A ris-assisted multi-user mimo downlink transmission method employing low-precision dac

Abstract

This invention discloses a RIS-assisted multi-user MIMO downlink transmission method using a low-precision DAC, comprising: deploying multiple RIS blocks to assist downlink communication between the base station and multiple antenna users; transmitting a portion of the precoded and low-precision DAC-quantized signal directly to the user, and the other portion arriving after reflection by the RIS; randomly configuring S groups of reflection coefficient phase shifts for the RIS; the base station using downlink channel state information obtained through channel estimation or feedback to calculate the precoding matrix using a dimensionality-reduced WMMSE algorithm; traversing the S groups of reflection coefficient phase shifts, recording the user weighted sum rate corresponding to each group of phase shifts, and the base station using the conditional sample mean method to calculate the optimal RIS reflection coefficients. This invention considers the nonlinear quantization of the low-precision DAC and the discrete phase shift constraints of the RIS, and reduces the computational complexity and pilot overhead of channel estimation.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology, and particularly relates to a RIS-assisted multi-user MIMO downlink transmission method using a low-precision DAC. Background Technology

[0002] In multi-user MIMO downlink systems, base station precoding can effectively reduce multi-user interference and improve the system's channel capacity. Furthermore, deploying multiple RIS (Reference Signal Processing) modules can improve the channel condition number and increase the multiplexing gain of signal transmission. In practical systems, to reduce the hardware cost and power consumption of the base station transmitter, the bit depth of the DAC (Digital Converter) is often limited to a low level. However, low-precision DACs can cause significant quantization errors, severely impacting system performance. Currently, in RIS-assisted wireless communication systems, the papers “Y. Xiu, J. Zhao, W. Sun and Z. Zhang, “Secrecy RateMaximization for Reconfigurable Intelligent Surface Aided Millimeter WaveSystem With Low-Resolution DACs,” in IEEE Commun. Lett., vol. 25, no. 7, pp.2166-2170, July 2021,” and “J. Dai, Y. Wang, C. Pan, K. Zhi, H. Ren and K. Wang, “Reconfigurable Intelligent Surface Aided Massive MIMO Systems With Low-Resolution DACs,” in IEEE Commun. Lett., vol. 25, no. 9, pp. 3124-3128, Sept.2021,” were the first to consider a design scheme using low-precision DACs in the base station, verifying its feasibility.

[0003] Unlike existing literature, this invention considers a scenario where users configure multiple antennas. To reduce the computational complexity of downlink precoding, this invention successfully extends the dimension-reduced WMMSE algorithm proposed in the paper "X. Zhao, S. Lu, Q. Shi and Z.-Q. Luo, "Rethinking WMMSE: Can Its Complexity Scale Linearly With the Number of BSAntennas?", in IEEE Trans. Signal Process., vol. 71, pp. 433-446, 2023" to a multi-purpose MIMO system using a low-precision DAC. Simultaneously, to reduce the pilot overhead of channel estimation, this invention utilizes the conditional sample mean method proposed in the paper "S. Ren, K. Shen, Y. Zhang, X. Li, X. Chen and Z.-Q. Luo, "Configuring Intelligent Reflecting Surface with PerformanceGuarantees: Blind Beamforming", in IEEE Transactions on Wireless Communications, 2022" to design the reflection coefficient matrix of the RIS, and designs it in conjunction with base station precoding. Summary of the Invention

[0004] The purpose of this invention is to provide a RIS-assisted multi-user MIMO downlink transmission method using a low-precision DAC. It jointly designs base station downlink precoding and RIS discrete phase shifting to improve the weighted sum rate of users. Unlike existing literature, this invention considers scenarios where users are configured with multiple antennas. To reduce the computational complexity of downlink precoding, this invention successfully extends the dimension-reduced WMMSE algorithm proposed in the paper "X. Zhao, S. Lu, Q. Shi and Z.-Q. Luo, "Rethinking WMMSE: Can Its Complexity Scale Linearly With the Number of BS Antennas?", in IEEE Trans.Signal Process., vol. 71, pp. 433-446, 2023" to multi-user MIMO systems using low-precision DACs. Meanwhile, to reduce the pilot overhead of channel estimation, this invention utilizes the conditional sample mean method proposed in the literature "S. Ren, K. Shen, Y. Zhang, X. Li, X. Chen and Z.-Q. Luo, "Configuring Intelligent Reflecting Surface with Performance Guarantees: Blind Beamforming," in IEEE Transactions on Wireless Communications, 2022" to design the reflection coefficient matrix of the RIS, and designs it in conjunction with base station precoding. This method, combined with RIS technology, can effectively reduce the DAC bit requirement, reduce computational complexity, and significantly improve system performance.

[0005] To solve the above-mentioned technical problems, the specific technical solution of the present invention is as follows:

[0006] A RIS-assisted multi-user MIMO downlink transmission method using a low-precision DAC includes the following steps:

[0007] Step 1: Deploy L blocks of RIS (Smart Metasurface) to assist downlink communication between a multi-antenna base station and K multi-antenna users. The base station has... The RIS establishes additional auxiliary links for downlink communication between the base station and the user, namely the base station to RIS link and the RIS to user link. Each user has Root receiving antenna, where variables The value satisfies ;No. Block RIS has Each reflection unit, where the variable The value satisfies ;

[0008] Step 2: The base station simultaneously sends modulated data symbol streams to the aforementioned users. These data symbol streams are first pre-coded by baseband and mapped to different antenna ports. The mapped transmitted signals are then quantized by a low-precision DAC (digital-to-analog converter). The resolution of the DAC is... 1 bit;

[0009] Step 3: Part of the quantized transmitted signal reaches the user via the direct link channel from the base station to the user, and the other part reaches the user via the base station to the RIS link, the adjustable RIS reflection, and the RIS to the user link. The phase shift of each reflection unit of the RIS is discrete, with a resolution of [missing value]. 1 bit;

[0010] Step 4: Randomly configure the phase shift of S groups of reflection coefficients for RIS. The phase shift of each group of reflection coefficients is maintained in multiple scheduling time slots. In this scheduling time slot, in order for the base station to obtain downlink channel state information, for TDD (Time Division Duplex) systems, the base station obtains it based on the pilot sequence sent by the user and utilizes channel reciprocity. For FDD (Frequency Division Duplex) systems, the user estimates the downlink channel state information based on the pilot sequence sent by the base station and feeds it back to the base station. At the same time, the base station uses the downlink channel state information to calculate the precoding matrix using the dimension-reduced WMMSE (Weighted Minimum Mean Square Error) algorithm with the user weighted sum rate maximization as the criterion.

[0011] Step 5: Traverse the phase shifts of the S groups of reflection coefficients, record the user weighted sum rate corresponding to each group of reflection coefficient phase shifts, and the base station uses the conditional sample mean method to calculate the optimal RIS reflection coefficient.

[0012] Furthermore, in step 2, the mapped transmitted signal is:

[0013]

[0014] in Representative sent to The data symbol vector of the nth user, for any nth user One user, Satisfying the autocorrelation matrix And the cross-correlation matrix ,symbol express 3D complex column vector, Indicates the first Number of data streams per user, symbol express The identity matrix, symbol express A matrix of all zeros, sign This represents taking the conjugate transpose of a vector or matrix; where It is the first Precoding matrices for each user, symbol express OK Column complex matrix;

[0015] go through Quantized transmit signal from a bit-resolution DAC for:

[0016]

[0017] in This refers to the quantization process of the DAC. This represents the quantization distortion factor, which is the reciprocal of the quantization signal-to-noise ratio; Is with Statistically uncorrelated quantization noise, with a mean of zero, has a covariance matrix that is approximately:

[0018]

[0019] in This means that the precoding matrices of all K users are arranged row-wise to form the total precoding matrix, and the total number of data streams is the sum of the individual data streams. ,symbol This indicates the creation of a diagonal matrix containing vector elements on the main diagonal.

[0020] Furthermore, in step 3, the quantized transmitted signal that has been reflected twice or more by the RIS will be ignored, and the base station will then transmit the signal to the next RIS. Valid channels for each user for:

[0021]

[0022] in, Indicates from the base station to the... Downlink channel for each user Indicates from the base station to the... The downlink channel of the block RIS. Indicates from the first Block RIS to the first Downlink channel for each user, index variable The value satisfies Subscript variables The value satisfies ; Indicates the first The diagonal reflection coefficient matrix of the block RIS, with main diagonal elements composed of Composed of several reflection coefficients;

[0023] The diagonal reflection coefficient matrix The The diagonal elements represent the reflection coefficient. Subscript variables The value satisfies ,symbol It is the imaginary unit, and its phase shift It is not continuously adjustable, that is:

[0024]

[0025] The set of phase shift values ​​is as follows: And the quantization interval is ;

[0026] No. Signal received by each user It is decomposed into four parts: useful signal, multi-user interference, quantization noise, and Gaussian white noise. Represented as:

[0027]

[0028] in It is the first Received noise for each user It is the first Noise power per user, symbol express The identity matrix.

[0029] Furthermore, in step 4, record the... The phase shift of the group reflection coefficient is , where variables The value satisfies superscript Represented as RIS configuration number The result after setting the reflection coefficient, yes The total number of reflection units in the block RIS; S sets of reflection coefficients are randomly generated, with phase shifts... Independently from the set according to a uniform distribution Generated in; configured for RIS After obtaining the set of reflection coefficients, the base station obtains the corresponding downlink channel state information, namely: the first set of reflection coefficients. One user effective channel subscript variables The value satisfies ;

[0030] Corresponding to the The reflection coefficients are used to concatenate the effective channels of all K users into a large channel matrix. ,in Establish the following weighted sum rate maximization problem :

[0031]

[0032]

[0033] Among them, weight The priority of the k-th user is represented by the equivalent noise covariance matrix. , This represents the maximum total power of all transmitting antennas at the base station;

[0034] The weighted sum rate maximization problem Optimal solution of the precoding matrix Constraints on the channel matrix In the row space, that is, satisfying ,in This is an introduced auxiliary matrix. The weighted sum rate maximization problem is solved using the dimension-reduced WMMSE algorithm to calculate the optimal precoding matrix. The specific steps include:

[0035] Step 4.1: Set the maximum number of iterations Convergence threshold Set the number of iterations ,initialization ,satisfy superscript Indicates the first The result of the next iteration;

[0036] Step 4.2: Calculate about Gram matrix ,symbol Indicates submatrix Arrange in columns;

[0037] Step 4.3: Fix Calculate the auxiliary matrix

[0038] ;

[0039] Step 4.4: Fix Update the receiver matrix ,symbol This means assigning the value on the right to the value on the left, updating the first... The result of the next iteration;

[0040] Step 4.5: Fix and Update the error weight matrix ;

[0041] Step 4.6: Fix and First calculate the auxiliary matrix Then calculate the auxiliary matrix respectively. and auxiliary matrix ;

[0042] Step 4.7: Update again

[0043]

[0044] in express The The submatrices are obtained by arranging them row by row. ;

[0045] Step 4.8: Determine if the convergence condition is met. or If not satisfied, let Return to step 4.3; if satisfied, calculate the precoding matrix corresponding to the s-th group of reflection coefficients. User-weighted sum rate .

[0046] Furthermore, in step 5, after traversing the set of reflection coefficient phase shifts... Afterwards, the weighted sum rate set was obtained. The optimal RIS reflectance coefficient is calculated using the conditional sample mean method, including the following steps:

[0047] Step 5.1: Calculate the first... Block RIS, the first Phase shift of reflection coefficient of each unit Exactly equal to sample index set and the number of elements in the set. , Perform calculations;

[0048] Step 5.2: Calculate the condition where... Under the condition, weighted sum rate The sample mean, i.e., the conditional sample mean: ;

[0049] Step 5.3: Select the phase shift with the largest conditional sample mean and calculate... ;

[0050] Step 5.4: [The sentence is incomplete and requires more context to be translated accurately.] As the optimal RIS reflection coefficient.

[0051] The RIS-assisted multi-user MIMO downlink transmission method using a low-precision DAC of the present invention has the following advantages:

[0052] 1. The present invention considers the nonlinear quantization of low-precision DAC when designing the downlink precoding, which improves the system performance, and uses the dimension-reduced WMMSE algorithm to further reduce the computational complexity.

[0053] 2. When designing the RIS reflection coefficient, this invention takes into account the discrete phase shift constraints that are more consistent with the actual system. It only needs to obtain the effective channel from the base station to each user, without needing to estimate the base station-RIS-user concatenated channel, which significantly reduces the pilot overhead of channel estimation. Attached Figure Description

[0054] Figure 1 This is a schematic diagram of the structure of a RIS-assisted multi-user MIMO downlink system using a low-precision DAC, provided in an embodiment of the present invention.

[0055] Figure 2(a) is a graph comparing the weighted sum rate of downlink data transmission with DAC accuracy provided by different precoding algorithms without RIS assistance, according to an embodiment of the present invention.

[0056] Figure 2(b) compares different precoding algorithms according to an embodiment of the present invention. A graph showing the weighted sum rate of downlink data transmission versus DAC accuracy in the case of BitRIS assistance.

[0057] Figure 3(a) compares the RIS of different phase shift accuracies in an embodiment of the present invention at the maximum total power of the base station. In this case, the cumulative distribution function curve of the weighted sum rate of the downlink data transmission is provided.

[0058] Figure 3(b) compares the RIS of different phase shift accuracies in an embodiment of the present invention at the maximum total power of the base station. In this case, the cumulative distribution function curve of the weighted sum rate of the downlink data transmission is provided.

[0059] Figure 4(a) is a graph comparing the weighted sum rate of downlink data transmission with the base station transmit power of DACs of different accuracies without RIS assistance, according to an embodiment of the present invention.

[0060] Figure 4(b) compares DACs of different precisions according to an embodiment of the present invention. The graph showing the weighted sum rate of downlink data transmission versus base station transmit power in the case of BitRIS assistance. Detailed Implementation

[0061] To better understand the purpose, structure, and function of this invention, the following detailed description, in conjunction with the accompanying drawings, provides a RIS-assisted multi-user MIMO downlink transmission method using a low-precision DAC.

[0062] This embodiment provides a RIS-assisted multi-user MIMO downlink transmission method using a low-precision DAC, wherein the structure of the RIS-assisted multi-user MIMO downlink system using a low-precision DAC is as follows: Figure 1 As shown.

[0063] The transmission method of this system specifically includes the following steps:

[0064] Step 1: Deploy L blocks of RIS to assist downlink communication between a multi-antenna base station and K multi-antenna users. The base station has... Root transmitting antenna;

[0065] In step 1, the RIS establishes additional auxiliary links for downlink communication between the base station and the user, namely the base station to RIS link and the RIS to user link. Each user has Root receiving antenna ( ), No. Block RIS has One reflective unit ( ).

[0066] Step 2: The base station simultaneously transmits modulated data symbol streams to K users. The data is first pre-coded by baseband and mapped to different antenna ports. The mapped transmitted signals are then quantized by a low-precision DAC, where the DAC resolution is [resolution missing]. 1 bit;

[0067] In step 2, the precoded transmission signal of the base station is as follows:

[0068]

[0069] in Representative sent to The data symbol vector of the nth user, for any nth user For each user, the autocorrelation matrix is ​​satisfied. And the cross-correlation matrix symbol express 3D complex column vector, Indicates the first Number of data streams per user, symbol express The identity matrix, symbol express A matrix of all zeros, sign This represents taking the conjugate transpose of a vector or matrix; where It is the first Precoding matrices for each user, symbol express OK Column complex matrix;

[0070] Transmitted signal after DAC quantization at B-bit resolution for:

[0071]

[0072] in This refers to the quantization process of the DAC. This represents the quantization distortion factor, which is the reciprocal of the quantization signal-to-noise ratio; Is with Statistically uncorrelated quantization noise, with a mean of zero, has a covariance matrix that is approximately:

[0073]

[0074] in This means that the precoding matrices of all K users are arranged row-wise to form the total precoding matrix, and the total number of data streams is the sum of the individual data streams. ,symbol This indicates the creation of a diagonal matrix containing vector elements on the main diagonal.

[0075] Step 3: Part of the quantized transmitted signal reaches the user through the base station-user direct link channel, and the other part reaches the user through the base station-RIS link, adjustable RIS reflection, and RIS-user link. The phase shift of each reflection unit of the RIS is discrete, with a resolution of [missing value]. 1 bit;

[0076] In step 3, signals reflected twice or more by the RIS are ignored, and the base station... Valid channels for each user for:

[0077]

[0078] in, , and No. Individual users, from base station to the first Block RIS and from the first Block RIS to the first Downlink channel for each user, index variable The value satisfies Subscript variables The value satisfies ; Indicates the first The diagonal reflection coefficient matrix of the block RIS, with main diagonal elements composed of It consists of a reflectance coefficient.

[0079] The diagonal matrix The The diagonal elements represent the reflection coefficient. Subscript variables The value satisfies ,symbol It is the imaginary unit, and its phase shift It is not continuously adjustable, that is:

[0080]

[0081] The set of phase shift values ​​is as follows: And the quantization interval is ;

[0082] No. The signal received by each user is for:

[0083]

[0084] in It is the first Received noise for each user It is the first Noise power per user, symbol express The identity matrix.

[0085] Step 4: Randomly configure the phase of S groups of reflection coefficients for RIS. The phase of each group of reflection coefficients is maintained for a short period of time. During this short period of time, for TDD systems, the base station obtains downlink channel state information based on the pilot sequence sent by the user and by utilizing channel reciprocity. For FDD systems, the user estimates the downlink channel state information based on the pilot sequence sent by the base station and feeds it back to the base station. At the same time, the base station uses the downlink channel state information to calculate the precoding matrix using the dimension-reduced WMMSE algorithm with the user weighted sum rate maximization as the criterion.

[0086] In step 4, record the first Group The phase shift of the reflection coefficient is , where variables The value satisfies superscript Represented as RIS configuration number The result after setting the reflection coefficient, yes The total number of reflection units in the block RIS; S sets of reflection coefficients are randomly generated, with phase shifts... Independently from the set according to a uniform distribution Generated in; configured for RIS After obtaining the set of reflection coefficients, the base station obtains the corresponding downlink channel state information, namely: the first set of reflection coefficients. One user effective channel subscript variables The value satisfies ;

[0087] Corresponding to the The reflection coefficients are used to concatenate the effective channels of all K users into a large channel matrix. ,in Establish the following weighted sum rate maximization problem :

[0088]

[0089]

[0090] Among them, weight The priority of the k-th user is represented by the equivalent noise covariance matrix. , This represents the maximum total power of all transmitting antennas at the base station;

[0091] The weighted sum rate maximization problem Optimal solution of the precoding matrix Constraints on the channel matrix In the row space, that is, satisfying ,in This is an introduced auxiliary matrix. The optimization problem described above is solved using the dimension-reduction WMMSE algorithm to calculate the optimal precoding matrix. The specific steps include:

[0092] Step 4.1: Set the maximum number of iterations Convergence threshold Set the number of iterations ,initialization ,satisfy superscript Indicates the first The result of the next iteration;

[0093] Step 4.2: Calculate about Gram matrix ,symbol Indicates submatrix Arrange in columns;

[0094] Step 4.3: Fix Calculate the auxiliary matrix

[0095] ;

[0096] Step 4.4: Fix Update the receiver matrix ,symbol This means assigning the value on the right to the value on the left, updating the first... The result of the next iteration;

[0097] Step 4.5: Fix and Update the error weight matrix ;

[0098] Step 4.6: Fix and First calculate the auxiliary matrix Then calculate the auxiliary matrix respectively. and auxiliary matrix ;

[0099] Step 4.7: Update again

[0100] in express The The submatrices are obtained by arranging them row by row. ;

[0101] Step 4.8: Determine if the convergence condition is met. or If not satisfied, let Return to step 4.3; if satisfied, calculate the precoding matrix corresponding to the s-th group of reflection coefficients. User-weighted sum rate .

[0102] Step 5: Traverse the S groups of reflection coefficient phases, record the user weighted sum rate corresponding to each group of reflection coefficient phases, and the base station uses the conditional sample mean method to calculate the optimal RIS reflection coefficient.

[0103] In step 5, after traversing the set of reflection coefficient phase shifts... Afterwards, the weighted sum rate set was obtained. The optimal RIS reflectance coefficient is calculated using the conditional sample mean method, including the following steps:

[0104] Step 5.1: Calculate the first... Block RIS, the first Phase shift of reflection coefficient of each unit Exactly equal to sample index set and the number of elements in the set. , Perform calculations;

[0105] Step 5.2: Calculate the condition where... Under the condition, weighted sum rate The sample mean, i.e., the conditional sample mean: ;

[0106] Step 5.3: Select the phase shift with the largest conditional sample mean and calculate... ;

[0107] Step 5.4: [The sentence is incomplete and requires more context to be translated accurately.] As the optimal RIS reflection coefficient.

[0108] To verify the correctness and advancement of this implementation method, a corresponding simulation experiment was conducted in this embodiment.

[0109] Specific simulation parameters include:

[0110] Consider two smart metasurfaces Set the base station to be located in At a distance of meters, each intelligent metasurface is located at... Mihe Rice, four users The height is Meters, and randomly distributed in positions. With meters as the center, radius Within a circular area of ​​meters. The threshold at which the WMMSE algorithm stops iterating. Weight All Unless otherwise specified, ,correspond Other system parameters are set to , , , and .

[0111] Specific simulation results include:

[0112] Figure 2(a) and Figure 2(b) show the results without RIS assistance and with... With the assistance of BitRIS, curves of the weighted sum rate of downlink data transmission versus DAC accuracy were plotted, and tests were conducted. and The performance of these two different base station transmit powers was compared with that of the classic WMMSE algorithm proposed in the literature "Q. Shi, M. Razaviyayn, Z. -Q.Luo and C. He, "An Iteratively Weighted MMSE Approach to Distributed Sum-Utility Maximization for a MIMO Interfering Broadcast Channel," in IEEE Trans. Signal Process., vol. 59, no. 9, pp. 4331-4340, Sept. 2011." Simulations were also performed on the performance of an unconstrained WMMSE algorithm equivalent to the classic WMMSE algorithm. The results show that, compared to the classic WMMSE and unconstrained WMMSE algorithms, even with lower-precision DACs, the performance loss of the dimensionality-reduced WMMSE algorithm is almost negligible. This algorithm avoids the performance degradation in each iteration. The inversion operation of an 1x1 matrix significantly reduces computational complexity.

[0113] Figures 3(a) and 3(b) show the base station transmit power at [value missing]. and In this case, the cumulative distribution function curves of the weighted sum rate of downlink data transmission were plotted and compared with the enhanced conditional sample mean method proposed in the literature "S. Ren, K. Shen, Y. Zhang, X. Li, X. Chen and Z.-Q. Luo, "Configuring Intelligent Reflecting Surface with Performance Guarantees: Blind Beamforming," in IEEE Transactions on Wireless Communications, 2022." The impact of different phase shift accuracies of the RIS on system performance was also investigated. The results show that the average performance of the conditional sample mean method is slightly lower than that of the enhanced conditional sample mean method, while the performance of the two methods is very close in the worst case. In particular, for the RIS with a 1-bit phase shift, the cumulative distribution function curves of the two methods almost overlap.

[0114] Figures 4(a) and 4(b) show the results without and without RIS assistance, respectively. With the aid of bit-based RIS (Reference-Based Receiver Assignment), the weighted sum rate of downlink data transmission was plotted as a function of the base station transmit power. Two different scenarios were tested: one without RIS and the other with 3-bit phase-shifted RIS. Performance was compared with the random maximum sampling method, and also with different DAC accuracies (1 bit to 4 bits and infinite bits). The results show that the weighted sum rate increases with increasing DAC accuracy; a DAC accuracy of 3 to 4 bits can approach the performance of an ideal DAC. Furthermore, regardless of whether a low-precision DAC or an ideal DAC is used, the enhanced conditional sample mean method outperforms the random maximum sampling method.

[0115] This invention proposes a RIS-assisted multi-user MIMO downlink transmission method using a low-precision DAC. In a RIS-assisted multi-user MIMO downlink communication system using a low-precision DAC, system performance is improved by jointly designing the base station precoding and the discrete phase shift of the RIS. The dimension-reduced WMMSE algorithm used in this invention guarantees a significant reduction in computational complexity with very little performance loss, and the conditional sample mean method can handle the discrete phase shift constraints of the RIS and reduce the pilot overhead of channel estimation. This invention has significant practical application value for improving low-cost, low-power, and high-performance multi-user MIMO downlink communication systems.

[0116] It is understood that the present invention has been described through some embodiments, and those skilled in the art will recognize that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the invention. Furthermore, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of the present invention.

Claims

1. A method for RIS-aided multi-user MIMO downlink transmission with low-precision DACs, characterized in that, Includes the following steps: Step 1: Deploy L blocks of RIS to assist downlink communication between a multi-antenna base station and K multi-antenna users. The base station has... The RIS establishes additional auxiliary links for downlink communication between the base station and the user, namely the base station to RIS link and the RIS to user link. Each user has Root receiving antenna, where variables The value satisfies ;No. Block RIS has Each reflection unit, where the variable The value satisfies ; Step 2: The base station simultaneously sends modulated data symbol streams to the aforementioned users. These data symbol streams are first pre-coded by baseband and mapped to different antenna ports. The mapped transmitted signals are then quantized by a low-precision DAC with a resolution of [missing information]. 1 bit; Step 3: Part of the quantized transmitted signal reaches the user via the direct link channel from the base station to the user, and the other part reaches the user via the base station to the RIS link, the adjustable RIS reflection, and the RIS to the user link. The phase shift of each reflection unit of the RIS is discrete, with a resolution of [missing value]. 1 bit; Step 4: Randomly configure the phase shifts of S groups of reflection coefficients for RIS. The phase shift of each group of reflection coefficients is maintained in multiple scheduling time slots. In order for the base station to obtain downlink channel state information, for TDD systems, the base station obtains it based on the pilot sequence sent by the user and utilizes channel reciprocity. For FDD systems, the user estimates the downlink channel state information based on the pilot sequence sent by the base station and feeds it back to the base station. At the same time, the base station uses the downlink channel state information to calculate the precoding matrix using the dimension-reduced WMMSE algorithm with the user weighted sum rate maximization as the criterion. Step 5: Traverse the phase shifts of the S groups of reflection coefficients, record the user weighted sum rate corresponding to each group of reflection coefficient phase shifts, and use the conditional sample mean method to calculate the optimal RIS reflection coefficient. In step 4, record the first... The phase shift of the group reflection coefficient is , where variables The value satisfies superscript Represented as RIS configuration number The result after setting the reflection coefficient, yes The total number of reflection units in the block RIS; S sets of reflection coefficients are randomly generated, with phase shifts... Independently from the set according to a uniform distribution Generated in; configured for RIS After obtaining the set of reflection coefficients, the base station obtains the corresponding downlink channel state information, namely: the first set of reflection coefficients. One user effective channel subscript variables The value satisfies ; Corresponding to the The reflection coefficients are used to concatenate the effective channels of all K users into a large channel matrix. ,in Establish the following weighted sum rate maximization problem : ； Among them, weight The priority of the k-th user is represented by the equivalent noise covariance matrix. ； This represents the maximum total power of all transmitting antennas at the base station; Indicates the quantization distortion factor; The weighted sum rate maximization problem Optimal solution of the precoding matrix Constraints on the channel matrix In the row space, that is, satisfying ,in This is an introduced auxiliary matrix. The weighted sum rate maximization problem is solved using the dimension-reduced WMMSE algorithm to calculate the optimal precoding matrix. The specific steps include: Step 4.1: Set the maximum number of iterations Convergence threshold Set the number of iterations ,initialization ,satisfy superscript Indicates the first The result of the next iteration; Step 4.2: Compute the Gram matrix with respect to the Gram matrix ； symbol Indicates submatrix Arrange in columns; Step 4.3: Fix Calculate the auxiliary matrix: ; Step 4.4: Fix Update the receiver matrix: ;symbol This means assigning the value on the right to the value on the left, updating the first... The result of the next iteration; Step 4.5: Fixing and updating the error weight matrix ； Step 4.6: Fix and First calculate the auxiliary matrix Then calculate the auxiliary matrix respectively. and auxiliary matrix ; Step 4.7: Update again in express The The submatrices are obtained by arranging them row by row: ; Step 4.8: Determine if the convergence condition is met: or If not satisfied, let Return to step 4.3; if satisfied, calculate the precoding matrix corresponding to the s-th group of reflection coefficients. Weighted sum rate and user-weighted sum rate: 。 2. The RIS-assisted multi-user MIMO downlink transmission method using a low-precision DAC according to claim 1, characterized in that, In step 2, the mapped transmission signal is: ； in Representative sent to The data symbol vector of the nth user, for any nth user One user, Satisfying the autocorrelation matrix And the cross-correlation matrix ,symbol express 3D complex column vector, Indicates the first Number of data streams per user, symbol express The identity matrix, symbol express A matrix of all zeros, sign This represents taking the conjugate transpose of a vector or matrix; where It is the first Precoding matrices for each user, symbol express OK Column complex matrix; After bit resolution DAC quantized transmit signal is: ； in The quantization process of a DAC is the reciprocal of the quantization signal-to-noise ratio. Is with Statistically uncorrelated quantization noise, with a mean of zero, has a covariance matrix that is approximately: ； in This means that the precoding matrices of all K users are arranged row-wise to form the total precoding matrix, and the total number of data streams is the sum of the individual data streams. ,symbol This indicates the creation of a diagonal matrix containing vector elements on the main diagonal.

3. The RIS-assisted multi-user MIMO downlink transmission method using a low-precision DAC according to claim 1, characterized in that, In step 3, the quantized transmitted signal that has been reflected twice or more by the RIS will be ignored. (The last part, "base station to the RIS," appears to be an unrelated fragment and is omitted from the translation.) Valid channels for each user for: ； in, Indicates from the base station to the... Downlink channel for each user Indicates from the base station to the... The downlink channel of the block RIS. Indicates from the first Block RIS to the first Downlink channel for each user, index variable The value satisfies Subscript variables The value satisfies ; Indicates the first The diagonal reflection coefficient matrix of the block RIS, with main diagonal elements composed of Composed of several reflection coefficients; The diagonal reflection coefficient matrix The The diagonal elements represent the reflection coefficient. Subscript variables The value satisfies ,symbol It is the imaginary unit, and its phase shift It is not continuously adjustable, that is: ； The set of phase shift values ​​is as follows: And the quantization interval is ; No. Signal received by each user It is decomposed into four parts: useful signal, multi-user interference, quantization noise, and Gaussian white noise. Represented as: ； in It is the first Received noise for each user It is the first Noise power per user, symbol express The identity matrix.

4. The RIS-assisted multi-user MIMO downlink transmission method with low-precision DACs of claim 1, wherein, In step 5, after traversing the set of reflection coefficient phase shifts... Afterwards, the weighted sum rate set was obtained. The optimal RIS reflectance coefficient is calculated using the conditional sample mean method, including the following steps: Step 5.1: Calculate the first... Block RIS, the first Phase shift of reflection coefficient of each unit Exactly equal to sample index set and the number of elements in the set. , Perform calculations; Step 5.2: Calculate the condition where... Under the condition, weighted sum rate The sample mean, i.e., the conditional sample mean: ； Step 5.3: Select the phase shift with the largest conditional sample mean and calculate... ; Step 5.4: [The sentence is incomplete and requires more context to be translated accurately.] As the optimal RIS reflection coefficient.

Images