055] In recent years, reconfigurable intelligent surfaces (RIS) have been proposed for wireless communication systems due to their low hardware cost and low energy consumption, to enhance the coverage and capacity of wireless communication systems, and RIS is also recognized as one of the key technologies for future sixth generation wireless communication networks (6G). In order to quantify channel characteristics (such as attenuation and phase), to compensate for the difference between the transmitted signal and the received signal, channel estimation is needed. Since the RIS-aided massive multiple-input multiple-output (MIMO) antenna system is composed of hundreds of antenna elements, and includes two-stage cascaded channels from user equipment (UE) to RIS and from RIS to base station (BS), the difficulty and complexity of channel estimation are dramatically increased.

[0056] Existing two-stage cascaded channel estimation methods propose that the cascaded channel shares completely common non-zero rows and partially common non-zero columns based on the different perspectives of different user equipment, in order to achieve channel estimation. However, this method does not consider the problem of direction mismatch, resulting in off-network error and violating the sparsity assumption. This not only leads to a sharp increase in pilot consumption and computational complexity, but also seriously affects the accuracy of cascaded channel estimation.

[0057] The above description, with reference to the accompanying drawings, illustrates a two-stage cascaded channel estimation method, apparatus, device, medium, and product according to embodiments of this application. By determining the index interval to which the angle of arrival belongs, the true angle of arrival can be obtained, transforming a non-convex problem into a convex optimization problem. Based on the sparsity of the angle domain, the sparse measurement matrix is ​​transformed into a sparse representation matrix. The sparse representation matrix only includes a dictionary containing the true angle of arrival, resulting in lower dimensionality. This not only avoids off-network errors but also reduces computational complexity and improves the computational accuracy and precision of the compressed cascaded channel matrix.

[0058] By transforming the calculation formulas for the first posterior parameter and the diagonal vector, the second posterior parameter in the calculation formula is moved to the other side of the equal sign. The inversion operation of the second posterior parameter itself is transformed into a method of solving equations, which greatly reduces the computational complexity, helps to solve the sparsity problem of the underlying block, introduces mode coupling prior, and reduces pilot overhead.

[0059] To facilitate understanding of this embodiment, a detailed description of the two-stage cascaded channel estimation method disclosed in this application embodiment will be provided first. The execution entity of the two-stage cascaded channel estimation method provided in this application embodiment is generally a computer device with certain computing capabilities. This computer device may include, for example, a terminal device, a server, or other processing devices. The terminal device may be a user equipment (UE), mobile device, user terminal, terminal, cellular phone, cordless phone, personal digital assistant (PDA), handheld device, computing device, vehicle-mounted device, wearable device, etc. In some possible implementations, this two-stage cascaded channel estimation method can be implemented by the processor calling computer-readable instructions stored in memory.

[0060] like Figure 1 The diagram shows a flowchart of a two-stage cascaded channel estimation method provided in an embodiment of this application. The method includes S1-S5:

[0061] S1: Compressed sensing is applied to the constructed initial system model to obtain the communication system model.

[0062] The communication system model comprises a two-stage concatenated channel of a user-to-reconfigurable intelligent surface side channel and a reconfigurable intelligent surface side-to-base station side channel, and is applied to a reconfigurable intelligent surface (RIS)-assisted massive multiple-input multiple-output communication system.

[0063] S1.1: generating a first dictionary matrix and a second dictionary matrix of discrete grid points based on an initial system model, and updating the concatenated channel in the initial system model.

[0064] The first dictionary matrix is a uniform discrete array of angles of arrival at the base station side, and the second dictionary matrix is a uniform discrete array of angles of arrival at the reconfigurable intelligent surface side.

[0065] S1.2: rewriting the initial system model based on the updated concatenated channel to obtain the communication system model.

[0066] S2: calculating a grid index value based on a base station measurement matrix and a sparse measurement matrix included in the communication system model.

[0067] S3: determining an index interval to which an angle of arrival belongs based on the grid index value.

[0068] The angle of arrival is an angle of arrival of a signal received at the base station side, and S3 specifically comprises the following steps:

[0069] S3.1: obtaining a right-side adjacent index value and a left-side adjacent index value of the grid index value based on the grid index value.

[0070] S3.2: calculating a first channel coefficient corresponding to the left-side adjacent index value and a second channel coefficient corresponding to the right-side adjacent index value.

[0071] S3.3: in the case where the first channel coefficient is greater than or equal to the second channel coefficient, determining that the index interval to which the angle of arrival belongs is an interval composed of the left-side adjacent index value and the grid index value.

[0072] In the case where the first channel coefficient is less than the second channel coefficient, it is determined that the index interval to which the angle of arrival belongs is an interval composed of the grid index value and the right-side adjacent index value.

[0073] S4: updating the sparse measurement matrix based on the index interval to obtain a sparse representation matrix. S4 specifically comprises the following steps:

[0074] S4.1: selecting a sample parameter in the index interval, and obtaining a support set of angles of arrival at the base station side based on the sample parameter.

[0075] S4.2: calculating an array vector based on the support set and the first dictionary matrix of discrete grid points generated by the initial system model.

[0076] S4.3: Based on the array vector, update the sparse measurement matrix in the communication system model to obtain the sparse representation matrix.

[0077] S5: Based on the sparse representation matrix and the cascaded channel matrix in the communication system model, obtain the compressed cascaded channel matrix.

[0078] S5 specifically includes the following steps:

[0079] S5.1: Based on the sparse representation matrix, the dimensionality of the cascaded channel matrix in the communication system model is reduced to obtain the mode formula including the compressed cascaded channel matrix;

[0080] S5.2: Calculate the compressed concatenated channel matrix based on the Expectation-Maximization (EM) algorithm and mode formula. Further, such as... Figure 2 As shown, S5.2 includes the following steps:

[0081] S5.2.1: Based on the expectation-maximization algorithm and the mode formula, the compressed concatenated channel matrix is ​​modeled to obtain the modeling formula of the compressed concatenated channel matrix;

[0082] S5.2.2: Based on the modeling formula, obtain the calculation formula for the first posterior parameter. Wherein, the first posterior parameter is the mean of the random variables of the compressed concatenated channel matrix;

[0083] S5.2.3: The calculation formula for the first posterior parameter is transformed so that the second posterior parameter in the calculation formula moves to the other side of the equality sign; and, according to the diagonal estimation criterion, the calculation formula for the diagonal vector of the second posterior parameter is obtained, and the calculation formula for the diagonal vector is transformed so that the second posterior parameter moves to the other side of the equality sign in the calculation formula for the diagonal vector. Wherein, the second posterior parameter is the covariance of the random variables of the compressed concatenated channel matrix;

[0084] S5.2.4: Calculate the first posterior parameter based on the expectation-maximization algorithm, the calculation formula of the first posterior parameter after deformation, and the calculation formula of the diagonal vector after deformation;

[0085] S5.2.5: Based on the first a posteriori parameter and the mode formula, the compressed concatenated channel matrix is ​​calculated.

[0086] like Figure 3 The diagram shown is another flowchart of the two-stage cascaded channel estimation method provided in this application embodiment. It should be noted that the symbols involved in this embodiment are explained as follows: (·) T Indicates transpose, (·) H This indicates the conjugate transpose, (·) -1 Indicates reversal. Indicates a false reversal; denotes the Kronecker product; vec(·) denotes the vectorization of a matrix; ||Δ||1and ||Δ||2denote the 1-norm and 2-norm of a matrix Δ; Tr(·) denotes the trace of a matrix; denotes a complex Gaussian distribution with mean vector μ and covariance matrix Σ; the operator diag(a) consists of a diagonal matrix from its vector elements; denotes the expected symbol. The method comprises steps 1-3:

[0087] Step 1: Establish a communication system model of RIS-aided massive MIMO.

[0088] Consider an uplink RIS-aided massive MIMO communication system, which consists of a base station (BS) with M (M ≥ 1) antennas, an RIS with N (N ≥ 1) elements of a uniform planar array, and a single-antenna user (UE). Without loss of generality, it is assumed that there is no direct link between the user and the BS, and only the reflected channel (BS-RIS channel and RIS-UE channel ) is considered, so that the effective uplink cascade channel of the user to the base station side is defined as:

[0089]

[0090] where G is the cascade channel, G RB is the BS-RIS channel, and h UR is the RIS-UE channel.

[0091] During channel estimation, different users (UEs) send orthogonal pilot sequences m t , for uplink cascade channel estimation, without loss of generality, it is assumed that m t = 1, and the initial system model can be represented as:

[0092] Y = GΓ + E (2)

[0093] where Y is the base station measurement matrix received at the BS from the user UE, Γ represents the phase reflection matrix, E represents an additive white Gaussian noise (AWGN) matrix, satisfying where σ represents the noise precision. According to the existing commonly used channel model, G RB and h UR can be represented as:

[0094] ​

[0095]

[0096] where L c is the number of scatterer paths of the RIS-BS, L g is the number of scatterer paths of the UE-RIS. is the complex gain of the l1th path, is the angle of arrival (AoA) at the BS, is the angle of departure (AoD) at the RIS, is the complex gain of the l2th path, is the angle of arrival (AoA) at the RIS, is the array steering vector associated with the RIS, is the array steering vector associated with the BS. As a function of χ and p, the array steering vector is defined as:

[0097]

[0098] where χ denotes the number of elements, θ denotes a function of the AOA or AOD, d denotes the antenna spacing, and typically satisfies λ is the carrier wavelength.

[0099] Since there are infinitely many atoms due to the random continuous values of AoAs / AoDs, the continuous parameter space and is discretized to a finite grid of points, in order to reduce the redundancy of the dictionary, a uniform sampling of the virtual angle domain is adopted, the AoA / AoD domain at the BS side and the RIS side contains a finite set of uniformly discrete grid points are defined as:

[0100]

[0101] where Θ (Bs) is the AoA / AoD domain at the BS side contains a finite set of uniformly discrete grid points, Θ (RIS) is the AoA / AoD domain at the RIS side contains a finite set of uniformly discrete grid points, is the mth grid value corresponding to the AoA / AoD domain at the BS side, sin(φ (n) ) is the nth grid value corresponding to the AoA / AoD domain at the RIS side, M G and N G are the grid resolutions at the BS side and the RIS side, respectively.

[0102] Therefore, the array response matrix at the BS side (i.e., the first dictionary matrix) of the uniformly discrete grid points is the array response matrix at the RIS side (the second dictionary matrix) of the uniformly discrete grid points Utilizing array response matrix and The cascaded channel G can be re-expressed in the virtual angle domain as:

[0103]

[0104] where, is the sparse representation of the cascaded channel, is the array response matrix of the uniform discrete grid points at the RIS side, further, the cascaded channel G can be expressed as:

[0105]

[0106] In the present embodiment, the following three properties are utilized to realize the two-stage cascaded channel estimation method:

[0107] Property 1, double structure property.

[0108] As can be seen from formula (9), there is only one non-zero element located at the position of the array response in the direction of and the direction of , therefore, the row and column indices of depend on and

[0109]

[0110] Property 2, block structure property.

[0111] Due to the dense position and close distance of the scatterers on the UE side, it leads to the similar or adjacent AoDs of the UE, thereby appearing sparse support in the block.

[0112] Property 3, BS-RIS angle is quasi-static compared with RIS-UE angle.

[0113] Because the positions of the BS and RIS are fixed and generally placed high, while the position of the mobile user UE is variable, the AoAs at the BS side can be regarded as static information, which means that it can be effectively estimated.

[0114] In order to fully utilize the significant sparsity of the cascaded channel G, the present embodiment re-formulates the channel estimation problem as a problem of recovering a sparse signal. Substituting formula (8) into formula (2), the initial system model is rewritten as a communication system model:

[0115]

[0116] where, is the array response matrix of the uniform discrete grid points at the BS side, is the sparse representation of the cascaded channel, is the array response matrix of the uniform discrete grid at the RIS side, Γ is the phase reflection matrix, and E represents the additive white Gaussian noise. With the help of sparse virtual representation and vector identity property, where A, B, C represent matrices respectively, and the communication system model is further rewritten as:

[0117] y = Φh + e (11)

[0118] where Φ represents a sparse measurement matrix, Based on Property 1 and Property 2, the can be represented as:

[0119]

[0120] where is the mth row of is the mth submatrix of Φ. To avoid confusion, h m and Φ m can be regarded as the mth group of h and Φ respectively.

[0121] Step 2: Construct the sparse representation matrix.

[0122] Since the true AoAs grid values are located around the coarsely divided grid and depend on the nearest grid points in the dictionary matrix. Therefore, the orthogonal projection strategy can be used to coarsely estimate the AoAs. Update the h non-zero group support set as:

[0123]

[0124] where is the cardinality of Finally contains all the non-zero group indices of h. is the maximum correlation index, in the nth iteration, the index with the maximum correlation is most likely to be a non-zero group of h, The calculation formula of is:

[0125]

[0126] where is the residual matrix obtained in the last iteration, is the submatrix containing the column support set selected from Φ. The channel coefficient is obtained by solving the least squares problem. The uniform discrete set Θ (BS) ​The most relevant index relative to the true AoAs is given by formula (15), which provides a coarse grid point estimate for each AoAs, i.e. However, considering the inherent grid mismatch problem in the grid method, this embodiment employs a simple numerical optimization algorithm to re-estimate the AoAs on the BS side through random search, in order to alleviate the energy leakage problem.

[0127] First, we use a rough estimate Limit the upper and lower limits of the search space to and in, For grid index values, and These are the grid index values. The left and right adjacent index values ​​are calculated using the least squares method. (Right now The first channel coefficient and (Right now The second channel coefficient The goal is to further narrow down the range of real-world physical perspectives. If... This means that the index range of the actual AOA is Otherwise For simplicity, this embodiment represents the index range of the AOA for each scatterer path as follows:

[0128] Secondly, in the index range Randomly select sample parameter Θ Δ Therefore, the initial support set of the re-estimated BS-side common AoAs is... Defined as:

[0129]

[0130] This means uniform sampling dictionary The columns only contain the re-estimated support set. The array response vector, i.e., the array vector in, It is an AoAs re-estimated support set The size of the sparse representation matrix. Reexpressed as:

[0131]

[0132] Based on the above description, the global optimum of the local optimization problem (i.e., Equation 18) can be solved for the BS-side AoA of each scattering path.

[0133]

[0134] wherein, is the residual matrix to be optimized, in order to accurately estimate the value of the next BS-side AoA, is connected vertically with as the initial estimated channel coefficient, i.e.,

[0135] In the process of finding the optimal sample parameter , the channel coefficient is defined to be static. Once the maximum iteration number J of finding the optimal sample parameter is met, the updated re-estimated support set is:

[0136]

[0137] Then, the corresponding channel coefficient is updated as:

[0138]

[0139] When the maximum iteration number Z is met, will gradually approach the true BS-side AoA. The re-estimated BS-side AoA is added at each iteration, gradually filling the sparse representation matrix

[0140] Since the scatterers on the BS side are sparse, the sparse representation dictionary realizes a substantial dimension reduction in the column space, thereby realizing the dimension reduction of the cascaded channel. Therefore, formula (11) can be rewritten as:

[0141]

[0142] wherein, is the compressed cascaded channel matrix.

[0143] In the case of the reduction of the cascaded channel matrix h, the proposed algorithm can effectively suppress the influence of off-network errors, and help to reduce the pilot overhead and computational complexity of the second-stage block structure SBL algorithm.

[0144] Step 3: solve formula (21) by using the EM algorithm to calculate the compressed cascaded channel matrix

[0145] On the one hand, in view of the block structure sparsity of the channel, the embodiment introduces a mode coupling prior model. On the other hand, in view of the high computational complexity problem caused by matrix inversion in the traditional SBL algorithm, the embodiment introduces a covariance-free method. Specifically as follows:

[0146] To obtain the characteristics of the transmitted symbol and its neighboring elements, a hyperparameter γ = {γ} is introduced. t-1 ,γ t ,γ t+1}, and will The model is as follows:

[0147]

[0148] in, γ t yes The accuracy of γ. To facilitate reasoning about γ, the elements of γ are usually modeled as independent gamma distributions, i.e.:

[0149]

[0150] Here, the parameters {a, b} take very small positive numbers. 0≤β≤1, where β is the coefficient. The pattern correlation parameter between γ and adjacent coefficients. t As the range increases (→+∞), the corresponding transmitted signal and its neighboring elements are also considered. Obviously, when β>0, the sparsity of each coefficient is controlled not only by its own hyperparameter but also by its nearest neighbor hyperparameters. If β=0, formula (22) degenerates into the traditional SBL model, and the likelihood function of the received signal is:

[0151]

[0152] Where 1 / β=σ -2 The noise accuracy is expressed as p(β) = Gamma(β|c,d) prior, controlled by parameters {c,d}. According to Bayes' theorem, and combining formulas (22) and (24), The posterior parameter distribution is given by the following formula:

[0153]

[0154] Wherein, the first posterior parameter μ is the compressed concatenated channel matrix. The mean of the random variable, and the second posterior parameter Σ is the compressed concatenated channel matrix. The covariance of a random variable.

[0155] Based on the above hierarchical model, the E-step of the EM algorithm uses formulas (26) and (27) to calculate μ and Σ respectively, in order to update posterior parameters. However, due to the high dimensionality of the cascaded channel G, the inverse computation process is very complex and tedious for each iteration when updating Σ. To reduce the computational complexity, this embodiment uses the solution of a linear system to replace the matrix inversion. A mode-coupled hierarchical Gaussian prior is introduced to capture the block-structured sparsity in the high-dimensional RIS cascaded channel. According to the need of M steps, the diagonal vector of Σ is first obtained using the diagonal estimation rule as follows:

[0156] Let be a square matrix, let p K be a random probing vector, where each entry is independently distributed such that The Rademacher distribution is used to characterize the random probing vector p k , each p k,t takes {-1, +1} with equal probability, and for p k , let d l = Mp k , each d t is an unbiased estimator of M t,t . Therefore, the diagonal vector of Σ can be expressed as: for all t = 1, 2,..., U,

[0157]

[0158] where each s t is an unbiased estimator of M t,t . Therefore, the diagonal vector of Σ can be expressed as:

[0159]

[0160] To avoid covariance inversion, this embodiment converts the formulas (26) and (29) into the form of solving the solution of a linear system, which are re-expressed as:

[0161]

[0162] The solution of the linear system AX = B is solved in parallel using the preconditioned conjugate gradient (PCG) algorithm, where and are as follows:

[0163]

[0164] However, for high-dimensional problems, more computational steps are required due to the large condition number of the matrix. This embodiment can accelerate the convergence by using a diagonal preconditioning matrix M = diag{Σ -1}, and the solution of the linear system obtained is divided into two parts, represented as μ = X(:, 1) and Thus, Σ t The calculation formula is:

[0165]

[0166] Using the above results, the M-step of the EM algorithm estimates the hyperparameters γ and β by maximizing the log marginal likelihood distribution, resulting in the following updates:

[0167]

[0168] where The time complexity of PCG is reduced from to where S is the number of PCG steps, K is the number of probing vectors, is the time of applying .

[0169] According to another aspect of the embodiments of the present application, a two-stage cascaded channel estimation device is provided, as shown in Figure 4 , the device comprises:

[0170] The compressed sensing module 101 is configured to perform compressed sensing on the constructed initial system model to obtain a communication system model, the initial system model comprising a two-stage cascaded channel of a user-to-reconfigurable intelligent surface side channel and a reconfigurable intelligent surface side-to-base station side channel.

[0171] The first calculation module 102 is configured to calculate a grid index value based on a base station measurement matrix and a sparse measurement matrix included in the initial system model.

[0172] The interval determination module 103 is configured to determine an index interval to which an angle of arrival belongs based on the grid index value, the angle of arrival being an angle of arrival of a signal received by the base station side.

[0173] The update module 104 is configured to update the sparse measurement matrix based on the index interval to obtain a sparse representation matrix.

[0174] The second calculation module 105 is configured to obtain a compressed cascaded channel matrix based on the sparse representation matrix and a cascaded channel matrix in the communication system model.

[0175] In one or more embodiments, the compressed sensing module 101 is configured to:

[0176] update the cascaded channel in the initial system model based on a first dictionary matrix and a second dictionary matrix of discrete grid points generated based on the initial system model, the first dictionary matrix being a uniform discrete array of angles of arrival at the base station side, and the second dictionary matrix being a uniform discrete array of angles of arrival at the reconfigurable intelligent surface side.

[0177] Rewrite the initial system model based on the updated cascade channel to obtain the communication system model.

[0178] In one or more embodiments, the interval determination module 103 is configured to:

[0179] Based on the grid index value, obtain a right adjacent index value and a left adjacent index value of the grid index value;

[0180] Calculate a first channel coefficient corresponding to the left adjacent index value and a second channel coefficient corresponding to the right adjacent index value;

[0181] In the case where the first channel coefficient is greater than or equal to the second channel coefficient, determine that the index interval to which the angle of arrival belongs is the interval composed of the left adjacent index value and the grid index value;

[0182] In the case where the first channel coefficient is less than the second channel coefficient, determine that the index interval to which the angle of arrival belongs is the interval composed of the grid index value and the right adjacent index value.

[0183] In one or more embodiments, the update module 104 is configured to:

[0184] Select a sample parameter in the index interval, and based on the sample parameter, obtain a support set of the angle of arrival at the base station side;

[0185] Based on the support set and a first dictionary matrix of the discrete grid points generated based on the initial system model, calculate an array vector;

[0186] Based on the array vector, update a sparse measurement matrix in the communication system model to obtain a sparse representation matrix.

[0187] In one or more embodiments, the second calculation module 105 is configured to:

[0188] Based on the sparse representation matrix, reduce the dimension of a cascade channel matrix in the communication system model to obtain a mode formula including the compressed cascade channel matrix;

[0189] Based on an expectation maximization algorithm and the mode formula, calculate the compressed cascade channel matrix.

[0190] In one or more embodiments, the second calculation module 105 is further configured to:

[0191] Based on an expectation maximization algorithm and the mode formula, model the compressed cascade channel matrix to obtain a modeling formula of the compressed cascade channel matrix;

[0192] obtain a calculation formula of a first posterior parameter based on the modeling formula, the first posterior parameter being a mean value of a random variable of the compressed concatenated channel matrix;

[0193] transform the calculation formula of the first posterior parameter to move a second posterior parameter in the calculation formula to the other side of an equal sign, and obtain a calculation formula of a diagonal vector of the second posterior parameter according to a diagonal estimation criterion, transform the calculation formula of the diagonal vector to move the second posterior parameter to the other side of the equal sign in the calculation formula of the diagonal vector, the second posterior parameter being a covariance of the random variable of the compressed concatenated channel matrix;

[0194] calculate the first posterior parameter based on the expectation maximization algorithm, the transformed calculation formula of the first posterior parameter, and the transformed calculation formula of the diagonal vector;

[0195] obtain the compressed concatenated channel matrix based on the first posterior parameter and the mode formula.

[0196] The two-stage concatenated channel estimation device provided by the embodiments of the present application and the two-stage concatenated channel estimation method provided by the embodiments of the present application have the same beneficial effects as the method adopted, run or implemented by them.

[0197] The embodiments of the present application further provide a computer device for executing the two-stage concatenated channel estimation method described above. Please refer to Figure 5 which shows a schematic diagram of a computer device provided by some embodiments of the present application. As shown in Figure 5 The computer device 8 comprises a processor 800, a memory 801, a bus 802 and a communication interface 803, the processor 800, the communication interface 803 and the memory 801 are connected through the bus 802; the memory 801 stores a computer program which can run on the processor 800, and the processor 800 runs the computer program to execute the two-stage concatenated channel estimation method provided by any of the preceding embodiments of the present application.

[0198] The memory 801 can include a high-speed random access memory (RAM: Random Access Memory) and can also include a non-volatile memory such as at least one disk memory. The communication between the device network element and at least one other network element is realized through at least one communication interface 803 (which can be wired or wireless), and the Internet, a wide area network, a local network, a metropolitan area network, etc. can be used.

[0199] The bus 802 can be an ISA bus, a PCI bus, an EISA bus, or the like. The bus can be divided into an address bus, a data bus, a control bus, and the like. The memory 801 is configured to store programs, and the processor 800 executes the programs after receiving execution instructions. The two-stage cascaded channel estimation method disclosed in any of the embodiments of the present application can be applied to the processor 800 or implemented by the processor 800.

[0200] The processor 800 can be an integrated circuit chip with processing capability. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor 800 or an instruction in the form of software. The processor 800 described above can be a general processor, including a central processing unit (CPU), a network processor (NP), and the like; or can be 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, discrete hardware components. Each method, step, and logic block diagram disclosed in the embodiments of the present application can be implemented or executed. The general processor can be a microprocessor or any conventional processor. The steps of the method disclosed in combination with the embodiments of the present application can be directly embodied as a hardware code processor for execution, or a combination of hardware and software modules in the code processor for execution. The software module can be located in a random access memory, a flash memory, a read-only memory, a programmable read-only memory, an electrically erasable programmable memory, a register, or other mature storage media in the art. The storage medium is located in the memory 801, and the processor 800 reads the information in the memory 801 and combines the hardware to complete the steps of the above method.

[0201] The computer device provided by the embodiments of the present application and the two-stage cascaded channel estimation method provided by the embodiments of the present application have the same beneficial effects as the method adopted, run, or implemented by them.

[0202] The embodiments of the present application also provide a computer readable storage medium corresponding to the two-stage cascaded channel estimation method provided by the preceding embodiments. The computer readable storage medium is an optical disc, and a computer program (i.e., a computer program product) is stored on the optical disc. When the computer program is run by a processor, the two-stage cascaded channel estimation method provided by any of the preceding embodiments is executed.

[0203] It should be noted that examples of the computer-readable storage medium can also include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory, or other optical, magnetic storage media, and the like, which will not be listed one by one here.

[0204] The computer-readable storage medium provided by the above embodiments of the present application has the same beneficial effects as the method adopted, run or implemented by the application program stored therein, based on the same inventive concept as the two-stage cascaded channel estimation method provided by the embodiments of the present application.

[0205] The embodiments of the present application also provide a computer program product, please refer to Figure 6 The computer program product 600 carries a program code, that is, a computer program 601, and the instructions included in the computer program 601 can be used to execute the steps of the two-stage cascaded channel estimation method described in the above method embodiments. For details, please refer to the above method embodiments, which will not be repeated here.

[0206] The computer program product can be specifically implemented by hardware, software or a combination thereof. In one optional embodiment, the computer program product is specifically embodied as a computer storage medium, and in another optional embodiment, the computer program product is specifically embodied as a software product, such as a software development kit (Software Development Kit, SDK) and the like.

[0207] The basic principles of the present application are described above in combination with specific embodiments, but it should be noted that the advantages, advantages, effects and the like mentioned in the present application are only examples and not limitations, and these advantages, advantages, effects and the like cannot be considered as the must-have of each embodiment of the present application. In addition, the above-mentioned specific details are only for the purpose of example and for the purpose of understanding, and are not limited to the present application, and the above-mentioned details do not limit the present application to the must-use of the above-mentioned specific details.

[0208] The block diagrams of the devices, apparatuses, equipment, systems referred to in this application are merely illustrative examples and are not intended to require or imply that the connection, arrangement, configuration must be as shown in the block diagrams. These devices, apparatuses, equipment, systems can be connected, arranged, configured in any manner as will be appreciated by those skilled in the art. Words such as "include," "contain," "have," etc. are open-ended words that are to be interpreted to mean "including but not limited to" and are to be used interchangeably. The words "or" and "and" as used herein are to be interpreted as the word "and / or" and are to be used interchangeably unless the context clearly indicates otherwise. The word "such as" as used herein is to be interpreted as the phrase "such as but not limited to" and is to be used interchangeably.

[0209] Also, as used herein, the term "or" when used in a list of two or more items indicates the separate lists such that, for example, the phrase "at least one of A, B, or C" means A or B or C or AB or AC or BC or ABC (i.e., A and B and C). Also, the phrase "example of" as used herein means "example, but not the only example of."

[0210] It is also important to note that the systems and methods of the present application can be embodied in a variety of forms without departing from the spirit or essential characteristics thereof. Likewise, the applications is not limited to the specific aspects described herein, but it is also intended to cover any and all alternatives, modifications, equivalents, and / or alternatives of the applications included within the spirit and / or scope of the present applications.

[0211] Various changes, modifications and alterations in the teachings and techniques described herein can be made without departing from the teachings and techniques defined in the appended claims. Furthermore, the scope of the following claims is not limited to the specific aspects described herein. The current existing or later developed processes, machines, manufactures, compositions of matter, means, methods, or steps that perform substantially the same function or achieve substantially the same results as the corresponding aspects described herein are intended to fall within the scope of the following claims. Accordingly, the appended claims are intended to cover all aspects of the application within the scope and spirit of the claims.

[0212] The above description of the disclosed aspects is intended to enable any person skilled in the art to make or use or perform the application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0213] The foregoing description has been presented for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of the application to the forms disclosed herein. Although various example aspects and embodiments have been discussed above, those of ordinary skill in the art will appreciate a variety of modifications, alternatives, permutations, additions, and sub-combinations of the described aspects and embodiments.

Claims

1. A two-stage cascaded channel estimation method, characterized in that, The method comprises the following steps: Compressive sensing is performed on an initial system model to obtain a communication system model, wherein the initial system model comprises a two-stage cascaded channel from a user to a reconfigurable intelligent surface side and from the reconfigurable intelligent surface side to a base station side; Based on the base station measurement matrix and the sparse measurement matrix included in the initial system model, a grid index value is calculated; Based on the grid index value, an index interval to which an angle of arrival belongs is determined, wherein the angle of arrival is an angle of arrival of a signal received by the base station side; Based on the index interval, the sparse measurement matrix is updated to obtain a sparse representation matrix; Based on the sparse representation matrix and the cascaded channel matrix in the communication system model, a compressed cascaded channel matrix is obtained.

2. The two-stage cascaded channel estimation method of claim 1, wherein, Compressive sensing is performed on an initial system model to obtain a communication system model, comprising: Based on the first dictionary matrix and the second dictionary matrix of the discrete grid points generated by the initial system model, the cascaded channel in the initial system model is updated, wherein the first dictionary matrix is a uniform discrete array of the angle of arrival of the base station side, and the second dictionary matrix is a uniform discrete array of the angle of arrival of the reconfigurable intelligent surface side; Based on the updated cascaded channel, the initial system model is rewritten to obtain the communication system model.

3. The two-stage cascaded channel estimation method of claim 1, wherein, Based on the grid index value, an index interval to which an angle of arrival belongs is determined, comprising: Based on the grid index value, a right adjacent index value and a left adjacent index value of the grid index value are obtained; A first channel coefficient corresponding to the left adjacent index value and a second channel coefficient corresponding to the right adjacent index value are calculated; In the case that the first channel coefficient is greater than or equal to the second channel coefficient, it is determined that the index interval to which the angle of arrival belongs is an interval composed of the left adjacent index value and the grid index value; In the case that the first channel coefficient is less than the second channel coefficient, it is determined that the index interval to which the angle of arrival belongs is an interval composed of the grid index value and the right adjacent index value.

4. The two-stage cascaded channel estimation method of claim 1, wherein, Based on the index interval, the sparse measurement matrix in the communication system model is updated to obtain a sparse representation matrix, comprising: In the index interval, a sample parameter is selected, and based on the sample parameter, a support set of the angle of arrival of the base station side is obtained; Based on the support set and the first dictionary matrix of the discrete grid points generated by the initial system model, an array vector is calculated; Based on the array vector, the sparse measurement matrix in the communication system model is updated to obtain a sparse representation matrix.

5. The two-stage cascaded channel estimation method of claim 1, wherein, Based on the sparse representation matrix and the cascaded channel matrix in the communication system model, a compressed cascaded channel matrix is obtained, comprising: Based on the sparse representation matrix, the cascaded channel matrix in the communication system model is reduced in dimension to obtain a mode formula including the compressed cascaded channel matrix; Based on the expectation maximization algorithm and the mode formula, the compressed cascaded channel matrix is calculated.

6. The two-stage cascaded channel estimation method of claim 5, wherein, Based on the expectation maximization algorithm and the mode formula, the compressed cascaded channel matrix is calculated, comprising: modeling the compressed cascaded channel matrix based on the expectation maximization algorithm and the mode formula, to obtain a modeling formula of the compressed cascaded channel matrix; obtaining a calculation formula of a first posterior parameter based on the modeling formula, the first posterior parameter being a mean value of a random variable of the compressed cascaded channel matrix; transforming the calculation formula of the first posterior parameter, so that a second posterior parameter in the calculation formula moves to the other side of an equal sign; and obtaining a calculation formula of a diagonal vector of the second posterior parameter according to a diagonal estimation criterion, transforming the calculation formula of the diagonal vector, so that the second posterior parameter moves to the other side of the equal sign in the calculation formula of the diagonal vector, the second posterior parameter being a covariance of the random variable of the compressed cascaded channel matrix; calculating the first posterior parameter based on the expectation maximization algorithm, the transformed calculation formula of the first posterior parameter, and the transformed calculation formula of the diagonal vector; obtaining the compressed cascaded channel matrix based on the first posterior parameter and the mode formula.

7. A two-stage cascaded channel estimation apparatus characterized by comprising: Comprise: The compressed sensing module is used for performing compressed sensing on the constructed initial system model to obtain a communication system model, the initial system model including a two-stage cascaded channel of a user-to-reconfigurable intelligent surface side channel and a reconfigurable intelligent surface side-to-base station side channel. The first calculation module is used for calculating a grid index value based on a base station measurement matrix and a sparse measurement matrix included in the initial system model. The interval determination module is used for determining an index interval to which an angle of arrival belongs based on the grid index value, the angle of arrival being an angle of arrival of a signal received by the base station side. The update module is used for updating the sparse measurement matrix based on the index interval to obtain a sparse representation matrix. The second calculation module is used for obtaining a compressed cascaded channel matrix based on the sparse representation matrix and a cascaded channel matrix in the communication system model.

8. A computer device comprising a memory, a processor, and a computer program stored on the memory, wherein the computer program comprises instructions that, when executed by the processor, cause the processor to perform the method of any one of claims 1-7. The processor executes the computer program to implement the method of any one of claims 1-6.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the method of any one of claims 1-6.

10. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the method of any one of claims 1-6.

Images