CN117200844A - Intelligent metasurface-assisted beamforming design methods, media, equipment and systems - Google Patents

Abstract

The invention discloses an intelligent super-surface-assisted beam forming design method, medium, equipment and system, wherein the method comprises the following steps: acquiring channel information of an intelligent super-surface-assisted communication system, and calculating equivalent channel state information from a base station to user equipment; setting an initial random phase shift matrix, and updating a precoding matrix through the initial random phase shift matrix; constructing an objective function containing a phase shift matrix, and inversely solving and updating the phase shift matrix by utilizing the objective function to obtain an updated phase shift matrix; alternately iterating and updating the phase shift matrix and the precoding matrix; setting a gradient rising factor, and updating the updated precoding matrix and the phase shift matrix; setting a convergence accuracy condition; when the phase shift matrix is satisfied, ending the iteration and outputting the phase shift matrix at the moment; if not, continuing iteration until convergence accuracy conditions are met. The method can rapidly and accurately calculate the phase shift matrix, and solves the problem of overlong calculation time of the traditional wave beam forming design algorithm.

Description

Intelligent metasurface-assisted beamforming design methods, media, equipment and systems

Technical field

The invention belongs to the field of wireless communications, and specifically relates to an intelligent metasurface-assisted beamforming design method, storage medium, computer equipment and device.

Background technique

In recent years, intelligent metasurface (Reconfigurable intelligent surface, RIS) is a new emerging dynamic electromagnetic parameter control technology. Its low cost, low energy consumption, programmability and easy deployment can greatly improve the future wireless communication will face. The dilemma of ultra-complexity, high cost and high energy consumption has greatly reshaped the wireless communication environment. RIS itself is a programmable two-dimensional metamaterial, which consists of a large number of passive emission units. Each unit can independently impose a controllable amplitude and phase on the incident signal. The most important feature of RIS technology is that it can adjust electromagnetic wave response, reshape the wireless channel between transceiver devices, improve the received signal strength of terminal devices, and achieve interference management.

In the existing technology, in order to give full play to the ability of intelligent reflective surfaces to flexibly control the signal propagation environment, it is necessary to quickly and accurately calculate the phase shift matrix of intelligent metasurfaces. However, the calculation time based on traditional beamforming algorithms is very long, due to the user's For mobility, the base station needs to calculate the phase shift matrices corresponding to users at different locations in real time. This results in that in practical applications, when the RIS unit is large, the traditional beamforming design algorithm will take too long to calculate online. Disadvantages, resulting in poor performance due to the phase shift matrix and the user's position being out of sync, affecting the user experience.

Contents of the invention

The purpose of the present invention is to overcome the shortcomings of the above existing technologies and provide an intelligent metasurface-assisted beamforming design method. This method accelerates iterative updates through the "gradient rise" factor and can be quickly and accurately in the shortest time. The phase shift matrix is calculated, thereby overcoming the problem of poor performance caused by the traditional beamforming design algorithm's long online calculation time and the phase shift matrix being out of sync with the user's position.

A second object of the present invention is to provide a storage medium.

A third object of the present invention is to provide a computer device.

The fourth object of the present invention is to provide an intelligent metasurface-assisted beamforming design system.

In order to achieve the above objects, the present invention can be achieved by adopting the following technical solutions:

An intelligent metasurface-assisted beamforming design method includes the following steps:

S1. Obtain the channel information of the intelligent metasurface-assisted communication system. The intelligent metasurface-assisted communication system includes a base station, an intelligent metasurface and a user equipment, and then calculate the base station in the intelligent metasurface-assisted communication system through the calculation. Equivalent channel state information to user equipment; the channel information includes a channel state information matrix from the base station to the smart metasurface reflector, and a channel state information matrix from the smart metasurface reflector to user equipment;

S2. Set an initial random phase shift matrix, and update the precoding matrix through the initial random phase shift matrix; the phase shift matrix will change in each iteration. At this time, the precoding matrix will also be updated; during iteration, the update expression of the precoding matrix is as follows:

Among them, W represents the precoding matrix at the base station, h represents the equivalent channel matrix from the base station to the user equipment, h ^H represents the conjugate transpose matrix of the matrix h, τ is an additional variable introduced using the secondary transformation technology, m ≥0 is a power-limited Lagrangian multiplier, I _N represents an N×N identity matrix, and α is an auxiliary vector introduced by the Lagrangian dual transformation;

S3. According to the precoding matrix obtained in step S2, construct an objective function including a phase shift matrix, use the inverse solution of the objective function to update the phase shift matrix, and obtain an updated phase shift matrix;

The expression of the updated phase shift matrix is as follows:

Among them, P _jF (·) represents the projection, q introduces an additional variable for θ to solve θ, q≠θ, ρ is a penalty parameter, represents the Lagrangian variable, t is the number of iterations;

S4. Alternately and iteratively update the phase shift matrix and precoding matrix to obtain the updated phase shift matrix and precoding matrix;

S5. Set the gradient rising factor, and update the updated precoding matrix and phase shift matrix described in step S4 by using the gradient rising factor to obtain the optimized precoding matrix and phase shift matrix;

S6. Set the convergence accuracy condition; construct the optimized precoding matrix and phase shift matrix obtained in step S5 into a function, determine whether the function satisfies the convergence accuracy condition, and when the convergence accuracy condition is met, end the iteration and output this phase shift matrix; if the convergence accuracy condition is not met, continue to return to step S4 and update again until the convergence accuracy condition is met.

Preferably, the expression of the gradient rising factor in step S5 is as follows:

Among them, λ represents the gradient ascent factor, a is the set parameter value, a ^t represents the parameter value of the t-1th iteration, and a ^{t +1} represents the parameter value of the t-th iteration;

The update expression of the set parameter value a is as follows:

Preferably, the optimized precoding matrix described in step S5 is expressed as follows:

W ^t+1 =W ^t +λ*(W ^t -W ^t-1 )

Among them, λ represents the gradient rising factor, W represents the precoding matrix at the base station, W ^t-1 represents the t-1 iteration value, W ^t represents the precoding matrix of the t iteration, and W ^t+1 represents The precoding matrix of the t+1 iteration.

Preferably, the optimized phase shift matrix described in step S5 is expressed as follows:

θ ^t+1 =θ ^t +λ*(θ ^t -θ ^t-1 )

Among them, λ represents the gradient rising factor, θ represents the phase matrix of the smart metasurface, and is arranged according to the column vector, θ ^t-1 represents the phase matrix value of the t-1th iteration of the smart metasurface, θ ^t represents The phase matrix value of the smart metasurface at the t-th iteration, and θ ^t+1 represents the phase matrix value of the smart metasurface at the t+1-th iteration.

Preferably, the specific steps of step S4 are as follows:

First, according to the input phase shift matrix, the corresponding precoding matrix is calculated using the method of step S2, and then the optimized and updated phase shift matrix is calculated according to the corresponding precoding matrix using the method of step S3, and finally the updated precoding is obtained. matrix and phase shift matrix.

Preferably, the optimized precoding matrix and phase shift matrix described in step S6 are constructed into a function f1 expressed as follows:

f ₁ (W,Θ)＝log ₂ (1+SNR)

Among them, the SNR expression is

Among them, W represents the precoding matrix, h represents the equivalent channel matrix from the base station to the user equipment, h ^H represents the conjugate transposed matrix of the matrix h, δ ² represents the noise, and G ₁ represents the network in the RIS-assisted communication system. The channel state information matrix from the device side to the RIS reflection end, G ₂ is the channel state information matrix from the RIS reflection end to the user equipment, and Θ is the phase matrix.

Preferably, the convergence accuracy condition is expressed as follows:

f ₁ (iter)-f ₁ (iter-1)＜σ

Among them, iter represents the number of iterations, σ is the convergence accuracy, and f ₁ is the maximum information transfer rate function.

A storage medium is used to store non-transitory computer instructions. When the non-transitory computer instructions are executed, the intelligent metasurface-assisted beamforming design method is executed.

A computer device, including a processor and a memory. Non-transitory computer instructions are stored on the memory. When the non-transitory computer instructions are executed by the processor, the intelligent metasurface-assisted beamforming design is executed. method.

An intelligent metasurface-assisted beamforming design system, used to implement the intelligent metasurface-assisted beamforming design method, the system includes:

The channel state information acquisition module is used to obtain the base station-intelligent metasurface reflector channel state information, the base station-user equipment channel state information, and the intelligent reflector-user equipment channel state information under different beams;

A phase shift matrix generation module, configured to alternately optimize the precoding matrix and the phase shift matrix by using a beamforming algorithm adding a gradient rising factor based on the channel data obtained by the channel state information acquisition module, and accelerate the generation of the precoding matrix and the phase shift matrix. Update iteratively to generate a phase shift matrix that meets the convergence conditions;

The network device regulates the intelligent metasurface phase module, which is used to generate the phase of the intelligent metasurface through the phase shift matrix generation module. The base station in the intelligent metasurface-assisted communication system controls the controller of the intelligent metasurface through the RF link. The voltage or current signal of each unit is adjusted to change the voltage or current signal of each unit of the smart metasurface, thereby changing the reflection coefficient of each unit in the smart metasurface.

The present invention has the following advantages over the prior art:

(1) The intelligent metasurface-assisted beamforming design method of the present invention accelerates iterative updates and convergence through the "gradient rising" factor, thereby reducing the response time of the algorithm and being able to quickly and accurately calculate the corresponding Phase shift matrix, thereby overcoming the problem of poor performance of traditional beamforming design algorithms due to too long online calculation time and the phase shift matrix and user position being out of sync, and improving efficiency.

(2) The intelligent metasurface-assisted beamforming design method of the present invention is suitable for various RIS communication scenarios, has good application prospects in current and future communication systems, and has high engineering application value.

Description of the drawings

Figure 1 is a method flow chart of an intelligent metasurface-assisted beamforming design method provided in Embodiment 1 of the present invention;

Figure 2 is a schematic structural diagram of an intelligent metasurface-assisted communication system provided in Embodiment 1 of the present invention;

Figure 3 is a spectrum efficiency change curve diagram of an algorithm of an intelligent metasurface-assisted beamforming design method provided in Embodiment 1 of the present invention;

Figure 4 is a response time comparison graph of an algorithm of an intelligent metasurface-assisted beamforming design method provided in Embodiment 1 of the present invention;

Figure 5 is a schematic structural diagram of an intelligent metasurface-assisted beamforming design system provided in Embodiment 4 of the present invention;

Figure 6 is a work flow chart of the phase shift matrix generation module provided in Embodiment 4 of the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments These are some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts fall within the scope of protection of the present invention. .

Example 1

As shown in Figures 1 and 2, an intelligent metasurface-assisted beamforming design method includes the following steps:

S1. Obtain the channel information of the intelligent metasurface-assisted communication system. The intelligent metasurface-assisted communication system includes a base station, an intelligent metasurface and a user equipment, and then calculate the base station in the intelligent metasurface-assisted communication system through the calculation. Equivalent channel state information to the user equipment; the channel information includes the channel state information matrix G ₁ from the base station to the smart metasurface reflector, and the channel state information matrix G ₂ from the smart metasurface reflector to the user equipment;

Specifically, the equivalent channel state information h from the base station to the user equipment in the intelligent metasurface-assisted communication system is expressed as follows:

h＝G ₂ ^H ΘG ₁ (1)

Among them, h represents the equivalent channel from the base station to the user equipment, G ₁ represents the channel state information matrix from the base station to the smart metasurface reflector, and G ₂ represents the channel state information matrix from the smart metasurface reflector to the user equipment.

S2. Set an initial random phase shift matrix, and update the precoding matrix through the initial random phase shift matrix; the phase shift matrix will change in each iteration. At this time, the precoding matrix Will also be updated.

Specifically, the specific process of step S2 is as follows:

S21. First set an initial random phase shift matrix. The phase shift matrix is an N-dimensional matrix, and the diagonal of the N-dimensional matrix is the reflection coefficient of each unit in the smart metasurface;

The N-dimensional phase shift matrix is expressed as follows:

Θ＝diag{θ} (2)

Among them, Θ represents the RIS phase shift matrix, θ = [θ ₁ , θ ₂ ,...θ _N ] ^T , represents the amplitude coefficient, θ represents the reflection coefficient of each unit, θ represents the column vector composed of all units, the dimension is Nx1, Θ is the diagonal form of θ, and the dimension is NxN.

S22. Set the precoding matrix on the network device side to W and the noise to δ ² . Then the signal-to-noise ratio SNR at the user receiving end is expressed as follows:

Then the maximum information transfer rate f ₁ is expressed as:

f ₁ (W, Θ)＝log ₂ (1+SNR) (4)

S23. Set the transmission power P, introduce an additional variable τ, and use quadratic technology to optimize the SNR structure W, a function of τ;

Specifically, since Θ has been initialized at this time, it is necessary to obtain the precoding matrix W. Therefore, the initialized Θ is substituted into the maximum information transmission rate function f _1. In order to more conveniently solve the variable precoding matrix W at this time, new additional variables are introduced. τ, so that function f ₁ is rewritten as function f _2. At this time, f ₂ is related to the encoding matrix W and the additional variable τ. The optimization function f ₂ with respect to W, τ is expressed as follows:

f ₂ (W,τ)＝|hW| ² -τ(|hW| ² +δ ² ) (5)

The precoding matrix W satisfies the following conditions:

tr(WW ^H )≤P (6)

Among them, W ^H represents the transpose of W.

S24. Then a power-limited Lagrangian multiplier can be derived based on the optimization function obtained in step S24; the Lagrangian multiplier The expression is as follows:

Among them, m≥0 is the power-limited Lagrangian multiplier, and m is the auxiliary variable.

S25. Calculate W according to formula (7) to update the phase shift matrix, and calculate the precoding matrix and additional variable τ when the maximum information transmission rate f ₁ is maximum by combining formulas (5) and (6).

The precoding matrix is expressed as follows:

Among them, W represents the precoding matrix at the base station, h represents the equivalent channel matrix from the base station to the user equipment, h ^H represents the conjugate transpose matrix of the matrix h, τ is an additional variable introduced using the secondary transformation technology, m ≥0 is a power-limited Lagrangian multiplier, _IN represents an N×N identity matrix, and α is an auxiliary vector introduced by the Lagrangian dual transformation.

The additional variable τ is expressed as follows:

Among them, α is an additional parameter, and the value is equal to α=SNR.

To sum up, during iteration, the update expression of the precoding matrix is as shown in formula (8).

S3. According to the precoding matrix obtained in step S2, construct the objective function Y(q, θ, κ _U , κ _I ) including the phase matrix from the function f ₁ , and use the inverse solution of the objective function to update the phase shift matrix to obtain the updated The phase shift matrix;

Specifically, the specific process of step S3 is as follows:

S31. According to the precoding matrix obtained in step S2, let the auxiliary variable ν = diag (G ₂ ^H ) G ₁ W and calculate the additional variable ρ; the additional variable ρ is expressed as follows:

S32. According to step S31, derive the additional variable Δ,

Δ＝|ρ| ² υυ ^H (11)

Among them, Δ and All are additional variables introduced to facilitate auxiliary calculations.

S33. Reconstruct the objective function and set constraints;

Specifically, since the precoding matrix W is known and what needs to be solved is Θ, the precoding matrix W is substituted into the function f ₁ for deformation and transformed into a function about θ. The reconstructed objective function f ₃ is specifically expressed as follows:

Among them, f ₃ is the objective function about θ transformed by f ₁ .

The constraints are expressed as follows:

S34. An additional variable q is introduced for θ, and q≠θ is set; therefore:

Among them, the constraint conditions are stq=θ, θ _i ∈[0,2π], ρ>0 is a penalty parameter.

S35. Construct a new function Y(q,θ,κ _U ,κ _I ). The new function Y(q,θ,κ _U ,κ _I ) is expressed as follows:

Among them, κ _R =[κ _R,1 ,…,κ _R,N ] ^T ,κ _I =[κ _I,1 ,…,κ _I,N ] ^T is a Lagrangian variable, Re{q-θ} =0, Im{q-θ}=0,

Then, the dual problem of formula (16) is transformed and expressed as follows:

S36 iterates the function formula (16) obtained in step S35 to obtain the updated phase shift matrix θ ^t+1 ; the θ ^t+1 is expressed as follows:

in, t is the number of iterations.

After optimizing formula (18), the final updated phase shift matrix is expressed as follows:

Among them, P _jF (·) represents the projection, q introduces an additional variable for θ to solve θ, q≠θ, ρ is a penalty parameter, represents the Lagrangian variable, and t is the number of iterations.

S4. Alternately and iteratively update the phase shift matrix and precoding matrix to obtain the updated phase shift matrix and precoding matrix;

The steps of step S4 are as follows:

First, according to the input phase shift matrix, the corresponding precoding matrix is calculated using the method of step S2, and then the optimized and updated phase shift matrix is calculated according to the corresponding precoding matrix using the method of step S3, and finally the updated precoding is obtained. matrix and phase shift matrix.

S5. Set the gradient rising factor, and update the updated precoding matrix and phase shift matrix described in step S4 by using the gradient rising factor to obtain the optimized precoding matrix and phase shift matrix;

The expression of the gradient ascent factor is as follows:

Among them, λ represents the gradient ascent factor, a is the set parameter value, a ^t represents the parameter value of the t-1th iteration, and a ^{t +1} represents the parameter value of the t-th iteration.

The update expression of the set parameter value a is as follows:

Specifically, the optimized precoding matrix is expressed as follows:

W ^t+1 ＝W ^t +λ*(W ^t -W ^t-1 ) (22)

Among them, λ represents the gradient rising factor, W represents the precoding matrix at the base station, W ^t-1 represents the t-1 iteration value, W ^t represents the precoding matrix of the t iteration, and W ^t+1 represents The precoding matrix of the t+1 iteration.

Specifically, the optimized phase shift matrix described in step S5 is expressed as follows:

θ ^t+1 =θ ^t +λ*(θ ^t -θ ^t-1 ) (23)

Among them, λ represents the gradient rising factor, θ represents the phase matrix of the smart metasurface, and is arranged according to the column vector, θ ^t-1 represents the phase matrix value of the t-1th iteration of the smart metasurface, θ ^t represents The phase matrix value of the smart metasurface at the t-th iteration, and θ ^t+1 represents the phase matrix value of the smart metasurface at the t+1-th iteration.

S6. Set the convergence accuracy condition; construct the optimized precoding matrix and phase shift matrix obtained in step S5 into a function, determine whether the function satisfies the convergence accuracy condition, and when the convergence accuracy condition is met, end the iteration and output this phase shift matrix; if the convergence accuracy condition is not met, continue to return to step S4 and update again until the convergence accuracy condition is met.

Specifically, the optimized precoding matrix and phase shift matrix described in step S6 are constructed into function f1, which is expressed as follows:

f ₁ (W,Θ)＝log ₂ (1+SNR)

Among them, the SNR expression is

Among them, W represents the precoding matrix, h represents the equivalent channel matrix from the base station to the user equipment, h ^H represents the conjugate transposed matrix of the matrix h, δ ² represents the noise, and G ₁ represents the network in the RIS-assisted communication system. The channel state information matrix from the device side to the RIS reflection end, G ₂ is the channel state information matrix from the RIS reflection end to the user equipment, and Θ is the phase matrix.

The convergence accuracy conditions are expressed as follows:

f ₁ (iter)-f ₁ (iter-1)＜σ

Among them, iter represents the number of iterations, and σ is the convergence accuracy.

As shown in Figure 3, when the set frequency is 28Ghz, the user is a single antenna, the number of base station antennas is 4, the number of RIS units is 20*20, the RIS unit size is 0.5λ*0.5λ, σ=1e-4, the operation of the computer Under the condition of 12GB of memory, the algorithm proposed in this embodiment is basically consistent with the fractional planning algorithm in improving spectral efficiency. Compared with the ADMM algorithm, the algorithm proposed in this embodiment improves the spectral efficiency significantly.

As shown in Figure 4, Figure 4 specifically evaluates the convergence performance of the algorithm proposed in this embodiment, the fractional planning algorithm, and the ADMM algorithm. In the figure, as the number of units in the RIS increases, the response times of the ADMM algorithm and the fractional planning algorithm increase. Significant increase, but the response time of the algorithm proposed in this embodiment is much lower than the first two. This is because this embodiment adds a gradient descent factor in the algorithm. This setting will greatly reduce the total number of iterations, so that the response time It will also be significantly reduced.

Specifically, after many experiments, when the gradient descent factor is set to an appropriate fixed value, the algorithm of this embodiment still converges faster and in a shorter time than the ADMM algorithm.

Example 2

A storage medium used to store non-transitory computer instructions. When the non-transitory computer instructions are executed, the intelligent metasurface-assisted beamforming design method described in Embodiment 1 is executed.

Example 3

A computer device, including a processor and a memory. Non-transitory computer instructions are stored on the memory. When the non-transitory computer instructions are executed by the processor, the intelligent metasurface-assisted beam described in Embodiment 1 is executed. Shape design method.

Example 4

As shown in Figure 5, an intelligent metasurface-assisted beamforming design system includes:

The channel state information acquisition module is used to obtain the base station-intelligent metasurface reflector channel state information, the base station-user equipment channel state information, and the intelligent reflector-user equipment channel state information under different beams;

A phase shift matrix generation module, configured to alternately optimize the precoding matrix and the phase shift matrix by using a beamforming algorithm adding a gradient rising factor based on the channel data obtained by the channel state information acquisition module, and accelerate the generation of the precoding matrix and the phase shift matrix. Update iteratively to generate a phase shift matrix that meets the convergence conditions;

The network device regulates the intelligent metasurface phase module, which is used to generate the phase of the intelligent metasurface through the phase shift matrix generation module. The base station in the intelligent metasurface-assisted communication system controls the controller of the intelligent metasurface through the RF link. The voltage or current signal of each unit is adjusted to change the voltage or current signal of each unit of the smart metasurface, thereby changing the reflection coefficient of each unit in the smart metasurface.

Specifically, the convergence accuracy condition is expressed as follows:

f ₁ (iter)-f ₁ (iter-1)＜σ

Among them, iter represents the number of iterations, σ is the convergence accuracy, and f ₁ is the maximum information transfer rate function.

It should be noted that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When executed, the computer program can include The procedures of the above method embodiments are as follows. Wherein, the computer-readable storage medium may include non-volatile and/or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory.

It can be understood that RAM has many forms, such as synchronous DRAM (SDRAM), enhanced SDRAM (ESDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), memory bus dynamic RAM ( RDRAM), static RAM (SRAM), dynamic RAM (DRAM), etc.

The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, etc. may be made without departing from the spirit and principles of the present invention. All simplifications should be equivalent substitutions, and are all included in the protection scope of the present invention.

Claims (10)

1. The intelligent super-surface assisted beam forming design method is characterized by comprising the following steps of:

s1, obtaining channel information of an intelligent super-surface auxiliary communication system, wherein the intelligent super-surface auxiliary communication system comprises a base station, an intelligent super-surface and user equipment, and then calculating equivalent channel state information from the base station to the user equipment in the intelligent super-surface auxiliary communication system; the channel information comprises a channel state information matrix from the base station to the intelligent super-surface reflecting end and a channel state information matrix from the intelligent super-surface reflecting end to user equipment;

s2, setting an initial random phase shift matrix, and updating a precoding matrix through the initial random phase shift matrix; the phase shift matrix changes in each iteration, and at this time, the precoding matrix is updated; in iteration, the update expression of the precoding matrix is as follows:

wherein W represents a precoding matrix of a base station end, h represents an equivalent channel matrix from the base station to user equipment, and h ^H The conjugate transpose of matrix h, τ is an additional variable introduced by the quadratic transformation technique, m.gtoreq.0 is a power-limited Lagrangian multiplier, I _N Representing an N x N identity matrix, α being the auxiliary vector introduced by the lagrangian dual transform;

s3, constructing an objective function containing a phase shift matrix according to the pre-coding matrix obtained in the step S2, and inversely solving and updating the phase shift matrix by utilizing the objective function to obtain an updated phase shift matrix;

the expression of the updated phase shift matrix is as follows:

wherein P is _jF (. Cndot.) represents projection, q introduces an additional variable for θ to solve for θ, q. Cndot. Is a penalty parameter,representing the Lagrangian variable, t being the number of iterations;

s4, alternately iterating and updating the phase shift matrix and the precoding matrix to obtain an updated phase shift matrix and an updated precoding matrix;

s5, setting a gradient ascending factor, and updating the updated precoding matrix and the phase shift matrix in the step S4 through the gradient ascending factor to obtain an optimized precoding matrix and an optimized phase shift matrix;

s6, setting convergence accuracy conditions; constructing the optimized precoding matrix and phase shift matrix obtained in the step S5 into a function, judging whether the function meets the convergence accuracy condition, ending iteration and outputting the phase shift matrix at the moment when the function meets the convergence accuracy condition; if the convergence accuracy condition is not satisfied, the process returns to step S4, and the process is updated again until the convergence accuracy condition is satisfied.

2. The intelligent super-surface assisted beam forming design method as claimed in claim 1, wherein the gradient rising factor in step S5 is expressed as follows:

wherein λ represents a gradient-increasing factor, a is a set parameter value, a ^t A represents the parameter value of the t-1 th iteration, and a ^t+1 Parameter values representing the t-th iteration;

the update expression of the set parameter value a is as follows:

3. the intelligent super-surface assisted beamforming design method according to claim 1, wherein the optimized precoding matrix in step S5 is represented as follows:

W ^t+1 ＝W ^t +λ*(W ^t -W ^t-1 )

wherein lambda represents a gradient rising factor, W represents a precoding matrix of a base station end, and W ^t-1 Represents the iteration value of the t-1 th time, W ^t Representing the precoding matrix of the t-th iteration, and W ^t+1 Representing the precoding matrix for the t+1th iteration.

4. The intelligent subsurface assisted beamforming design method according to claim 1, wherein the optimized phase shift matrix in step S5 is represented as follows:

θ ^t+1 ＝θ ^t +λ*(θ ^t -θ ^t-1 )

wherein lambda represents a gradient rising factor, theta represents a phase matrix of the intelligent super-surface, and is arranged according to column vectors, and theta ^t-1 Representing the phase matrix value, theta, of the t-1 th iterative intelligent super surface ^t Phase matrix values representing the intelligent subsurface of the t-th iteration, and θ ^t+1 Representing the phase matrix value of the intelligent subsurface for the t+1st iteration.

5. The intelligent super-surface-assisted beam forming design method as claimed in claim 1, wherein the specific steps of step S4 are as follows:

firstly, according to an input phase shift matrix, a corresponding precoding matrix is calculated by utilizing the method of the step S2, then, according to the corresponding precoding matrix, an optimized and updated phase shift matrix is calculated by utilizing the method of the step S3, and finally, the updated precoding matrix and the phase shift matrix are obtained.

6. The intelligent super-surface assisted beamforming design method according to claim 1, wherein the optimized precoding matrix and phase shift matrix in step S6 are configured as the expression of the function f1 as follows:

f ₁ (W,Θ)＝log ₂ (1+SNR)

wherein the SNR expression is

Wherein W is expressed as a precoding matrix, h is expressed as an equivalent channel matrix from the base station to the user equipment, and h ^H Conjugate transpose of matrix h, delta ² Is noise, G ₁ G, as a channel state information matrix from a network equipment side to an RIS reflecting side in the RIS auxiliary communication system ₂ For the RIS reflection end to user equipment channel state information matrix, Θ is the phase matrix.

7. The intelligent subsurface assisted beamforming design method according to claim 1, wherein the convergence accuracy condition is expressed as follows:

f ₁ (iter)-f ₁ (iter-1)＜σ

wherein iter represents iteration number, sigma is convergence accuracy, f ₁ As a function of the maximum information transfer rate.

8. A storage medium storing non-transitory computer instructions that, when executed, perform the intelligent super-surface assisted beamforming design method according to any of claims 1-7.

9. A computer device comprising a processor and a memory having stored thereon non-transitory computer instructions that, when executed by the processor, perform the intelligent super-surface assisted beamforming design method according to any of claims 1-7.

10. An intelligent subsurface assisted beamforming design system for implementing the intelligent subsurface assisted beamforming design method as recited in any of claims 1-7, said system comprising:

the channel state information acquisition module is used for acquiring the channel state information of the base station-intelligent super-surface reflecting end, the channel state information of the base station-user equipment and the channel state information of the intelligent reflecting surface-user equipment under different wave beams;

the phase shift matrix generation module is used for alternately optimizing the precoding matrix and the phase shift matrix by adopting a beam forming algorithm added with gradient ascending factors according to the channel data acquired by the channel state information acquisition module, accelerating the iterative update of the precoding matrix and the phase shift matrix and generating a phase shift matrix meeting convergence conditions;

the network equipment regulates and controls the intelligent super-surface phase module, which is used for generating the phase of the intelligent super-surface through the phase shift matrix generation module, and a base station in the intelligent super-surface auxiliary communication system regulates the voltage or current signal of each unit through a controller of the RF link control intelligent super-surface, so that the voltage or current signal of each unit of the intelligent super-surface is changed, and the reflection coefficient of each unit in the intelligent super-surface is changed.