A Multi-Channel Cooperative Control Method Based on RIS Phased Array

Abstract

This invention discloses a multi-channel cooperative control method based on a RIS phased array, relating to the field of phased array technology. The method includes: Step S1: Arranging several independent channels in the target RIS phased array and calculating a first efficiency value representing the ideal beamforming efficiency; Step S2: Activating the independent channels into cooperative multi-channels and constructing a beam coupling model for the cooperative multi-channels; Step S3: Predicting the predicted beam parameters of the cooperative multi-channels using the beam coupling model; Step S4: Calculating a second efficiency value representing the predicted beamforming efficiency based on the predicted beam parameters, and comparing the second efficiency value with the first efficiency value to determine whether the calibration is effective. This invention, by constructing a beam coupling model after closed-loop calibration and simultaneously introducing a comparison and discrimination mechanism for beam generation efficiency values, performs real-time evaluation of the multi-channel calibration results. It has the advantages of reducing beam sidelobes, improving main lobe focusing, and significantly reducing the amplification effect of multi-channel phase errors.

Description

Technical Field

[0001] This invention relates to the field of phased array technology, and specifically to a multi-channel cooperative control method based on a RIS phased array. Background Technology

[0002] Existing channel signal calibration techniques for RIS phased arrays are mostly based on the assumption of channel independence, meaning that during single-channel calibration, other channels are assumed to be in a closed or ideally matched state. The calibration process typically involves injecting test signals channel by channel using an external measurement system (such as a vector network analyzer) and adjusting the amplitude and phase parameters of each element based on the feedback from the reflected signals. This method is applicable to sparse arrays or low-frequency scenarios, but in high-frequency millimeter-wave or terahertz dense array deployments of RIS phased arrays, near-field electromagnetic coupling effects can easily cause nonlinear fluctuations in the multi-channel coupling strength due to non-uniform changes in the element spacing during multi-channel collaborative control when the RIS phased array is in a bent state.

[0003] Under the traditional linear system assumption, the joint response of multiple channels is a vector superposition of the single-channel responses. However, experimental results show that in a 128-element RIS phased array, when 16 channels are activated simultaneously, the multi-channel phase error of the synthesized beam reaches 6.3 times the single-channel calibration error, resulting in an average beam gain decrease of 7.2 dB, which is significantly different from theoretical expectations. Current technologies for correcting beam gain loss due to multi-channel phase errors typically involve fine-tuning the amplitude and phase parameters of each individual channel, which can easily lead to neglecting the cross-influence of coupling interference under multi-channel coordination. Existing methods, after adjusting for single-channel phase errors, may actually increase the error or further decrease the beam gain during multi-channel coordination because the single-channel calibration process does not take effect on multi-channel coordination. This results in poor correction of the beam gain loss and consequently reduces the energy efficiency of dense RIS phased array deployments. Summary of the Invention

[0004] This invention provides a multi-channel cooperative control method based on RIS phased arrays, which solves the problem that in dense RIS phased array deployments, existing technologies for calibrating individual channel errors lack interference correction for multi-channel cooperation, resulting in a decrease in multi-channel cooperative beam gain.

[0005] This invention is achieved through the following technical solution:

[0006] A multi-channel cooperative control method based on RIS phased array, the method comprising:

[0007] Step S1: Divide the array elements of the target RIS phased array into several independent channels, preset ideal beam parameters for each independent channel, and calculate the first efficiency value representing the ideal beamforming efficiency based on the ideal beam parameters.

[0008] Step S2: Label the original reflection coefficient of each independent channel, perform closed-loop calibration on all original reflection coefficients, activate all independent channels after closed-loop calibration as collaborative multi-channel, and construct a beam coupling model for collaborative multi-channel.

[0009] Step S3: Use the beam coupling model to predict and generate the synthetic beam of the cooperative multi-channel array control, and label the generated data of the synthetic beam as the predicted beam parameters representing the tendency of the cooperative multi-channel beam results.

[0010] Step S4: Calculate a second efficiency value representing the predicted beamforming efficiency based on the predicted beam parameters, and compare the second efficiency value with the first efficiency value. If the first efficiency value is lower than the second efficiency value, the calibration is determined to be effective; if the first efficiency value is higher than the second efficiency value, the calibration is determined to be ineffective, and the process returns to step S2.

[0011] Furthermore, the closed-loop calibration of all original reflection coefficients includes: detecting and labeling the coupling strength of different independent channels, allocating calibration time slot lengths in a positive correlation manner according to the coupling strength between channels; constructing a TDM schedule table, and using the TDM schedule table to perform closed-loop parallel calibration of the original reflection coefficients in each independent channel according to the calibration time slot length.

[0012] Furthermore, the contents of the TDM scheduling table include: dividing the independent channel into several channel groups of equal number, measuring the electromagnetic coupling strength between each pair of channel groups and representing it as a coupling strength matrix, calculating the coupling weight value of each channel group through the coupling strength matrix, and allocating calibration time to each channel group using the coupling weight value.

[0013] Furthermore, the original reflection coefficients are calibrated in a closed-loop parallel manner using an iterative least squares method, the process of which includes:

[0014] Let k be the iteration number in the calibration process, set the calibration reflection coefficient Γ, and set the original reflection coefficient as the initial value of the iterative reflection coefficient; preset the target reflection coefficient Γ for each independent channel. t Given the control parameter θ, set the Jacobian matrix and represent it as J; set the error threshold for the iterative process;

[0015] The formula for calculating the calibration reflection coefficient Г in each iteration is expressed as: ,

[0016] Wherein, the θ k-1 The control parameter value representing the previous iteration number, θ k The control parameter value representing the current iteration sequence number; the Г k The calibration reflection coefficient representing the current iteration order, Γ k-1The J represents the calibration reflection coefficient of the previous iteration sequence; k-1 The Jacobian matrix representing the ordinal number of the previous iteration;

[0017] When calibrating the reflection coefficient Г k With the target reflection coefficient Г t When the difference is less than the error threshold, the iteration process stops and calibration is completed.

[0018] Furthermore, the correction error is defined as ε, and its calculation formula is as follows: , where ε k This represents the correction error for the current iteration sequence.

[0019] Let the least squares objective value of the control parameter be denoted as Δθ, and solve it using the standard least squares formula.

[0020] The formula for calculating the least squares objective value Δθ is as follows: ,

[0021] J H It is the conjugate transpose of the Jacobian matrix.

[0022] Furthermore, the formula for calculating the Jacobian matrix J is expressed as follows: J k This represents the value of the Jacobian matrix at the current iteration ordinal number.

[0023] Furthermore, the beam coupling model includes a Riemannian manifold algorithm and a beam generator; the beam coupling model's operation includes:

[0024] The original reflection coefficients after closed-loop calibration are constructed into complex reflection coefficients for cooperative multi-channel operation. The phase amplitude space is modeled as a complex Hilbert manifold. The complex reflection coefficients are mapped to the complex Hilbert manifold using a nonlinear mapping. A manifold loss function is constructed based on the complex reflection coefficients, and the Riemann gradient of the manifold loss function is calculated.

[0025] The control parameters of the beam generator are obtained. The Riemann gradient and control parameters are mapped to the tangent space of the complex Hilbert manifold for geodesic update analysis. The control parameters are updated using exponential mapping in the tangent space. The updated control parameters are input into the beam generator to output the synthesized beam. The generated data of the synthesized beam are collected and labeled as the predicted beam parameters.

[0026] Furthermore, an inverse distortion parameter is added to the updated control parameters and then input into the beam generator to output a synthesized beam.

[0027] Furthermore, assume that the ideal beam parameters and the predicted beam parameters are identical in content; the data content of the ideal beam parameters and the predicted beam parameters includes:

[0028] The main lobe gain value represents the radiation intensity in the main lobe direction corresponding to the peak value of the beam gain.

[0029] The main lobe pointing error value represents the degree of deviation of the actual pointing direction of the main lobe from the predetermined direction;

[0030] Sidelobe level value represents the ratio of radiated power in the sidelobe direction to the peak power of the main lobe;

[0031] The sidelobe pseudolobe value represents the parasitic high-gain lobe value caused by the phase quantization error of the independent channel.

[0032] Furthermore, the calculation process for the first efficiency value and the second efficiency value is set as follows:

[0033] After normalizing the data content of both the ideal beam parameters and the predicted beam parameters, a scoring function is generated. The scoring functions for the main lobe gain value, main lobe pointing error value, side lobe level value, and side lobe pseudolobe value are respectively labeled as the first function f1, the second function f2, the third function f3, and the fourth function f4, and corresponding weight values ​​m1, m2, m3, and m4 are set. Let the beam efficiency function be represented by η.

[0034] The beam efficiency function η is then expressed as: η = m1∙f1 + m2∙f2 + m3∙f3 + m4∙f4,

[0035] Among the four weight values, the first weight value m1 has the largest value; the calculation process of the first efficiency value and the second efficiency value is the same as the calculation formula of the beam efficiency function η.

[0036] Compared with existing technologies, this invention constructs a beam coupling model after closed-loop calibration to explicitly characterize the near-field electromagnetic coupling between channels and the nonlinear changes under curved structures. This overcomes the shortcomings of traditional methods that neglect the mutual interference of multiple channels. At the same time, it introduces a comparison and discrimination mechanism for beam generation efficiency values ​​to evaluate the multi-channel calibration results in real time. This avoids the accumulation of new errors that may be caused by single fine-tuning of the RIS phased array, and can reduce beam sidelobes and improve main lobe focusing. It has the advantages of improving the multi-channel cooperative beam gain of the RIS phased array and the beneficial effect of significantly reducing the multi-channel phase error amplification effect. Attached Figure Description

[0037] The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and form part of this application, do not constitute a limitation thereof. In the drawings:

[0038] Figure 1 This is a structural block diagram of the present invention. Detailed Implementation

[0039] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention. Example

[0040] like Figure 1 As shown, this embodiment is a multi-channel cooperative control method based on a RIS phased array, which includes:

[0041] Step S1: Divide the array elements of the target RIS phased array into several independent channels, preset ideal beam parameters for each independent channel, and calculate the first efficiency value representing the ideal beamforming efficiency based on the ideal beam parameters.

[0042] Step S2: Label the original reflection coefficient of each independent channel, perform closed-loop calibration on all original reflection coefficients, activate all independent channels after closed-loop calibration as collaborative multi-channel, and construct a beam coupling model for collaborative multi-channel.

[0043] Step S3: Use the beam coupling model to predict and generate the synthetic beam of the cooperative multi-channel array control, and label the generated data of the synthetic beam as the predicted beam parameters representing the tendency of the cooperative multi-channel beam results.

[0044] Step S4: Calculate a second efficiency value representing the predicted beamforming efficiency based on the predicted beam parameters, and compare the second efficiency value with the first efficiency value. If the first efficiency value is lower than the second efficiency value, the calibration is determined to be effective; if the first efficiency value is higher than the second efficiency value, the calibration is determined to be ineffective, and the process returns to step S2.

[0045] Dividing the array elements of the target RIS phased array into several independent channels means dividing the entire large-scale array into several small unit groups that can be driven, calibrated, and optimized independently at the physical or control level. In practice, common partitioning methods include partitioning according to physical topology, RF links, and beam segments, and usually one or more hybrid methods are used for array channel partitioning. The original reflection coefficient refers to the reflection amplitude and phase parameters of each independent channel before precise adjustment or collaborative calibration. The closed-loop calibration refers to the introduction of a measurement-feedback-adjustment-verification closed-loop feedback control mechanism to continuously optimize the original reflection coefficient until the system performance meets the preset target or convergence condition. Activating independent channels as collaborative multi-channel means that after completing the closed-loop amplitude and phase calibration of each independent channel, each channel is no longer tested or controlled individually, but all channels participate together in a complete beamforming task. In practical implementation, there are various conventional technical methods to activate independent channels into cooperative multi-channel systems. For example, unified addressing by the controller connects independent channels to the central controller via address mapping and drive circuitry; or a unified power supply network is used, with phase shifters converging the bus to a single main control chip to complete the conversion from single-channel to multi-channel operation. Using a beam coupling model to predict the synthesized beam means using the beam coupling model to predict the synthesized beam formed by multiple simultaneously activated channels under given array control parameters. This prediction evaluates whether the current cooperative multi-channel synthesized beam meets the requirement of reducing beam gain degradation in a RIS phased array.

[0046] The beam coupling model described is a computational simulation model constructed in this embodiment, used to evaluate the impact of coupling interference between channels in a cooperative multi-channel system on the generated composite beam. Both the first efficiency value and the second efficiency value are beamforming efficiency values. The first efficiency value represents the theoretical beamforming efficiency calculated based on ideal beam parameters; the second efficiency value represents the beamforming efficiency predicted by the beam coupling model for the target RIS phased array after this calibration process. When the first efficiency value is greater than the second efficiency value, considering the calibration configuration after coupling interference, the predicted beamforming performance is better than or exceeds the ideal state, indicating that the current closed-loop calibration combined with coupling modeling has achieved an optimization effect. In practical implementation, to avoid excessively long overall operating time for the target RIS phased array, the ideal beam parameters and the first efficiency value can be set to the lower limit of the ideal state, i.e., each parameter is set to relatively low values. This improves the pass rate of the predicted beam parameters and avoids excessively long calibration time while ensuring stable beam gain. Conversely, to guarantee the generation quality of the target RIS phased array, the ideal beam parameters and the first efficiency value can be set to a higher level of the ideal state, i.e., each parameter is set to relatively high values. This maximizes the guarantee of the predicted beam quality, significantly stabilizes the beam gain, and avoids beam gain degradation. This implementation ensures that the target RIS phased array always uses optimal or suboptimal amplitude and phase settings in high-frequency, dense, and structurally variable scenarios, reducing manual intervention and testing costs, and improving the efficiency of beam generation verification and calibration for array elements.

[0047] Furthermore, as a feasible implementation method, the closed-loop calibration of all original reflection coefficients includes: detecting and marking the coupling strength of different independent channels, allocating calibration time slot lengths in a positive correlation manner according to the coupling strength between channels; constructing a TDM schedule table, and using the TDM schedule table to perform closed-loop parallel calibration of the original reflection coefficients in each independent channel according to the calibration time slot length.

[0048] The coupling strength of different independent channels is detected and labeled. This typically involves measuring S-parameters and estimating them using EM simulation for each independent channel. The S-parameter amplitude is used as the index data for coupling strength. The field coupling coefficient of neighboring cells is calculated using EM for different channel excitations. The calibration time slot length is allocated in a positive correlation based on the coupling strength between channels; that is, the greater the coupling strength, the longer the calibration time slot allocated to that channel. This provides more measurement and adjustment time for highly coupled channels, reducing interference errors during calibration. TDM stands for Time Division Multiplexing, which divides time into several consecutive and non-overlapping time slots, allowing different channels or signals to use different time slots sequentially to avoid interference. The TDM scheduling table specifies when each channel receives time resources for calibration operations. It is typically a time series table, set in the form of "Channel 1 occupies time slots 1-3, Channel 2 occupies time slots 4-6, ...". The calibration time slot length is dynamically allocated based on the channel coupling strength; channels with higher coupling strength receive longer time slots to ensure more sufficient calibration time.

[0049] Furthermore, as a feasible implementation method, the contents of the TDM scheduling table include: dividing the independent channel into several channel groups of equal number, measuring the electromagnetic coupling strength between each pair of channel groups and representing it as a coupling strength matrix, calculating the coupling weight value of each channel group through the coupling strength matrix, and allocating calibration time to each channel group using the coupling weight value.

[0050] Electromagnetic coupling strength describes the degree of electromagnetic field interaction between two channel groups, reflecting the strength of interference or energy transfer caused by the signal from one channel group to the other in space. In practice, the coupling coefficient can be obtained by measuring the transmission parameters between the two channel groups using a vector network analyzer, and the coupling strength can be indirectly calculated using electromagnetic simulation with signal injection and feedback measurement results. The purpose of calculating the coupling weight value for each channel group is to first quantify the coupling influence of each channel group in the entire array, starting from the coupling relationship between groups, and then allocate the total calibration time to each channel group according to the proportion of coupling influence.

[0051] Example 2

[0052] In this embodiment, the iterative least squares method is used to perform closed-loop parallel calibration of the original reflection coefficients. The process includes:

[0053] Let k be the iteration number in the calibration process, set the calibration reflection coefficient Γ, and set the original reflection coefficient as the initial value of the iterative reflection coefficient; preset the target reflection coefficient Γ for each independent channel. t Given the control parameter θ, set the Jacobian matrix and represent it as J; set the error threshold for the iterative process;

[0054] The formula for calculating the calibration reflection coefficient Г in each iteration is expressed as: ,

[0055] Wherein, the θ k-1 The control parameter value representing the previous iteration number, θ k The control parameter value representing the current iteration sequence number; the Г k The calibration reflection coefficient representing the current iteration order, Γ k-1 The J represents the calibration reflection coefficient of the previous iteration sequence; k-1 The Jacobian matrix representing the ordinal number of the previous iteration;

[0056] When calibrating the reflection coefficient Г k With the target reflection coefficient Г t When the difference is less than the error threshold, the iteration process stops and calibration is completed.

[0057] The calibration reflection coefficient represents the reflection coefficient value of the original reflection coefficient in each iteration. The target reflection coefficient represents the reference target value for the reflection coefficient iteration, which can be set based on historical data and empirical rules. It includes, for example, the reflection coefficient amplitude representing the reflected energy intensity and the reflection coefficient phase determining the directionality of the reflected wave. The control parameters represent the physically adjustable physical inputs, representing the physical parameters of the corresponding independent channels in the target RIS phased array, used for initializing the control configuration for calibration or simulation modeling. The physical quantity represented by the control parameters can be one or more, depending on the control complexity of the reflection characteristics of each RIS unit or channel. For simple RIS units, such as those controlling only the reflection phase or having only one adjustable device, such as a PIN switch or a 1-bit capacitive load; for complex RIS units, such as dual-control units where amplitude and phase can be independently adjusted, the control parameters are in the form of multivariable functions. The Γ... k-1 This represents the current estimated calibration reflection coefficient obtained in the previous iteration of calibration based on the electromagnetic response under the current control parameters. The θ... k -θ k-1 This represents the adjustment amount of the control parameters for each independent channel in the actual system, such as a voltage increase of 0.2V or a phase modulation bit flip. The J... k-1 The Jacobian matrix represents the sensitivity of the reflection response to the control variable. Its value is the value of the previous iteration. Essentially, it is the gradient of the reflection coefficient caused by a small change in each control parameter.

[0058] Furthermore, as a feasible implementation method, the correction error is defined as ε, and its calculation formula is as follows: , where ε k This represents the correction error for the current iteration sequence.

[0059] Let the least squares objective value of the control parameter be denoted as Δθ. Using the standard least squares formula, the calculation formula for the least squares objective value Δθ is as follows: J H It is the conjugate transpose of the Jacobian matrix.

[0060] The correction error ε represents the error that the calibration system still needs to compensate for, and is used to drive the next step of control parameter updates. This is used to deduce the optimal change in the control variables, so that the correction error value in each iteration sequence is... It can reach a minimum. Therefore, the formula for calculating Δθ can be substituted into the formula for calculating the calibration reflection coefficient Г to obtain: This completes the update of the calculation formula for the calibration reflection coefficient Г.

[0061] More specifically, as a feasible implementation, the Jacobian matrix J is expressed as follows: J k This represents the value of the Jacobian matrix at the current iteration order. Its meaning is to express the sensitivity of the j-th control parameter to the i-th reflection channel at its current value. It is in the form of a partial derivative matrix, which is the mathematical basis of the system's local linear model, representing the effect of θ... k The values ​​are taken at each position to ensure that the Jacobian matrix is ​​consistent with the current state, which helps the iteration converge.

[0062] Example 3

[0063] In this embodiment, the beam coupling model includes a Riemannian manifold algorithm and a beam generator; the working process of the beam coupling model includes:

[0064] The original reflection coefficients after closed-loop calibration are constructed into complex reflection coefficients for cooperative multi-channel operation. The phase amplitude space is modeled as a complex Hilbert manifold. The complex reflection coefficients are mapped to the complex Hilbert manifold using a nonlinear mapping. A manifold loss function is constructed based on the complex reflection coefficients, and the Riemann gradient of the manifold loss function is calculated.

[0065] The control parameters of the beam generator are obtained. The Riemann gradient and control parameters are mapped to the tangent space of the complex Hilbert manifold for geodesic update analysis. The control parameters are updated using exponential mapping in the tangent space. The updated control parameters are input into the beam generator to output the synthesized beam. The generated data of the synthesized beam are collected and labeled as the predicted beam parameters.

[0066] In this process, an inverse distortion parameter is added to the updated control parameters and then input into the beam generator to output a synthesized beam.

[0067] Constructing a collaborative multi-channel complex reflection coefficient refers to unifying or merging the reflection coefficients of all independent channels into a unified multi-channel complex vector structure, serving as the foundational input for subsequent multi-channel coupling analysis and modeling. The RIS reconfigurable smart surface consists of numerous sub-units, each with its own reflection parameters, including phase and amplitude, typically represented as complex numbers. The reflection responses of each calibrated channel are concatenated into a complete system-level complex vector input for subsequent modeling of collaborative behavior. In conventional implementations, the original reflection coefficients in vector or matrix form are typically transmitted via a control bus to the coupling modeling or beamforming module, where they are encapsulated in the software as complex arrays in a high-level language for direct use in subsequent nonlinear mapping, manifold optimization, and other processes. The complex Hilbert manifold is a mathematical structure combining a differential manifold and a complex Hilbert space; in the RIS phased array calibration process of this embodiment, it represents the parameter space formed by the complex reflection coefficients of N units after satisfying complex analytic properties. Using a nonlinear mapping to map the complex reflection coefficients onto a complex Hilbert manifold means embedding the calibrated planar complex reflection coefficient vector into the manifold space of a complex Hilbert manifold through a nonlinear function, so that Riemann optimization can be performed on this manifold subsequently. The synthesized beam is usually given in the form of a two-dimensional or three-dimensional radiation map, or it can be a sampled point cloud or a polar coordinate data table. The data for generating the synthesized beam can be a set of scalars or a discrete vector of a beam pattern. The beam generator can specifically be a vector network analyzer or an antenna measurement system, which receives these control parameters and outputs a complete beam pattern through physical measurement or numerical simulation. The inverse distortion parameter is used to introduce a pre-compensated modulation amount at the control parameter level to compensate for distortion effects, such as nonlinear distortion and electromagnetic coupling interference, so that the beam output is as close as possible to the theoretical target. Then, the complete control parameters with inverse distortion correction are sent to the beam simulation or measurement system to make the generated beam even closer to the ideal target effect, thereby improving the control accuracy and energy efficiency of the entire system.

[0068] Furthermore, as a feasible implementation, it is assumed that the ideal beam parameters and the predicted beam parameters are identical in content; the data content of the ideal beam parameters and the predicted beam parameters includes:

[0069] The main lobe gain value represents the radiation intensity in the main lobe direction corresponding to the peak value of the beam gain.

[0070] The main lobe pointing error value represents the degree of deviation of the actual pointing direction of the main lobe from the predetermined direction;

[0071] Sidelobe level value represents the ratio of radiated power in the sidelobe direction to the peak power of the main lobe;

[0072] The sidelobe pseudolobe value represents the parasitic high-gain lobe value caused by the phase quantization error of the independent channel.

[0073] The main lobe gain value is the maximum gain value in the main lobe direction of the beam pattern, and it is the core standard value for RIS beamforming. The main lobe pointing error value represents the deviation angle between the actual direction of the main lobe and the target direction, and is used to measure the accuracy of beamforming. The smaller the value, the better, reflecting the RIS control accuracy. The sidelobe level value is the ratio of the radiation intensity in other directions in the beam pattern (excluding the main lobe) to the main lobe. It is used to control spurious interference, noise immunity, and confidentiality. Under normal circumstances, an ideal beam should have low sidelobes. The sidelobe level value is an abnormally high-gain lobe caused by inter-channel phase quantization error or non-ideality, which may be located in an unexpected direction. In characterizing the amplification effect of non-ideal errors in the control system, it is used to evaluate the actual RIS accuracy loss or hardware impact.

[0074] The calculation process for the first efficiency value and the second efficiency value is set as follows:

[0075] After normalizing the data content of both the ideal beam parameters and the predicted beam parameters, a scoring function is generated. The scoring functions for the main lobe gain value, main lobe pointing error value, side lobe level value, and side lobe pseudolobe value are respectively labeled as the first function f1, the second function f2, the third function f3, and the fourth function f4, and corresponding weight values ​​m1, m2, m3, and m4 are set. Let the beam efficiency function be represented by η.

[0076] The beam efficiency function η is then expressed as: η = m1∙f1 + m2∙f2 + m3∙f3 + m4∙f4,

[0077] Among the four weight values, the first weight value m1 has the largest value; the calculation process of the first efficiency value and the second efficiency value is the same as the calculation formula of the beam efficiency function η.

[0078] Normalization maps the four indices of the ideal beam parameters and the predicted beam parameters to the same dimensional range, eliminating the influence of differences in the dimensions and numerical spans of the indices themselves. Each normalized index is then converted into a score through a mapping function, which can be a linear or nonlinear mapping in practice. The first weight m1 is the largest, meaning that the main lobe gain is the most important in this embodiment, followed by the other related indices. These four scoring functions are combined with their corresponding weights in a linear weighted manner to obtain the beam efficiency function η, which represents the overall efficiency. The larger the value of η, the closer the beam performance of the predicted beam parameters is to the ideal configuration. By using normalization and weighted scoring, the four key beam performance indices—main lobe gain, pointing error, sidelobe level, and pseudolobe—are fused into a beam efficiency function η that represents the overall efficiency. At the same time, the beam efficiency function η is used to measure the quality of the ideal beam and the predicted beam, thus providing a unified and adjustable quantitative standard for determining the effectiveness of multi-channel collaborative calibration.

[0079] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A multi-channel cooperative control method based on a RIS phased array, characterized in that, The method includes: Step S1: Divide the array elements of the target RIS phased array into several independent channels, preset ideal beam parameters for each independent channel, and calculate the first efficiency value representing the ideal beamforming efficiency based on the ideal beam parameters. Step S2: Label the original reflection coefficient of each independent channel, perform closed-loop calibration on all original reflection coefficients, activate all independent channels after closed-loop calibration as collaborative multi-channel, and construct a beam coupling model for collaborative multi-channel. Step S3: Use the beam coupling model to predict and generate the synthetic beam of the cooperative multi-channel array control, and label the generated data of the synthetic beam as the predicted beam parameters representing the tendency of the cooperative multi-channel beam results. Step S4: Calculate a second efficiency value representing the predicted beamforming efficiency based on the predicted beam parameters, and compare the second efficiency value with the first efficiency value. If the first efficiency value is lower than the second efficiency value, the calibration is determined to be effective; if the first efficiency value is higher than the second efficiency value, the calibration is determined to be ineffective, and the process returns to step S2. The beam coupling model includes a Riemannian manifold algorithm and a beam generator; the working process of the beam coupling model includes: The original reflection coefficients after closed-loop calibration are constructed into complex reflection coefficients for cooperative multi-channel operation. The phase amplitude space is modeled as a complex Hilbert manifold. The complex reflection coefficients are mapped to the complex Hilbert manifold using a nonlinear mapping. A manifold loss function is constructed based on the complex reflection coefficients, and the Riemann gradient of the manifold loss function is calculated. The control parameters of the beam generator are obtained. The Riemann gradient and control parameters are mapped to the tangent space of the complex Hilbert manifold for geodesic update analysis. The control parameters are updated using exponential mapping in the tangent space. The updated control parameters are input into the beam generator to output the synthesized beam. The generated data of the synthesized beam are collected and labeled as the predicted beam parameters.

2. The multi-channel cooperative control method based on a RIS phased array according to claim 1, characterized in that, The closed-loop calibration of all original reflection coefficients includes: detecting and marking the coupling strength of different independent channels, allocating calibration time slot lengths in a positive correlation manner according to the coupling strength between channels; constructing a TDM schedule table, and using the TDM schedule table to perform closed-loop parallel calibration of the original reflection coefficients in each independent channel according to the calibration time slot length.

3. The multi-channel cooperative control method based on a RIS phased array according to claim 2, characterized in that, The TDM scheduling table includes: dividing the independent channel into several channel groups of equal number, measuring the electromagnetic coupling strength between each pair of channel groups and representing it as a coupling strength matrix, calculating the coupling weight value of each channel group through the coupling strength matrix, and allocating calibration time to each channel group using the coupling weight value.

4. The multi-channel cooperative control method based on a RIS phased array according to claim 2, characterized in that, The original reflection coefficients are calibrated in a closed-loop parallel manner using an iterative least squares method. The process includes: Let k be the iteration number in the calibration process, set the calibration reflection coefficient Γ, and set the original reflection coefficient as the initial value of the iterative reflection coefficient; preset the target reflection coefficient Γ for each independent channel. t Given the control parameter θ, set the Jacobian matrix and represent it as J; set the error threshold for the iterative process; The formula for calculating the calibration reflection coefficient Г in each iteration is expressed as: , Wherein, the θ k-1 The control parameter value representing the previous iteration number, θ k The control parameter value representing the current iteration sequence number; the Г k The calibration reflection coefficient representing the current iteration order, Γ k-1 The calibration reflection coefficient represents the previous iteration number; the J k-1 The Jacobian matrix representing the ordinal number of the previous iteration; When calibrating the reflection coefficient Г k With the target reflection coefficient Г t When the difference is less than the error threshold, the iteration process stops and calibration is completed.

5. The multi-channel cooperative control method based on a RIS phased array according to claim 4, characterized in that, The correction error is defined as ε, and its calculation formula is as follows: , where ε k This represents the correction error for the current iteration sequence. Let the least squares objective value of the control parameter be denoted as Δθ, and solve it using the standard least squares formula. The formula for calculating the least squares objective value Δθ is as follows: , J H It is the conjugate transpose of the Jacobian matrix.

6. The multi-channel cooperative control method based on a RIS phased array according to claim 5, characterized in that, The formula for calculating the Jacobian matrix J is expressed as follows: J k This represents the value of the Jacobian matrix at the current iteration ordinal number.

7. The multi-channel cooperative control method based on a RIS phased array according to claim 1, characterized in that, After adding the inverse distortion parameter to the updated control parameters, input it into the beam generator to output the synthesized beam.

8. A multi-channel cooperative control method based on a RIS phased array according to claim 1, characterized in that, ... The ideal beam parameters and the predicted beam parameters are identical in content; the data content of the ideal beam parameters and the predicted beam parameters includes: The main lobe gain value represents the radiation intensity in the main lobe direction corresponding to the peak value of the beam gain. The main lobe pointing error value represents the degree of deviation of the actual pointing direction of the main lobe from the predetermined direction; Sidelobe level value represents the ratio of radiated power in the sidelobe direction to the peak power of the main lobe; The sidelobe pseudolobe value represents the parasitic high-gain lobe value caused by the phase quantization error of the independent channel.

9. A multi-channel cooperative control method based on a RIS phased array according to claim 8, characterized in that, The calculation process for the first efficiency value and the second efficiency value is set as follows: After normalizing the data content of both the ideal beam parameters and the predicted beam parameters, a scoring function is generated. The scoring functions for the main lobe gain value, main lobe pointing error value, side lobe level value, and side lobe pseudolobe value are respectively labeled as the first function f1, the second function f2, the third function f3, and the fourth function f4, and corresponding weight values ​​m1, m2, m3, and m4 are set. Let the beam efficiency function be represented by η. The beam efficiency function η is then expressed as: η = m1∙f1 + m2∙f2 + m3∙f3 + m4∙f4, Among the four weight values, the first weight value m1 has the largest value; the calculation process of the first efficiency value and the second efficiency value is the same as the calculation formula of the beam efficiency function η.

Images