Intelligent metasurface staged phase locking optimization method for MIMO (Multiple Input Multiple Output) system

Abstract

The invention discloses an intelligent metasurface staged phase locking optimization method for an MIMO (Multiple Input Multiple Output) system, which adopts a staged phase fixing and weight adjusting strategy, generates a random phase codebook in an initial stage, and performs phase locking by normalizing an effective rank and a reachable rate index of a channel. Weighting the normalized effective rank and the reachable rate based on a preset weight to obtain a comprehensive score, screening the configuration of the first two bits of the comprehensive score, and locking a phase overlapping unit; in the secondary stage, a secondary codebook is generated for the remaining units, a score is calculated by adopting an adjusted weight coefficient, and a phase locking position is updated and expanded; and in the final stage, generating a random codebook for an unfixed unit, and outputting the configuration with the highest comprehensive score as an optimal solution in combination with the weight coefficient. According to the method, hierarchical optimization of intelligent metasurface phase configuration is realized by gradually reducing the optimization space and comprehensively balancing multi-target indexes, the MIMO system performance and the calculation efficiency are effectively improved, and an efficient solution is provided for 6G intelligent wireless environment construction.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communication, and particularly relates to an intelligent metasurface phased locking optimization method for MIMO systems. Background Art

[0002] With the popular application of the millimeter-wave band in 5G cellular systems, the demand for wireless data transmission in existing wireless communication systems is increasing continuously. However, due to the short-wavelength characteristics of the millimeter-wave band, it faces challenges such as limited spectrum resources, signal interference, and coverage range limitations. MIMO technology can achieve higher spectral efficiency, reduce transmission interference, and increase system capacity by adding multiple antennas at the transmitter and receiver and utilizing the signal propagation characteristics in the spatial domain. However, traditional MIMO systems are still limited by the number of antennas and the acquisition of channel state information. In this context, as a new type of passive wireless device, RIS can change the surface emission characteristics by adjusting its own phase, thereby realizing the phase and amplitude adjustment of wireless signals. This precise adjustment enables RIS to effectively control the shape, direction, and power distribution of signal beams, thus achieving the directional transmission and enhancement of signals. RIS brings new opportunities to wireless communication systems and can significantly improve the coverage range, signal quality, and energy efficiency of the system.

[0003] Currently, RIS has been applied to MIMO wireless communication systems by many researchers to improve system performance in various scenarios. However, existing research mainly focuses on improving the overall achievable rate of the system and does not propose better optimization methods starting from improving characteristics such as the effective rank of the channel. At the same time, there are few effective optimization algorithms for the phase of RIS, and in a real intelligent metasurface RIS-assisted communication system, it exhibits high complexity and high codebook overhead and cannot be directly applied to practical systems. Therefore, there is an urgent need to propose an efficient RIS phase optimization method to simultaneously optimize the effective rank and achievable rate of the MIMO channel using RIS. Summary of the Invention

[0004] The purpose of the present invention is to solve the pain point of the complex calculation of existing RIS optimization methods, provide an effective solution for building a 6G intelligent wireless environment, and provide an intelligent metasurface phased locking optimization method for MIMO systems;

[0005] To achieve the above purpose, the present invention adopts the following technical solutions: An intelligent metasurface phased locking optimization method for MIMO systems, including:

[0006] Step 1: In the initial phase configuration stage, generate an initial codebook set, perform maximum normalization on the achievable rate and channel effective rank of each configuration in the initial codebook set, and then weight the maximum-normalized achievable rate and channel effective rank of each configuration in the initial codebook set based on a first preset weight coefficient to obtain the comprehensive scores of all configurations in the initial codebook set. Select the top two configurations with the highest comprehensive scores, record the set of unit positions with the same phase in these two configurations, fix the phase values of these positions, and the remaining unfixed positions enter the secondary phase optimization stage;

[0007] Step 2: In the secondary phase optimization stage, generate a secondary codebook for the remaining unfixed positions in Step 1, perform maximum normalization on the achievable rate and channel effective rank of each configuration in the secondary codebook, and then weight the maximum-normalized achievable rate and channel effective rank of each configuration in the secondary codebook based on a second preset weight coefficient to obtain the comprehensive scores of all configurations in the secondary codebook. Select the top two configurations with the highest comprehensive scores, update the set of overlapping phase positions, fix the phase values of the new overlapping positions, and the remaining unfixed positions enter the final phase determination stage;

[0008] Step 3: In the final phase determination stage, generate a final codebook for the remaining unfixed positions in Step 2, perform maximum normalization on the achievable rate and channel effective rank of each configuration in the final codebook, and then weight the maximum-normalized achievable rate and channel effective rank of each configuration in the final codebook based on a third preset weight coefficient to obtain the comprehensive scores of all configurations in the final codebook. Select the configuration with the highest comprehensive score as the optimal RIS phase configuration to complete the phased phase locking optimization and achieve the weighted maximization of the channel effective rank and achievable rate in the intelligent metasurface-assisted MIMO system.

[0009] Furthermore, the weighted maximization of the channel effective rank and achievable rate in the intelligent metasurface-assisted MIMO system is expressed as:

[0010]

[0011] where ω1 + ω2 = 1 and ω1, ω2 ≥ 0 are the weight coefficients of the effective rank and achievable rate; is the effective rank of the equivalent cascaded channel H between the transceiver ends, where the equivalent cascaded channel H is determined by the phase matrix and represents the reflection coefficient of the electromagnetic unit U n . Let φ = [φ1, φ2,..., φ N be the vector composed of the reflection coefficient phases of each electromagnetic unit of the RIS, which is the configuration vector of the RIS and satisfies where B = 2 b, b is the number of quantization bits of the phase of the reflection coefficient of each electromagnetic unit; rank(H) is the rank of the channel, σ′ i = σ i / Σ i σ i is the normalized value of each singular value σ i of the channel H; is the achievable rate of the system, where ρ is the transmit signal-to-noise ratio, [·] ii represents the i-th diagonal element of the matrix.

[0012] Furthermore, the expression for weighting the achievable rate and the effective channel rank after normalizing the maximum value of each configuration in the initial codebook set based on the first preset weight coefficient is:

[0013]

[0014] where α1 and β1 are the first preset weight coefficients, satisfying α1 + β1 = 1, C′ i is the achievable rate after normalizing the maximum value of each configuration in the initial codebook set, is the effective channel rank after normalizing the maximum value of each configuration in the initial codebook set.

[0015] Furthermore, the expression for weighting the achievable rate and the effective channel rank after normalizing the maximum value of each configuration in the secondary codebook based on the second preset weight coefficient is:

[0016]

[0017] where α2 and β2 are the second preset weight coefficients, satisfying α2 + β2 = 1, C″ i is the achievable rate after normalizing the maximum value of each configuration in the secondary codebook, is the effective channel rank after normalizing the maximum value of each configuration in the secondary codebook.

[0018] Furthermore, the expression for weighting the achievable rate and the effective channel rank after normalizing the maximum value of each configuration in the final codebook based on the third preset weight coefficient is:

[0019]

[0020] where α2 and β2 are the second preset weight coefficients, satisfying α2 + β2 = 1, C″′ i is the achievable rate after normalizing the maximum value of each configuration in the final codebook, is the effective channel rank after normalizing the maximum value of each configuration in the final codebook.

[0021] Further, the specific steps for screening the top two configurations with the highest comprehensive scores from all the configured comprehensive scores in the initial codebook set, recording the set of unit positions with the same phase in these two configurations, and fixing the phase values of these positions are as follows:

[0022] Select the top K (1) i highest two configurations from i.e., compare with for their phase configurations, record the set of unit positions with the same phase and fix the phase values of these positions as where K (1) i is the comprehensive score of all configurations in the initial codebook set, and Φ1 is the initial codebook set.

[0023] Further, the specific steps for screening the top two configurations with the highest comprehensive scores from the secondary codebook, updating the set of overlapping phase positions, and fixing the phase values of the new overlapping positions in step 2 are as follows:

[0024] Select the top K (2) i highest two configurations from i.e., and update the overlapping phases where K (2) i is the comprehensive score of all configurations in the secondary codebook, Φ2 is the secondary codebook, and is the remaining unfixed positions from step 1.

[0025] Further, the specific steps for screening the configuration with the highest comprehensive score from the final codebook as the optimal RIS phase configuration in step 3 are as follows:

[0026] Select the configuration φ with the highest K (3) i from all the configured comprehensive scores in the final codebook * i.e., and output it as the optimal RIS phase configuration, where K (3) i is the comprehensive score of all configurations in the final codebook, and Φ3 is the final codebook.

[0027] Beneficial effects: The present invention provides an intelligent metasurface phased phase-locking optimization method for MIMO systems. Based on the phased phase fixing and comprehensive index optimization strategy, it optimizes the intelligent metasurface RIS configuration through multi-stage random codebook generation and phase-locking mechanism, and designs a phased weight adjustment strategy to balance the effective channel rank and system spectral efficiency. The innovatively proposed phased phase fixing and comprehensive index optimization strategy in the present invention locks overlapping phase units through multi-stage iteration, gradually reduces the optimization space, significantly reduces the algorithm complexity, and differentially sets weight coefficients at different stages to balance the optimization objectives of the effective channel rank and spectral efficiency. Combining the normalized index and linear weighted scoring, it realizes multi-objective collaborative optimization, improves the overall performance of the MIMO system, and achieves a significant balance in the improvement of the effective channel rank and system spectral efficiency while reducing the algorithm complexity. Moreover, this method is not limited to a certain frequency band and has a wide application range. Therefore, the technical solution provided by the present invention solves the pain points of high computational complexity, single index, and large codebook overhead in existing RIS optimization methods, and provides an efficient solution for the construction of 6G intelligent wireless environments. Description of the Drawings

[0028] Figure 1 It is a schematic diagram of a RIS-assisted MIMO communication system in an embodiment of the present invention

[0029] Figure 2 It is a top view of a two-dimensional scenario of a RIS-assisted MIMO communication system in an embodiment of the present invention

[0030] Figure 3 It is a flowchart of the steps of an intelligent metasurface phased phase-locking optimization method for MIMO systems in an embodiment of the present invention

[0031] Figure 4 It is a schematic diagram of the pseudocode of the IFPA algorithm in an embodiment of the present invention

[0032] Figure 5 It is a parameter setting diagram of the simulation experiment in an embodiment of the present invention

[0033] Figure 6 It is a result diagram of the simulation of the RIS improving the effective channel rank of the MIMO channel based on the IFPA algorithm in an embodiment of the present invention

[0034] Figure 7 It is a result diagram of the simulation of the RIS improving the achievable rate of the MIMO channel based on the IFPA algorithm in an embodiment of the present invention. Detailed Embodiments

[0035] The following further explains the present invention with reference to the accompanying drawings.

[0036] An intelligent metasurface phased locking optimization method for MIMO systems according to the present invention. The implementation of this method includes the following steps:

[0037] Construct an RIS-assisted MIMO system model.

[0038] As Figure 1 described, the specific steps for constructing an RIS-assisted MIMO system model are as follows:

[0039] S1-1: The system includes a transmitter Tx equipped with L antennas, a receiver Rx equipped with Q antennas, and an intelligent metasurface RIS with N programmable electromagnetic units. The intelligent metasurface RIS generates a phase matrix Γ by an external controller to freely adjust the phase configuration of each unit of the RIS. There are LoS paths and NLoS paths between the transmitter and receiver.

[0040] S1-2: The effective channel H between the two ends of the communication system transceiver is modeled as:

[0041]

[0042] Among them, respectively represent the channel from the transmitter to the RIS and the channel from the RIS to the receiver, respectively represent the NLoS path and the LoS path between the two ends of the transceiver without RIS assistance; α ∈ (0, 1) represents the power ratio of the channel assisted by the RIS in the entire cascaded channel; K represents the Rice factor of the channel without RIS assistance;

[0043] S1-3: The channel from the transmitter to the RIS is expressed as:

[0044]

[0045] Among them, represents the channel from the transmitting antenna l (1 ≤ k ≤ L) to the nth electromagnetic unit U n . Considering the wireless signal propagation loss between them, the corresponding channel can be modeled as:

[0046]

[0047] Among them, G t represents the gain of the transmitting antenna, respectively represent the angles of arrival from the transmitting antenna l to the electromagnetic unit U n , represents the scattering pattern of the electromagnetic unit U n ; d x , d z respectively represent the length and width of the electromagnetic unit; represents the distance from the transmitting antenna l to the electromagnetic unit Un the distance; λ represents the carrier wavelength, i.e., c is the speed of light and f is the carrier frequency. The channel from the RIS to the receiver Similarly;

[0048] The diagonal matrix Γ composed of the reflection coefficients of each electromagnetic unit of S1-4, RIS is expressed as:

[0049]

[0050] where, represents the reflection coefficient of the electromagnetic unit U n and let φ = [φ1, φ2,..., φ N be the vector composed of the phases of the reflection coefficients of each electromagnetic unit of the RIS, which is the configuration vector of the RIS and satisfies where B = 2 b , and b is the quantization bit number of the phases of the reflection coefficients of each electromagnetic unit;

[0051] S1-5. The LoS path channel without RIS assistance is modeled as:

[0052]

[0053] where, represents the channel between the transmitting antenna l (1 ≤ l ≤ L) and the receiving antenna q (1 ≤ q ≤ Q), G r represents the receiving antenna gain, and d q,l represents the distance between the transmitting antenna l and the receiving antenna q;

[0054] S1-6. The channel of the NLoS path is modeled as:

[0055]

[0056] where, represents the minimum distance between the transceiver,

[0057] S1-7. The effective rank of the equivalent cascaded channel H between the transceiver ends can be expressed as:

[0058]

[0059] where, rank(H) is the rank of the channel, and σ′ i = σ i / ∑ i σ i is the normalized value of each singular value σ i of the channel H;

[0060] S1-8. The achievable rate of the system is:

[0061]

[0062] where ρ is the transmit signal-to-noise ratio, [·] ii represents the i-th diagonal element of the matrix;

[0063] S1-9. The weighted maximization of the channel effective rank and achievable rate of the intelligent metasurface-assisted MIMO system is expressed as:

[0064]

[0065] where ω1 + ω2 = 1 and ω1, ω2 ≥ 0 are the weight coefficients of the effective rank and achievable rate;

[0066] Optimize the RIS configuration in the RIS-assisted MIMO system model by using the intelligent metasurface phased locking optimization method for MIMO systems.

[0067] As Figure 3 shown, the process of the intelligent metasurface phased locking optimization method for MIMO systems is as follows: Generate a random phase codebook in the initial stage, normalize the channel effective rank and achievable rate indicators, obtain a comprehensive score by weighting the normalized effective rank and achievable rate based on preset weights, select the top two configurations with the comprehensive score, and lock the phase overlapping units.

[0068] In the secondary stage, generate a secondary codebook for the remaining units, calculate the score using the adjusted weight coefficients, and update and expand the phase locking positions.

[0069] In the final stage, generate a random codebook for the unfixed units, and output the configuration with the highest comprehensive score as the optimal solution in combination with the weight coefficients.

[0070] The specific implementation manner of the above process is as Figure 4 shown, including:

[0071] S2-1. First, generate an initial codebook set containing T1 random RIS phase configurations and measure the corresponding achievable rate for each configuration and the channel effective rank To eliminate the dimension difference, normalize the performance index vectors and Specifically, calculate the maximum value C (1) max = max(C (1) ) and and divide each element by the maximum value of its corresponding vector to obtain the normalized vectors C' and R', where,

[0072] S2-2. Define the linearly weighted comprehensive score for each group of configurations: Among them, α1 and β1 are the first preset weight coefficients, satisfying α1 + β1 = 1; select the comprehensive score K of all configurations in the initial codebook set (1) i The top two configurations with the largest Compare with the phase configurations of Compare with and record the set of unit positions with the same phase and fix the phase values of these positions to The remaining unfixed positions (denoted as that is, excluding the set from the set {1, 2,..., N} of elements) are iteratively optimized in the subsequent steps.

[0073] S2-3. Generate T2 random phase configurations for the units in to form a secondary codebook Among them the phase at the position is fixed, the phase at the position is random; measure all secondary codebooks of and and then perform maximum normalization to obtain the maximum-normalized achievable rate C″ of each configuration in the final codebook i and the channel effective rank and combine with the second preset weight coefficients α2, β2 through to calculate the comprehensive score K of all configurations in the secondary codebook (2) i , and satisfy α2 + β2 = 1, select K (2) i The two phase configurations with the highest weighted value and update the overlapping phases

[0074] S2-4. Generate T3 phase configurations Φ3, where the phase at the position is fixed, the phase at the position is random; measure all final codebooks of and and then perform maximum normalization to obtain the maximum-normalized achievable rate C″′ of each configuration in the final codebook i and the channel effective rank and combine with the third preset weight coefficients α3, β3 through to calculate the comprehensive score K of all configurations in the final codebook (3) i, and finally output the optimal configuration And satisfying α3 + β3 = 1, which is the optimal RIS phase configuration solved by this algorithm.

[0075] To further verify the present invention, the following simulation experiments are carried out. Based on the previously constructed RIS-assisted MIMO channel model, the transceiver is set in the near field of the RIS, and the centers of all the transmitting antennas, the centers of the receiving antennas and the RIS center are all set at the same height. The top view of the scenario is as Figure 2 shown.

[0076] To more intuitively reflect the optimization effect of IFPA on the channel effective rank and achievable rate, the Maximum Cross Algorithm (MCA) and the RIS with all-zero phase coding (i.e., the case where the RIS is not configured) are introduced in this simulation experiment for comparative experiments. The optimization objective of the MCA algorithm is only the effective rank. Its basic process is to first generate T random RIS phase configuration sets, after configuring the RIS, obtain the effective rank of the channel under the corresponding configuration, sort the T RIS configurations according to the corresponding effective rank values, and select the first N max configurations as the parent set. Then cross the phases of the first N new electromagnetic units in the RIS configurations of the parent set to generate a new round of offspring phase configuration sets. Finally, configure the RIS according to the newly generated offspring phase configuration sets to obtain the corresponding effective rank of the channel, and select the optimal RIS phase configuration in the offspring set according to the obtained effective rank result. The parameter settings related to the simulation are as Figure 5 shown

[0077] As Figure 6 shown, Figure 6 shows the variation of the effective rank of the RIS-optimized MIMO channel with the power ratio of the RIS-assisted channel in the entire cascaded channel under the optimization of the IFPA algorithm and the MCA algorithm. Under the optimization of the IFPA algorithm and the MCA algorithm, as the power ratio α of the RIS-assisted channel increases, the effective rank quickly improves to 2; when α is in the interval of 0 - 0.1, the optimization effect of the IFPA algorithm on the effective rank is slightly better than that of the MCA algorithm, with an average increase of 0.03. The effective rank of the all-zero phase coding is significantly inferior to the two algorithms, and as the channel power ratio α increases, the effective rank significantly decreases, and the maximum difference in the effective rank is nearly 0.65. The simulation results show that the effective rank of the MIMO channel under the RIS configuration optimized by the IFPA algorithm is significantly better than that of the RIS-assisted MIMO channel without configuration, and the optimization effect is comparable to that of the MCA algorithm with only the effective rank as the single optimization target.

[0078] As Figure 7 shown, Figure 7It shows the optimization of the phase configuration of RIS based on the IFPA algorithm and the MCA algorithm, and simulates the change of the achievable rate of the MIMO channel improved by RIS with the power ratio of the RIS-assisted channel in the entire cascaded channel. Under the optimization of the IFPA algorithm and the MCA algorithm, as the power ratio α of the RIS-assisted channel increases, the spectral efficiency is significantly improved. The optimization effect of the IFPA algorithm on the spectral efficiency is significantly better than that of the MCA algorithm, and it becomes more obvious as α increases, with an average increase of nearly 2.5 bps / Hz. The spectral efficiency of the all-zero phase coding is significantly inferior to that of the two algorithms, and as the channel power ratio α increases, the spectral efficiency significantly decreases. Compared with the IFPA algorithm, the maximum difference is nearly 9 bps / Hz. The simulation results show that the optimization effect of the IFPA algorithm on the achievable rate of the MIMO channel is significantly better than that of the MCA algorithm. Under the same codebook overhead, the IFPA algorithm can achieve a significant improvement in the balance of the effective rank of the channel and the system spectral efficiency, showing obvious advantages in channel optimization.

[0079] The above are only the preferred embodiments of the present invention. It should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. An intelligent metasurface phased phase-locking optimization method for MIMO systems, characterized in that Including: Step 1: In the initial phase configuration stage, generate an initial codebook set, normalize the achievable rate and channel effective rank of each configuration in the initial codebook set to the maximum value, then weight the achievable rate and channel effective rank of each configuration in the initial codebook set after maximum normalization based on the first preset weight coefficient to obtain the comprehensive scores of all configurations in the initial codebook set, and select the top two configurations with the highest comprehensive scores from them, record the set of unit positions with the same phase in these two configurations, and fix the phase values of these positions. The remaining unfixed positions enter the secondary phase optimization stage; Step 2: In the secondary phase optimization stage, generate a secondary codebook for the remaining unfixed positions in Step 1, normalize the achievable rate and channel effective rank of each configuration in the secondary codebook to the maximum value, then weight the achievable rate and channel effective rank of each configuration in the secondary codebook after maximum normalization based on the second preset weight coefficient to obtain the comprehensive scores of all configurations in the secondary codebook, and select the top two configurations with the highest comprehensive scores from them, update the set of overlapping phase positions, fix the phase values of the new overlapping positions. The remaining unfixed positions enter the final phase determination stage; Step 3: In the final phase determination stage, generate a final codebook for the remaining unfixed positions in Step 2, normalize the achievable rate and channel effective rank of each configuration in the final codebook to the maximum value, then weight the achievable rate and channel effective rank of each configuration in the final codebook after maximum normalization based on the third preset weight coefficient to obtain the comprehensive scores of all configurations in the final codebook, and select the configuration with the highest comprehensive score as the optimal RIS phase configuration to complete the phased phase locking optimization and achieve the weighted maximization of the channel effective rank and achievable rate in the intelligent metasurface-assisted MIMO system.

2. The intelligent metasurface phased locking optimization method for MIMO systems according to claim 1, wherein The weighted maximization of the channel effective rank and achievable rate in the intelligent metasurface-assisted MIMO system is expressed as: where ω1 + ω2 = 1 and ω1, ω2 ≥ 0 are the weight coefficients of the effective rank and the achievable rate; is the effective rank of the equivalent cascaded channel H between the transmitter and the receiver, where the equivalent cascaded channel H is determined by the phase matrix ; represents the reflection coefficient of the electromagnetic unit U n , and let φ = [φ1, φ2,..., φ N Let the vector composed of the phases of the reflection coefficients of each electromagnetic unit of the RIS be the configuration vector of the RIS, satisfying where B = 2 b , b is the quantization bit number of the phase of the reflection coefficient of each electromagnetic unit; rank(H) is the rank of the channel, σ′ i = σ i / ∑ i σ i is the normalized value of each singular value σ i of the channel H; is the achievable rate of the system, where ρ is the transmit signal-to-noise ratio, and [·] ii represents the i-th diagonal element of the matrix.

3. The intelligent metasurface phased locking optimization method for MIMO systems according to claim 1, wherein The expression for weighting the achievable rate and channel effective rank of each configuration in the initial codebook set after maximum normalization based on the first preset weight coefficient is: where α1 and β1 are the first preset weight coefficients, satisfying α1 + β1 = 1, C′ i is the achievable rate after normalizing the maximum value of each configuration in the initial codebook set, and is the effective channel rank after normalizing the maximum value of each configuration in the initial codebook set.

4. The intelligent metasurface phased-locking optimization method for MIMO systems according to claim 1, wherein The expression for weighting the achievable rate and channel effective rank of each configuration in the secondary codebook after maximum normalization based on the second preset weight coefficient is: where α2 and β2 are the second preset weight coefficients, satisfying α2 + β2 = 1, C″ i is the achievable rate after normalizing the maximum value of each configuration in the secondary codebook, and is the effective channel rank after normalizing the maximum value of each configuration in the secondary codebook.

5. The intelligent metasurface phased locking optimization method for MIMO systems according to claim 1, characterized in that The expression for weighting the achievable rate and channel effective rank of each configuration in the final codebook after maximum normalization based on the third preset weight coefficient is: where α2 and β2 are second preset weight coefficients satisfying α2 + β2 = 1, and C″′ i is the achievable rate after normalizing the maximum value of each configuration in the final codebook, and is the effective channel rank after normalizing the maximum value of each configuration in the final codebook.

6. The intelligent metasurface phased phase-locking optimization method for MIMO systems according to claim 1, wherein The specific steps for selecting the top two configurations with the highest comprehensive scores from the comprehensive scores of all configurations in the initial codebook set in Step 1, recording the set of unit positions with the same phase in these two configurations, and fixing the phase values of these positions are as follows: Select the top K configurations with the highest comprehensive scores from all the configurations in the initial codebook set (1) i The top two configurations with the largest scores And That is Compare With For the phase configurations, record the set of unit positions with the same phase And fix the phase values of these positions to be Where K (1) i Is the comprehensive score of all configurations in the initial codebook set, and Φ1 is the initial codebook set 7. The intelligent metasurface phased phase-locking optimization method for MIMO systems according to claim 1, wherein The specific steps for selecting the top two configurations with the highest comprehensive scores from the secondary codebook in Step 2, updating the set of overlapping phase positions, and fixing the phase values of the new overlapping positions are as follows: Select the top K configurations with the highest combined scores from all the configurations in the secondary codebook (2) i The top two configurations with the largest scores and i.e., and update the overlapping phases where K (2) i is the combined score of all the configurations in the secondary codebook, Φ2 is the secondary codebook, is the remaining unfixed position in step 1 8. The intelligent metasurface phased locking optimization method for MIMO systems according to claim 1, characterized in that The specific steps for selecting the configuration with the highest comprehensive score from the final codebook in Step 3 as the optimal RIS phase configuration are: Select the K with the highest comprehensive score among all the configurations in the final codebook (3) i The largest configuration φ * , that is And output it as the optimal RIS phase configuration, where K (3) i is the comprehensive score of all the configurations in the final codebook, and Φ3 is the final codebook

Images