Probe weight allocation method, apparatus, medium and computer program product

Abstract

The invention relates to the technical field of communication, and particularly provides a probe weight distribution method and device, a medium and a computer program product. The probe weight distribution method comprises the following steps: acquiring working condition parameters of a to-be-detected channel model, and constructing an initial objective function corresponding to the to-be-detected channel model; introducing a replica variable, and constructing a target function based on the replica variable and the initial target function; based on the objective function, respectively carrying out iterative optimization on the replica variable and the probe power weight vector; and under the condition that a preset stop criterion is met, obtaining an optimized probe power weight vector, and taking the optimized probe power weight vector as target probe weight distribution information of the to-be-detected channel model. According to the method, the original problem is decomposed and the alternative updating strategy is adopted, so that the problem of numerical instability caused by directly facing a huge Hessian matrix (Hessian Matrix) is avoided, the calculation efficiency is greatly improved, and the convergence of the algorithm is effectively ensured.

Description

Technical Field

[0001] The present disclosure relates to the field of communication technology, and in particular to a probe weight allocation method, device, medium and computer program product. Background Art

[0002] MIMO (Multiple-Input Multiple-Output) technology not only significantly improves channel capacity through the transmission and reception of multiple antennas, but also reduces the bit error rate and improves channel reliability, meeting the urgent needs of communication technology for high data rates and high reliability in recent years. Therefore, the demand for verification of MIMO terminal equipment performance is increasing, and MIMO OTA (Multiple-Input Multiple-Output Over-the-Air) technology, as a test method specifically for multi-antenna technology, has received widespread attention.

[0003] The 3D multi-probe chamber is an ideal environment for testing and measuring MIMO OTA performance. In different 3D multi-probe chamber configurations, the layout and structure of the probe can adopt a variety of models, such as the three-ring model, the double-ring model, and the fan-shaped model. The design goal of these models is to ensure the accuracy and efficiency of the measurement while maximizing the spatial angle coverage.

[0004] In the related art, solvers (such as CVX (Convex Optimization)) are usually used to determine the weight allocation of probes. However, in the context of massive MIMO, due to the significant increase in the number of probes and the large sample size of the test area, using CVX to calculate the probe weights will result in high resource consumption and low computational efficiency, which in turn leads to poor real-time performance of probe weight allocation. Summary of the invention

[0005] In view of this, exemplary embodiments of the present disclosure provide a probe weight allocation method, device, medium and computer program product to solve the problems existing in the related art.

[0006] In one aspect of an exemplary embodiment of the present disclosure, a method for allocating probe weights is provided, the method comprising:

[0007] Acquire the working condition parameters of the channel model to be detected, and construct an initial objective function corresponding to the channel model to be detected; the initial objective function includes a probe power weight vector;

[0008] Introducing a replication variable, and constructing an objective function based on the replication variable and the initial objective function;

[0009] Based on the objective function, iteratively optimizing the replication variables and the probe power weight vector respectively;

[0010] When a preset stop criterion is met, an optimized probe power weight vector is obtained, and the optimized probe power weight vector is used as target probe weight allocation information of the channel model to be detected.

[0011] Another aspect of the exemplary embodiments of the present disclosure provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the method described in the exemplary embodiments of the present disclosure.

[0012] According to another aspect of the exemplary embodiments of the present disclosure, a computer-readable storage medium is provided, on which a computer program / instruction is stored. When the computer program / instruction is executed by a processor, the method described in the exemplary embodiments of the present disclosure is implemented.

[0013] According to another aspect of the exemplary embodiments of the present disclosure, a computer program product is provided, including a computer program / instruction. When the computer program / instruction is executed by a processor, the method described in the exemplary embodiments of the present disclosure is implemented.

[0014] As described in detail below, according to a probe weight allocation method of an embodiment of the present disclosure, the operating parameters of the channel model to be detected are obtained, and the initial objective function corresponding to the channel model to be detected is constructed; a replication variable is introduced, and the objective function is constructed based on the replication variable and the initial objective function; based on the objective function, the replication variable and the probe power weight vector are iteratively optimized respectively; when the preset stopping criteria are met, the optimized probe power weight vector is obtained, and the optimized probe power weight vector is used as the target probe weight allocation information of the channel model to be detected. Therefore, the probe weight allocation method provided by the present disclosure avoids the numerical instability problem caused by directly facing the huge Hessian Matrix by decomposing the original problem and adopting an alternating update strategy, thereby greatly improving the computational efficiency and effectively ensuring the convergence of the algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The above and other purposes, features and advantages of the present disclosure will become more apparent by describing the embodiments of the present disclosure in more detail in conjunction with the accompanying drawings. The accompanying drawings are used to provide a further understanding of the embodiments of the present disclosure and constitute a part of the specification. Together with the embodiments of the present disclosure, they are used to explain the present disclosure and do not constitute a limitation of the present disclosure. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0016] Figure 1a A schematic diagram of a probe configuration of a three-ring model provided as an example in the present disclosure;

[0017] Figure 1b A schematic diagram of a probe configuration of a double-ring model provided as an example in the present disclosure;

[0018] Figure 1c A schematic diagram of a probe configuration of a fan-shaped model provided as an example in the present disclosure;

[0019] Figure 2 A schematic diagram of a probe weight allocation model provided as an example in the present disclosure;

[0020] Figure 3 A flowchart of a probe weight allocation method provided by an exemplary embodiment of the present disclosure;

[0021] Figure 4 A schematic diagram of spatial correlation error of different probe weighting algorithms in three probe configurations provided as an example in the present disclosure;

[0022] Figure 5 A schematic diagram showing a comparison of probe weighted time of CVX and DAU-ALM algorithms in three probe configurations provided as an example in the present disclosure;

[0023] Figure 6 A schematic diagram of comparing probe weighted time of PD-IPM and DAU-ALM algorithms in three probe configurations provided as an example in the present disclosure;

[0024] Figure 7 A schematic diagram of weight comparison of probe weighting using different methods under a fan-shaped model condition provided as an example of the present disclosure;

[0025] Figure 8 A schematic block diagram of functional modules of a probe weight distribution device provided as an example in the present disclosure;

[0026] Fig. 9 A structural block diagram of an electronic device provided as an example in the present disclosure;

[0027] Fig.10 A schematic diagram of a computer program product provided by an exemplary embodiment of the present disclosure. DETAILED DESCRIPTION

[0028] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0029] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0030] The term "including" and its variations used in this document are open inclusions, that is, "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one other embodiment"; the term "some embodiments" means "at least some embodiments". Relevant definitions of other terms will be given in the description below. It should be noted that the concepts of "first", "second", etc. mentioned in this disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0031] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0032] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0033] It is understandable that before using the technical solutions disclosed in the embodiments of the present disclosure, the types, scope of use, usage scenarios, etc. of the personal information involved in the present disclosure should be informed to the user and the user's authorization should be obtained in an appropriate manner in accordance with relevant laws and regulations.

[0034] For example, in response to receiving an active request from a user, a prompt message is sent to the user to clearly prompt the user that the operation requested to be performed will require obtaining and using the user's personal information. Thus, the user can autonomously choose whether to provide personal information to software or hardware such as an electronic device, application, server, or storage medium that performs the operation of the technical solution of the present disclosure according to the prompt message.

[0035] As an optional but non-limiting implementation, in response to receiving an active request from the user, the method of sending a prompt message to the user may be, for example, a pop-up window, in which the prompt message may be presented in text. In addition, the pop-up window may also carry a selection control for the user to choose "agree" or "disagree" to provide personal information to the electronic device. It is understandable that the above notification and the process of obtaining user authorization are only illustrative and do not constitute a limitation on the implementation of the present disclosure. Other methods that meet relevant laws and regulations may also be applied to the implementation of the present disclosure.

[0036] Before introducing the embodiments of the present disclosure, the following definitions are given for the terms involved in the embodiments of the present disclosure:

[0037] Channel model: Channel models are used to simulate the propagation behavior of electromagnetic waves or other signals in a specific environment. In the design and testing of communication systems, different channel models can reflect different propagation characteristics and interference effects.

[0038] 3D Multi-Probe Anechoic Chamber: A 3D multi-probe anechoic chamber is an environment for testing and measurement in which probes (or antennas) are configured at different positions and angles to collect data in all directions.

[0039] PD-IPM (Primal-Dual Interior Point Method): PD-IPM is an algorithmic framework for solving linear and nonlinear programming problems. It achieves efficient optimization by solving the primal and dual forms of the problem simultaneously and using interior point techniques to approximate the optimal solution.

[0040] MIMO technology not only significantly improves channel capacity through multi-antenna transmission and reception, but also reduces bit error rate and improves channel reliability, meeting the urgent demand for high data rate and high reliability in communication technology in recent years. Therefore, the demand for verification of MIMO terminal equipment performance is increasing, and MIMO OTA technology, as a test method specifically for multi-antenna technology, has received widespread attention.

[0041] The 3D multi-probe chamber is an ideal environment for testing and measuring MIMO OTA performance. In different 3D multi-probe chamber configurations, the layout and structure of the probe can adopt a variety of models, such as the three-ring model, the double-ring model, and the fan-shaped model. The design goal of these models is to ensure the accuracy and efficiency of the measurement while maximizing the spatial angle coverage.

[0042] In the related art, solvers (such as CVX) are usually used to determine the weight allocation of probes. However, in the context of massive MIMO, due to the significant increase in the number of probes and the large sample size of the test area, using CVX to calculate the probe weights will result in high resource consumption and low computational efficiency, which in turn leads to poor real-time performance of probe weight allocation.

[0043] Therefore, in order to solve the above problems, an exemplary embodiment of the present invention provides a probe weight allocation method. For various channel models, according to DAU-ALM (Decomposed and Alternating Updated Augmented Lagrangian Multiplier, decomposed alternating updated augmented Lagrangian method probe weight allocation method), by constructing an augmented Lagrangian function, decomposing the original problem and alternately updating parameters to iteratively optimize the probe weights, thereby determining the optimal probe combination weights.

[0044] A 3D multi-probe chamber is a facility used for radio frequency (RF) measurements, especially for testing antenna performance, such as directivity, gain, and radiation efficiency. In different 3D multi-probe chamber configurations, the layout and structure of the probes can adopt a variety of different models, such as the three-ring model, the double-ring model, and the fan-shaped model. These models are designed to maximize the coverage of spatial angles while ensuring measurement accuracy and efficiency.

[0045] For example, Figure 1a A schematic diagram of a probe configuration of a three-ring model provided as an example in the present disclosure; Figure 1b A schematic diagram of a probe configuration of a double-ring model provided as an example in the present disclosure; Figure 1c A schematic diagram of a probe configuration of a fan-shaped model provided as an example in the present disclosure.

[0046] For example, Figure 2 This is a schematic diagram of the structure of a probe weight allocation model provided by an exemplary embodiment of the present disclosure, such as Figure 2 As shown, the probe weight allocation model may include: constructing an augmented Lagrangian function submodule, introducing a replication variable submodule, a decentralized variable optimization strategy submodule, a dynamic penalty adjustment submodule and an output submodule.

[0047] Among them, the augmented Lagrangian function submodule introduces Lagrangian multipliers to transform a constrained optimization problem into an unconstrained problem. In addition, by adding a strong convex penalty term, the augmented Lagrangian function is regarded as a strong convex function of w, thereby enhancing the convexity of the function and improving the convergence.

[0048] The replication variable submodule is introduced to ensure that each variable appears at most once in any given constraint by adding local substitution variables of the probe power weight vector (the original global optimization variable). Local substitution variables Q and J of the global variable w are introduced to replace the original variable w in the objective function and equality and inequality constraints respectively. At the same time, new equality constraints w=Q and w=J need to be introduced to ensure consistency.

[0049] The decentralized variable optimization strategy submodule decomposes the original problem into three subproblems containing independent variables and optimizes them separately. The conjugate gradient method is used to optimize the unconstrained problem of w, the Lagrange multiplier method is used to optimize Q, and the truncation process is used to satisfy the non-negative constraint of J.

[0050] The dynamic penalty adjustment submodule takes the difference between w and Q and J in the current iteration as the original residual, and takes the difference between Q and J in the current iteration as the dual residual. The adaptive update strategy uses the ratio of the original residual to the dual residual as the stopping criterion to dynamically adjust the penalty parameter, control the trade-off between the original and dual problems, and ensure that the algorithm can balance the update speed and accuracy under different residual conditions.

[0051] The output submodule outputs the optimized probe weights.

[0052] For example, Figure 3 A flow chart of a probe weight allocation method provided by an exemplary embodiment of the present disclosure is shown in FIG. Figure 3 As shown, the following steps may be specifically included:

[0053] Step S310: Acquire the operating parameters of the channel model to be detected, and construct an initial objective function corresponding to the channel model to be detected.

[0054] Exemplarily, the operating parameters of the channel model to be detected may include: power, ASA (Angle of Arrival Spread), AOA (Angle of Arrival), ESA (Elevation Spread of Arrival) and EoA (Elevation of Arrival).

[0055] The initial objective function is constructed using the operating condition parameters, which may include:

[0056] First, the spatial correlation is measured. Assuming that the receiving antennas in the test area have isotropic characteristics, taking the single cluster target channel model as a reference, the theoretical spatial correlation of a pair of antennas (u, v) in the mth position pair in the test area can be expressed as:

[0057]

[0058] Where a represents the wave number, a=2π / λ; λ represents the wavelength; j represents the imaginary unit in Euler's formula; and Respectively represent the position vectors of sampling points u and v in the mth position pair; represents the unit direction vector of the signal; P(Ω) is the three-dimensional spherical power angle spectrum; (·) represents the dot product operator.

[0059] Then, the power weight of the OTA (Over-the-Air) probe is adjusted.

[0060] During the MIMO OTA configuration process, the power weight of the OTA probe needs to be precisely adjusted to minimize the deviation between the theoretical spatial correlation of the target continuous spherical power spectrum and the simulated spatial correlation of the simulated discrete spherical power spectrum. By adjusting the power distribution of the probe, the simulated power spectrum is made as close as possible to the ideal continuous spherical power spectrum, thus ensuring the accuracy and consistency of the measurement results.

[0061] Corresponding to formula (1), in a three-dimensional multi-probe darkroom composed of K probes, the simulated spatial correlation factor obtained for the mth position pair is:

[0062]

[0063] in, represents the power weight vector of K probe antennas; j represents the imaginary unit in Euler's formula; a represents the wave number, a=2π / λ; and Respectively represent the position vectors of sampling points u and v in the mth position pair; w k is the power weight of the kth probe; is the unit position vector of the kth probe antenna.

[0064] Then, set the optimization goal.

[0065] Further assume that there are M antenna pairs on the test body, so there are To achieve the best reconstruction effect, an optimization goal of minimizing the sum of reconstruction errors can be set. This goal can be expressed as an objective function:

[0066]

[0067] in, represents the spatial correlation of reconstruction; according to formula (3), we can get in, represents the power weight vector to be optimized; represents the target spatial correlation vector; e represents an all-one lateral quantity with a length of K, that is, e = [1,1,...,1]; ‖·‖2 represents the 2-norm operation; 0 = [0,0,...,0] T , with a length of K.

[0068] To simplify the solution process, the complex value problem in equation (4) is equivalently expressed as a real value problem, and the initial objective function is obtained as follows:

[0069]

[0070] Among them, F Re and ρ Re represent the real part of F and ρ respectively; F Im and ρ Im denote the imaginary parts of F and ρ respectively; e denotes an all-one transverse quantity with a length of K, i.e., e = [1,1,...,1]; ‖·‖2 denotes the 2-norm operation; Represents the power weight vector of K probe antennas.

[0071] In the present disclosure, the average correlation error in the test area is measured by the root mean square error (RMSE). The RMSE between the actual measurement results and the theoretical expectations is calculated to measure the average correlation error in the test area. The weighted average can reasonably reflect the overall error of the test area. The root mean square error can be expressed as:

[0072]

[0073] in, represents the spatial correlation of reconstruction; represents the target spatial correlation vector; M represents the total number of antenna pairs.

[0074] Step S320: introducing replication variables, and constructing an objective function based on the replication variables and the initial objective function.

[0075] Exemplarily, the replicated variables include a first replicated variable Q and a second replicated variable J. The replicated variables Q and J can be introduced as local substitute variables for the probe power weight vector w, thereby ensuring that each variable appears at most once in any given constraint.

[0076] Specifically, the replication variables Q and J are introduced to replace the original global optimization variable w in the objective function and equality and inequality constraints respectively. In addition, new equality constraints w=Q and w=J need to be introduced to ensure consistency. By introducing replication variables, the initial objective function is constructed into the following augmented Lagrangian function form, which makes it easy to decompose and solve. The objective function can be expressed as:

[0077]

[0078] Among them, Z1 and Z2 represent the introduced Lagrange multipliers, μ represents the introduced penalty parameter; Q and J represent replication variables respectively; represents the power weight vector of K probe antennas; F Re and ρ Re represent the real part of F and ρ respectively; F Im and ρIm represent the imaginary parts of F and ρ respectively; e represents an all-one transverse quantity with a length of K, that is, e = [1,1,...,1].

[0079] Step S330: Based on the objective function, iteratively optimize the replication variables and the probe power weight vector respectively.

[0080] Step S331: Solve the objective function by using the conjugate gradient method and iteratively optimize the probe power weight vector.

[0081] Specifically, w is optimized by fixing the first replicated variable Q and the second replicated variable J, and the problem is converted into an unconstrained optimization problem about w, and the conjugate gradient method is used to solve it. The objective function (the above formula (6)) is converted into a standard quadratic form You can get:

[0082]

[0083] Among them, Z1 and Z2 represent the introduced Lagrange multipliers, μ represents the introduced penalty parameter; Q and J represent replication variables respectively; represents the power weight vector of K probe antennas; F Re and ρ Re represent the real part of F and ρ respectively; F Im and ρ Im They represent the imaginary parts of F and ρ respectively; I represents the K×K identity matrix.

[0084] According to the quadratic form, we can get the symmetric positive definite matrix H and vector b, and then use the conjugate gradient method to solve the linear equation system Hw=b, iteratively update the solution vector w, and optimize w:

[0085] H=F Re T F Re +F Im T F Im +μI (8)

[0086]

[0087] Where H represents a symmetric positive definite matrix; b represents a vector; Z1 and Z2 represent the introduced Lagrange multipliers, μ represents the introduced penalty parameter; Q and J represent the replication variables; F Re and ρ Re represent the real part of F and ρ respectively; F Im and ρ Im They represent the imaginary parts of F and ρ respectively; I represents the K×K identity matrix.

[0088] The conjugate gradient method takes advantage of the mathematical conjugation property, that is, in each iteration, the newly generated search direction is conjugate (orthogonal) to all previous search directions, thus ensuring that each step is optimized in a new direction, effectively avoiding duplication of work and accelerating the convergence process. In addition, the conjugate gradient method automatically calculates the optimal step size to minimize the quadratic function of the residual modulus, thereby improving the automation and convergence speed of the solution process without manual intervention.

[0089] Specifically, the detailed steps of the conjugate gradient method may include:

[0090] First, an initial solution vector w_0 and an initial residual r_0 = b-Hw_0 are selected and used as the first search direction p_0. In each iteration, the solution vector w is updated according to the current residual r and the searched direction p.

[0091] Then, new search directions are generated, ensuring that they are conjugate with all previous search directions and mathematically orthogonal, which helps ensure independent optimization in each dimension.

[0092] Furthermore, the optimal step length of each step is automatically calculated using a quadratic function that minimizes the residual modulus, thereby ensuring that the optimal solution is approached quickly and accurately.

[0093] Furthermore, according to the convergence status of the iteration process, the iteration parameters, such as the search direction update strategy and step size calculation accuracy, are dynamically adjusted.

[0094] Finally, specific convergence conditions are set, such as when the residual is small enough or a predetermined number of iterations is reached, the iteration is automatically terminated and an approximate solution is output.

[0095] Step S332: Solve the objective function by Lagrange multiplier method and iteratively optimize the first replicated variable.

[0096] Specifically, the first replicated variable Q is optimized by fixing the probe power weight vector w and the second replicated variable J. Based on the above formula (6), ignoring the constant term, we can get:

[0097]

[0098] ste·Q=1 (10)

[0099] Among them, Z1 represents the introduced Lagrange multiplier, μ represents the introduced penalty parameter; Q represents the replication variable; w represents the probe power weight vector; e represents an all-one lateral quantity with a length of K, that is, e = [1,1,...,1].

[0100] Expand the inner product and norm square terms in formula (10), reorganize and simplify them according to the relevant terms containing Q, separate the constant terms that do not contain Q and remove them. Since these terms do not affect the optimization results, they can be ignored. By expanding and reorganizing the terms to simplify the objective function, we can get:

[0101]

[0102] ste·Q=1 (11)

[0103] Among them, Z1 represents the introduced Lagrange multiplier, μ represents the introduced penalty parameter; Q represents the replication variable; w represents the probe power weight vector; e represents an all-one lateral quantity with a length of K, that is, e = [1,1,...,1].

[0104] Then, we can solve the constrained minimization problem of formula (11). Due to the existence of equality constraints, we can directly use the Lagrange multiplier method to find the optimal solution Q:

[0105]

[0106] Among them, Z1 represents the introduced Lagrange multiplier, μ represents the introduced penalty parameter; Q represents the replication variable; w represents the probe power weight vector; e represents an all-one transverse quantity with a length of K, that is, e = [1,1,...,1], e T represents the transpose of e; K represents the number of probes.

[0107] Step S333: Solve the objective function by truncating the non-negative conditions and perform iterative optimization on the second replicated variable.

[0108] Specifically, the second replicated variable J is optimized by fixing the probe power weight vector w and the first replicated variable Q. Based on the above formula (6), ignoring the constant term, we can get:

[0109]

[0110] stJ≥0 (13)

[0111] Among them, Z2 represents the introduced Lagrange multiplier, μ represents the introduced penalty parameter; J represents the replication variable; w represents the probe power weight vector; 0 = [0, 0, ..., 0] T , with a length of K.

[0112] In order to solve J in perfect square form, we can achieve this by changing the constant based on formula (13). We can first use a constant factor Multiply by formula (13) and add Since these are all constant differences, they will not affect the solution to the optimization problem about J. Then, we reorganize and simplify according to different J-related terms, separate the constant terms that do not contain J and remove them, and simplify the objective function by expanding and recombining the terms, and we get:

[0113]

[0114] stJ≥0 (14)

[0115] Among them, Z2 represents the introduced Lagrange multiplier, μ represents the introduced penalty parameter; J represents the replication variable; w represents the probe power weight vector; 0 = [0, 0, ..., 0] T , with a length of K.

[0116] The above formula (14) is a least squares problem with non-negative constraints. The objective function is a convex function, and the non-negative constraints are linear. Therefore, the entire problem is a convex optimization problem with a unique global minimum. For the unconstrained least squares problem, the optimal solution is However, due to the non-negative constraint J ≥ 0, we need to Truncate to satisfy the non-negative condition. Specifically, for For each element in , if it is greater than or equal to 0, it remains unchanged; if it is less than 0, it is truncated to 0. The final result of J is shown in formula (15):

[0117]

[0118] Among them, Z2 represents the introduced Lagrange multiplier, μ represents the introduced penalty parameter; J represents the replication variable; w represents the probe power weight vector.

[0119] Based on this, w is optimized by the conjugate gradient method, Q is optimized by the Lagrange multiplier method, and J is optimized by truncating the non-negative conditions, so as to obtain the optimal solutions of w, Q and J respectively.

[0120] Step S340: When a preset stop criterion is met, an optimized probe power weight vector is obtained, and the optimized probe power weight vector is used as target probe weight allocation information of the channel model to be detected.

[0121] In an embodiment, in order to control the iterative process and terminate it in time, a preset stopping criterion based on the residual ratio can be designed: when the ratio of the original residual to the dual residual is lower than a preset threshold, it is considered that the current solution is close enough to the optimal solution and the iteration can be stopped, thereby avoiding the waste of computing resources caused by excessive iteration.

[0122] When the preset stopping criterion is met, the iteratively optimized probe power weight vector w is obtained, and the optimized probe power weight vector is used as the target probe weight allocation information of the channel model to be detected.

[0123] When solving optimization problems, especially when multiple variables or complex constraints are involved, it is crucial to effectively control the iterative process and ensure the accuracy and convergence speed of the solution. In step S340, a strategy of adaptively updating the penalty parameter and a preset stopping criterion are adopted to optimize the iterative solution process, thereby efficiently and stably approaching the optimal solution.

[0124] First, define the primal residual and the dual residual. The primal residual refers to the difference between the solution vector w and the variables Q and J in the current iteration (i.e., (wQ) and (wJ)), while the dual residual is the difference between Q and J (i.e., (QJ)). These two residuals reflect the state of the current solution and the degree of satisfaction of the constraints.

[0125] Exemplarily, the initialization parameters may be expressed as: μ>0; s>0; rho=1.1126; ε=1e-8.

[0126] Among them, μ represents the introduced penalty parameter; s represents the ratio of the original residual to the dual residual; rho represents the step size used to adaptively update the penalty parameter; ε represents the convergence tolerance; e represents an all-one horizontal quantity with a length of K, that is, e = [1,1,...,1].

[0127] For example, the original residual can be expressed as:

[0128] p=||wQ||2+||wJ||2 (16)

[0129] Where p represents the original residual; w represents the probe power weight vector; Q and J represent the replicated variables respectively.

[0130] For example, the dual residual can be expressed as:

[0131] d=μ*||QJ||2 (17)

[0132] Among them, μ represents the introduced penalty parameter; Q and J represent the replication variables respectively.

[0133] After the current iteration is finished, if the preset stopping criterion is not met, the original residual, the dual residual, and the ratio between the original residual and the dual residual of the current iteration are obtained;

[0134] Based on the primal residual, the dual residual and the ratio, the penalty parameter in the objective function is updated.

[0135] Exemplarily, in the process of adaptively updating the penalty parameter, if p>s×d, then μ=μ×rho; if d>s×p, then μ=μ / rho.

[0136] By amplifying the role of the penalty parameter, the enforcement of the constraints can be strengthened, thereby better guiding the iterative direction of the solution. The adaptive update strategy adjusts the penalty parameter according to the ratio of the original residual to the dual residual. This dynamic adjustment mechanism allows the algorithm to flexibly adjust the penalty intensity according to the specific needs of the current solution state, effectively balancing the relationship between the original problem and the dual problem.

[0137] Based on this, the two constraints of the probe weight are satisfied by optimizing the sub-problems related to Q and J. During the optimization process, the unconstrained probe weight variable w optimized by the conjugate gradient method gradually approaches Q and J during the iteration process to ensure that the constraints of the probe weight are satisfied until the optimal probe weight solution vector w is output. In addition, the preset stopping criterion is combined with the adaptive penalty parameter update, so that the algorithm can ensure the convergence speed while also ensuring that the algorithm can balance the update speed and accuracy under different residual conditions.

[0138] Based on the above embodiment, this embodiment also provides a comparison between the CVX solver, the PD-IPM algorithm and the DAU-ALM algorithm proposed in the above embodiment under three different probe configurations and two different channel model conditions. In order to quantify the operating efficiency of each algorithm, the running time of the three probe weighted algorithms is compared in seconds.

[0139] For example, Figure 4 The figure is a schematic diagram of the spatial correlation error of different probe weighting algorithms in three probe configurations provided as an example in the present disclosure. The compared algorithms include CVX, PD-IPM and DAU-ALM. The ordinate represents the root mean square error (RMSE) which can be calculated by the above formula (5). Figure 4 As shown in the figure, compared with CVX and PD-IPM algorithms, the spatial correlation errors of DAU-ALM are the same, indicating that the weight allocation accuracy of DAU-ALM is not inferior to that of the other two algorithms.

[0140] For example, Figure 5 The following is a schematic diagram of the probe weighted time comparison of the CVX and DAU-ALM algorithms in the three probe configurations provided by the present disclosure. Figure 5 As shown in Figure 1, the comparison algorithms include CVX and DAU-ALM. Under two models and three probe configurations, the time comparison between DAU-ALM and CVX probe weighting is performed. Figure 5It can be seen that the total weighted duration of the probes of DAU-ALM is much shorter than that of CVX, which indicates that DAU-ALM can greatly improve the computational efficiency.

[0141] For example, Figure 6 The following is a schematic diagram of the probe weighted time comparison of the PD-IPM and DAU-ALM algorithms in the three probe configurations provided by the present disclosure. Figure 6 As shown in Figure 1, the comparison algorithms include PD-IPM and DAU-ALM. Under two models and three probe configurations, the time comparison between DAU-ALM and PD-IPM probe weighting is performed. Figure 5 It can be seen that the total weighted duration of the probes of DAU-ALM is smaller than that of PD-IPM, which indicates that DAU-ALM can improve the computational efficiency.

[0142] For example, Figure 7 The following is a schematic diagram of comparing the weights of probe weighting using different methods under the condition of a fan-shaped model provided by the present disclosure. Figure 7 As shown, the compared algorithms include CVX, PD-IPM and DAU-ALM. For each probe corresponding to the single cluster channel model and the CDL-B channel model respectively, the probe weighted values ​​determined by the three algorithms are the same. It can be seen that the weight allocation accuracy of DAU-ALM is not inferior to that of the other two algorithms.

[0143] One or more technical solutions provided in the exemplary embodiments of the present disclosure perform variable optimization by combining the augmented Lagrange multiplier method and the conjugate gradient method with decomposition and alternating updates. First, by decomposing the original problem and adopting an alternating update strategy, the numerical instability problem caused by the huge Hessian matrix is ​​avoided, while the convergence of the algorithm is effectively guaranteed. Secondly, the linear approximation part is introduced and the coefficients of the linear terms are dynamically adjusted, which significantly improves the numerical stability. In addition, by adaptively adjusting the penalty parameter and dynamically adjusting the step size according to the original residual and the dual residual, the convergence speed is accelerated.

[0144] Therefore, the probe weight allocation method provided in the exemplary embodiment of the present disclosure accelerates the algorithm convergence speed while considering the non-negative constraint of the probe weight, thereby improving the computational efficiency of the weight allocation method.

[0145] The above mainly introduces the solutions provided by the exemplary embodiments of the present disclosure. It is understandable that in order to implement the above functions, the electronic device includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should easily realize that, in combination with the units and algorithm steps of each example described in the embodiments disclosed herein, the present disclosure can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present disclosure.

[0146] The exemplary embodiments of the present disclosure can divide the functional units of the electronic device according to the above method examples. For example, each functional module can be divided according to each function, or two or more functions can be integrated into one processing module. The above integrated modules can be implemented in the form of hardware or in the form of software functional modules. It should be noted that the division of modules in the exemplary embodiments of the present disclosure is schematic and is only a logical function division. There may be other division methods in actual implementation.

[0147] In the case of dividing each functional module according to each function, an exemplary embodiment of the present disclosure provides a probe weight allocation device, which may be a server or a chip applied to a server. Figure 8 This is a schematic block diagram of the functional modules of a probe weight distribution device provided by an exemplary embodiment of the present disclosure. Figure 8 As shown, the probe weight distribution device 800 includes:

[0148] The data acquisition module 810 is used to acquire the operating parameters of the channel model to be detected and construct an initial objective function corresponding to the channel model to be detected; the initial objective function includes a probe power weight vector;

[0149] A data processing module 820 is used to introduce a replication variable and construct an objective function based on the replication variable and the initial objective function;

[0150] The data processing module 820 is further used to iteratively optimize the replication variables and the probe power weight vector respectively based on the objective function;

[0151] The data processing module 820 is further configured to obtain an optimized probe power weight vector when a preset stop criterion is met, and use the optimized probe power weight vector as target probe weight allocation information of the channel model to be detected.

[0152] In another embodiment provided by the present disclosure, the replicated variables include a first replicated variable and a second replicated variable, and the data processing module 820 is further used to solve the objective function by a conjugate gradient method, and iteratively optimize the probe power weight vector; solve the objective function by a Lagrange multiplier method, and iteratively optimize the first replicated variable; solve the objective function by truncating non-negative conditions, and iteratively optimize the second replicated variable.

[0153] In another embodiment provided by the present disclosure, the data processing module 820 is further used to convert the objective function into a linear equation by fixing the first replicated variable and the second replicated variable; solve the linear equation using a conjugate gradient method, and iteratively update the probe power weight vector;

[0154] The linear equation is:

[0155] Hw=b

[0156] in:

[0157] H=F Re T F Re +F Im T F Im +μI

[0158]

[0159] H represents a symmetric positive definite matrix; b represents a vector; Z1 and Z2 represent the introduced Lagrange multipliers, μ represents the introduced penalty parameter; Q represents the first replicated variable; J represents the second replicated variable; F Re and ρ Re represent the real part of F and ρ respectively; F Im and ρ Im They represent the imaginary parts of F and ρ respectively; I represents the K×K unit matrix; w represents the probe power weight vector.

[0160] In another embodiment provided by the present disclosure, the data processing module 820 is further used to transform the objective function by fixing the second replica variable and the probe power weight vector; solve the transformed objective function using the Lagrange multiplier method, and iteratively update the first replica variable;

[0161] The first replication variable is:

[0162]

[0163] Among them, Z1 represents the introduced Lagrange multiplier, μ represents the introduced penalty parameter; Q represents the first replicated variable; w represents the probe power weight vector; e represents an all-one transverse quantity with a length of K, that is, e = [1,1,...,1], e T represents the transpose of e; K represents the number of probes.

[0164] In another embodiment provided by the present disclosure, the data processing module 820 is further used to convert the objective function by fixing the first replication variable and the probe power weight vector; Truncation processing, iteratively updating the second copy variable;

[0165] The second replication variable is:

[0166]

[0167] Among them, Z2 represents the introduced Lagrange multiplier, μ represents the introduced penalty parameter; J represents the second replication variable; w represents the probe power weight vector.

[0168] In another embodiment provided by the present disclosure, the data processing module 820 is also used to obtain the original residual, dual residual and their ratio of the current iteration when a preset stopping criterion is not met; the original residual is the difference between the probe power weight vector and the replicated variable, and the dual residual is the difference between the first replicated variable and the second replicated variable; based on the original residual, the dual residual and their ratio, the penalty parameter in the objective function is updated.

[0169] In another embodiment provided by the present disclosure, the data processing module 820 further includes: the objective function is:

[0170]

[0171] stw=Q,w=J,e·Q=1,J≥0

[0172] Among them, Z1 and Z2 represent the introduced Lagrange multipliers, μ represents the introduced penalty parameter; Q and J represent replication variables respectively; represents the power weight vector of K probe antennas; F Re and ρ Re represent the real part of F and ρ respectively; F Im and ρ Im represent the imaginary parts of F and ρ respectively; e represents an all-one transverse quantity with a length of K, that is, e = [1,1,...,1].

[0173] The exemplary embodiment of the present disclosure also provides an electronic device, comprising: at least one processor; and a memory connected to the at least one processor in communication. The memory stores a computer program that can be executed by the at least one processor, and the computer program is used to cause the electronic device to perform the method according to the embodiment of the present disclosure when executed by the at least one processor.

[0174] The exemplary embodiments of the present disclosure also provide a non-transitory computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor of a computer, is used to cause the computer to perform the method according to the embodiments of the present disclosure.

[0175] Fig. 9 The block diagram of the electronic device provided as an example of the present disclosure, the block diagram of the electronic device 900 that can be used as the server or client of the present disclosure will now be described, which is an example of a hardware device that can be applied to various aspects of the present disclosure. The electronic device is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices 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 disclosure described and / or required herein.

[0176] like Fig. 9 As shown, the electronic device 900 includes a computing unit 901, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 902 or a computer program loaded from a storage unit 908 into a random access memory (RAM) 903. In the RAM 903, various programs and data required for the operation of the electronic device 900 can also be stored. The computing unit 901, the ROM 902, and the RAM 903 are connected to each other via a bus 904. An input / output (I / O) interface 905 is also connected to the bus 904.

[0177] A plurality of components in the electronic device 900 are connected to the I / O interface 905, including: an input unit 906, an output unit 907, a storage unit 908, and a communication unit 909. The input unit 906 may be any type of device capable of inputting information to the electronic device 900, and the input unit 906 may receive input digital or character information, and generate key signal inputs related to user settings and / or function control of the electronic device. The output unit 907 may be any type of device capable of presenting information, and may include, but is not limited to, a display, a speaker, a video / audio output terminal, a vibrator, and / or a printer. The storage unit 908 may include, but is not limited to, a disk, an optical disk. The communication unit 909 allows the electronic device 900 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks, and may include, but is not limited to, a modem, a network card, an infrared communication device, a wireless communication transceiver, and / or a chipset, such as a Bluetooth™ device, a WiFi device, a WiMax device, a cellular communication device, and / or the like.

[0178] The computing unit 901 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 901 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 901 performs the various methods and processes described above. Each of the methods described above may be implemented as a computer software program, which is tangibly included in a machine-readable medium, such as a storage unit 908. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 900 via the ROM 902 and / or the communication unit 909.

[0179] Fig.10 This is a schematic diagram of an exemplary computer program product provided by the present disclosure. The exemplary embodiment of the present disclosure also provides a computer program product 1000, including a computer program 1001, wherein the computer program 1001 is used to enable the computer to perform a method according to an embodiment of the present disclosure when executed by a processor of the computer.

[0180] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code 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.

[0181] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. 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.

[0182] As used in this disclosure, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, apparatus, and / or device (e.g., disk, optical disk, memory, programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0183] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer 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 computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0184] The systems and techniques described herein may be implemented in a computing system that includes back-end 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 front-end 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 back-end components, middleware components, or front-end 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), and the Internet.

[0185] A computer system may include clients and servers. Clients and servers are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship to each other.

[0186] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer programs or instructions. When the computer program or instruction is loaded and executed on a computer, the process or function described in the embodiment of the present disclosure is executed in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, a terminal, a user device or other programmable device. The computer program or instruction may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer program or instruction may be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server, data center, etc. that integrates one or more available media. The available medium may be a magnetic medium, for example, a floppy disk, a hard disk, a tape; it may also be an optical medium, for example, a digital video disc (DVD); it may also be a semiconductor medium, for example, a solid state drive (SSD).

[0187] Although the present disclosure has been described in conjunction with specific features and embodiments thereof, it is apparent that various modifications and combinations may be made thereto without departing from the spirit and scope of the present disclosure. Accordingly, this specification and the drawings are merely exemplary illustrations of the present disclosure as defined by the appended claims, and are deemed to have covered any and all modifications, variations, combinations or equivalents within the scope of the present disclosure. Obviously, those skilled in the art may make various modifications and variations to the present disclosure without departing from the spirit and scope of the present disclosure. Thus, if these modifications and variations of the present disclosure fall within the scope of the claims of the present disclosure and their equivalents, the present disclosure is also intended to include these modifications and variations.

Claims

1. A probe weight allocation method, characterized in that: The method comprises: Acquire the working condition parameters of the channel model to be detected, and construct an initial objective function corresponding to the channel model to be detected; the initial objective function includes a probe power weight vector; Introducing a replication variable, and constructing an objective function based on the replication variable and the initial objective function; Based on the objective function, iteratively optimizing the replication variables and the probe power weight vector respectively; When a preset stop criterion is met, an optimized probe power weight vector is obtained, and the optimized probe power weight vector is used as target probe weight allocation information of the channel model to be detected.

2. The method according to claim 1, characterized in that: The replicated variables include a first replicated variable and a second replicated variable; and the iterative optimization of the replicated variables and the probe power weight vector based on the objective function includes: Solving the objective function by the conjugate gradient method, and iteratively optimizing the probe power weight vector; Solving the objective function by Lagrange multiplier method, and iteratively optimizing the first replicated variable; The objective function is solved by truncating non-negative conditions, and the second replicated variable is iteratively optimized.

3. The method according to claim 2, characterized in that Solving the objective function by the conjugate gradient method and iteratively optimizing the probe power weight vector comprises: By fixing the first replicated variable and the second replicated variable, converting the objective function into a linear equation; Solving the linear equation using the conjugate gradient method, and iteratively updating the probe power weight vector; The linear equation is: Hw=b in: H=F Re T F Re +F Im T F Im +μI H represents a symmetric positive definite matrix; b represents a vector; Z1 and Z2 represent the introduced Lagrange multipliers, μ represents the introduced penalty parameter; Q represents the first replicated variable; J represents the second replicated variable; F Re and ρ Re represent the real part of F and ρ respectively; F Im and ρ Im They represent the imaginary parts of F and ρ respectively; I represents the K×K unit matrix; w represents the probe power weight vector.

4. The method according to claim 2, characterized in that: The step of solving the objective function by using the Lagrange multiplier method and iteratively optimizing the first replicated variable includes: Transforming the objective function by fixing the second replicated variable and the probe power weight vector; Solve the converted objective function using the Lagrange multiplier method, and iteratively update the first replica variable; The first replication variable is: Among them, Z1 represents the introduced Lagrange multiplier, μ represents the introduced penalty parameter; Q represents the first replicated variable; w represents the probe power weight vector; e represents an all-one transverse quantity with a length of K, that is, e = [1,1,...,1], e T represents the transpose of e; K represents the number of probes.

5. The method according to claim 2, characterized in that: The step of solving the objective function by truncating the non-negative condition and iteratively optimizing the second replicated variable includes: Transforming the objective function by fixing the first replicated variable and the probe power weight vector; The converted objective function is Truncation processing, iteratively updating the second copy variable; The second replication variable is: Among them, Z2 represents the introduced Lagrange multiplier, μ represents the introduced penalty parameter; J represents the second replication variable; w represents the probe power weight vector.

6. The method according to claim 1, characterized in that The method further comprises: When the preset stopping criteria are not met, the original residual, the dual residual and the ratio thereof of the current iteration are obtained; the original residual is the difference between the probe power weight vector and the replicated variable, and the dual residual is the difference between the first replicated variable and the second replicated variable; Based on the primal residual, the dual residual and their ratio, a penalty parameter in the objective function is updated.

7. The method according to claim 1, characterized in that The objective function is: stw=Q,w=J,e·Q=1,J≥0 Among them, Z1 and Z2 represent the introduced Lagrange multipliers, μ represents the introduced penalty parameter; Q and J represent replication variables respectively; represents the power weight vector of K probe antennas; F Re and ρ Re represent the real part of F and ρ respectively; F Im and ρ Im represent the imaginary parts of F and ρ respectively; e represents an all-one transverse quantity with a length of K, that is, e = [1,1,...,1].

8. A computer device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the method of claim 1.

9. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instructions are executed by a processor, the method of claim 1 is implemented.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the method of claim 1 is implemented.

Images