Irs-assisted noma network transmission method based on hardware damage

Abstract

The present invention relates to the technical field of wireless communications, and particularly relates to an intelligent reflecting surface (IRS)-assisted non-orthogonal multiple access (NOMA) network transmission method based on hardware damage. The method comprises: building an IRS-assisted NOMA network system; building a base station transmission power minimization model on the basis of a user quality of service constraint, a successive interference cancellation constraint and a reflection phase shift constraint; solving the base station transmission power minimization model to obtain an optimal transmission solution; and the system performing transmission on the basis of the optimal transmission solution. An optimal transmission power result for multiple users under the NOMA technology and with the assistance of the IRS can be reflected, enhancing the robustness of the system.

Description

An IRS-assisted NOMA network transmission method based on hardware damage Technical Field

[0001] The present invention belongs to the field of wireless communication technology, and in particular relates to an IRSTO-assisted NOMA network transmission method based on hardware damage. Background Art

[0002] To meet the development needs of the mobile internet and the Internet of Everything, future 6G networks will support holographic and high-precision communications. To achieve this, enormous amounts of data must be processed in near real time, which will lead to significant energy consumption and communication instability in communication systems. Against this backdrop, achieving low-power transmission while improving the robustness of communication systems has become a key research issue in the field of wireless communications. In recent years, non-orthogonal multiple access (NOMA) technology has been recognized as an effective technique for improving spectrum efficiency. Specifically, NOMA utilizes superposition coding and successive interference cancellation (SIC) to serve multiple users within the same resource block. However, NOMA networks are still susceptible to communication instability. Intelligent reflecting surfaces (IRS) are considered a key technology for future wireless communication systems. An IRS is a plane composed of a large number of reconfigurable passive reflective elements, each of which can independently modify the amplitude and phase of the incident signal, effectively improving the signal propagation environment and enhancing spectrum utilization and robustness.

[0003] Currently, the use of radio frequency technology and large-scale base stations can partially solve the above problems. However, due to the active nature of radio frequency technology, it does not meet the current requirements for low energy consumption, and large-scale base stations also face the same problem. In existing technologies, there are some examples of combining smart reflective surfaces with non-orthogonal multiple access technologies. For example, the IRS-assisted uplink NOMA network model maximizes the number of users and data rates under user quality of service (QoS) constraints, and reveals the relationship between the number of IRS reflective elements and users and data rates. The IRS-assisted NOMA downlink network system, which integrates multicast and unicast, maximizes the unicast data rate by optimizing the IRS's passive beamforming vectors while ensuring the multicast data rate.

[0004] However, all of the aforementioned studies assume perfect channel state information (CSI) and transceiver hardware impairments (HWI), which are simply not possible in reality. Imperfect CSI exists during signal propagation due to issues like obstruction that interfere with the signal propagation environment, and CSI update delays that affect channel estimation. Furthermore, in real-world applications, issues such as transceiver hardware aging, oscillator noise, and low-resolution digital-to-analog converters often arise, leading to hardware impairments in both base stations and user equipment.

[0005] Summary of the Invention

[0006] In view of the shortcomings of the existing technology, the present invention proposes an IRS-assisted NOMA network transmission method based on hardware damage, which includes:

[0007] S1: Build an IRS-assisted NOMA network system;

[0008] S2: Construct a base station transmission power minimization model based on user service quality constraints, serial interference cancellation constraints, and reflection phase shift constraints;

[0009] S3: Solve the base station transmission power minimization model to obtain the optimal transmission plan; the system transmits according to the optimal transmission plan.

[0010] Preferably, the IRS-assisted NOMA network system includes: a base station equipped with N antennas, an intelligent reflecting surface and K single-antenna users; the intelligent reflecting surface is equipped with L reflecting units.

[0011] Preferably, the base station transmission power minimization model is expressed as:

[0012] Among them, w k represents the active beamforming vector sent by BS (base station) to user k, K represents the total number of users, e represents the passive beamforming vector of IRS, R k→k represents the decoding rate of user k, R min Indicates the user minimum rate threshold, R j→k represents the decoding rate of user j decoding user k’s signal, Ω(j) represents the decoding order of user j, Ω(k) represents the decoding order of user k, e l represents the lth element of the passive beamforming vector of the IRS.

[0013] Preferably, the process of solving the base station transmission power minimization model includes:

[0014] S31: Constructing a composite channel uncertainty model;

[0015] S32: rewrite the base station transmission power minimization model based on the composite channel uncertainty model;

[0016] S33: The rewritten base station transmission power minimization model is divided into the active beamforming vector optimization subproblem and the passive beamforming vector optimization subproblem;

[0017] S34: Solve the active beamforming vector optimization subproblem and the passive beamforming vector optimization subproblem to obtain the optimal transmission solution.

[0018] Furthermore, the composite channel uncertainty model is expressed as:

[0019] Among them, H k represents the composite channel matrix from the base station to user k, represents the estimated value of the composite channel, ΔH k represents the estimation error of the composite channel, ξ h,k Indicates the radius of the error area.

[0020] Furthermore, the active beamforming vector optimization subproblem is expressed as:

[0021] Among them, w k represents the active beamforming vector sent by the BS to user k, K represents the total number of users, R k→k represents the decoding rate of user k decoding its own signal, R j→k represents the decoding rate of user j decoding user k’s signal, Ω(j) represents the decoding order of user j, Ω(k) represents the decoding order of user k, and I (L×N) represents the L×N order unit matrix, A k,k 、a k,k 、C k Represent the first, second and third intermediate parameters respectively, α a,k represents the proportional coefficient of HWI at the receiving end of user k, λ k represents the interference plus noise power of user k, δ represents the variance of Gaussian white noise, e represents the passive beamforming vector of IRS, represents the estimated composite channel matrix of user k, ω k represents the active beamforming matrix, 0 1×N Represents a 1×N order zero matrix, I (K-k) represents the identity matrix of order Kk, ξ h,k Indicates the radius of the error area, I N represents the N×N order unit matrix, η h =[ηh,1 ,...,η h,K ] T ≥0 indicates the first slack variable, η h,K represents the first slack variable of user K, μ h =[μ h,1 ,...,μ h,K ] T ≥0 indicates the second slack variable, μ h,K represents the second slack variable of user K.

[0022] Furthermore, the passive beamforming vector optimization subproblem is expressed as:

[0023] Among them, p k represents the SINR residual of user k, K represents the total number of users, e represents the passive beamforming vector of IRS, e l The lth element of the IRS passive beamforming vector, I (L×N) represents the L×N order unit matrix, A k,k 、a k,k 、C k Represent the first, second and third intermediate parameters respectively, α a,k represents the proportional coefficient of HWI at the receiving end of user k, λ k represents the interference plus noise power of user k, δ represents the variance of Gaussian white noise, represents the estimated composite channel matrix of user k, ω k represents the active beamforming matrix, 0 1×N Represents a 1×N order zero matrix, I (K-k) represents the identity matrix of order Kk, ξ h,k Indicates the radius of the error area, I N represents the N×N order unit matrix, η h =[η h,1 ,...,η h,K ] T ≥0 indicates the first slack variable, η h,K represents the first slack variable of user K, μ h =[μ h,1 ,...,μ h,K ] T ≥0 indicates the second slack variable, μ h,K represents the second slack variable of user K, p represents the SINR residual matrix, H k represents the composite channel matrix from BS to user k, w i represents the active beamforming vector sent by the base station to user i, Λ k represents the total noise power of user k, Hj represents the composite channel matrix from the base station to user j.

[0024] Furthermore, the process of solving the active beamforming vector optimization subproblem and the passive beamforming vector optimization subproblem includes: in the active beamforming vector optimization subproblem, the non-convex term C2 is transformed by linear approximation and successive convex approximation methods to obtain a standard semidefinite programming problem, and the corresponding active beamforming vector value is obtained using a convex optimization toolbox; in the passive beamforming vector optimization subproblem, the non-convex term C2 is transformed by a penalized convex-concave process algorithm to obtain a convex optimization problem, and the corresponding passive beamforming vector value is calculated using a convex optimization toolbox; under the alternating optimization framework, the two optimization subproblems are iteratively solved to obtain the optimal solution of the original problem.

[0025] The beneficial effects of the present invention are as follows: the present invention takes into account the problems of transceiver hardware aging, oscillator noise, and low-resolution digital-to-analog converters that lead to impaired quality of transmitted and received signals, and takes into account the channel uncertainty that occurs in actual applications, and constructs a non-orthogonal multiple access model assisted by a smart reflector. This model can reflect the optimal transmission power results of multiple users under non-orthogonal multiple access technology and the assistance of smart reflectors, and achieves low-power transmission under the premise of considering high-power Wi-Fi and imperfect center-span interface (CSI). It can minimize energy consumption while ensuring system robustness, and is more in line with actual application scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] FIG1 is a flow chart of the IRS-assisted NOMA network transmission method based on hardware damage in the present invention;

[0027] FIG2 is a diagram of the IRS-assisted NOMA network system model based on hardware damage in the present invention;

[0028] FIG3 is a flow chart of solving the optimal transmission scheme in a preferred embodiment of the present invention;

[0029] FIG4 is a diagram of a simulation reference model in the present invention;

[0030] FIG5 is a diagram showing the relationship between the user minimum rate threshold and the base station transmission power of the present invention and the comparative solution. DETAILED DESCRIPTION

[0031] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0032] The present invention proposes an IRS-assisted NOMA network transmission method based on hardware damage, as shown in FIG1 . The method includes the following contents:

[0033] S1: Build an IRS-assisted NOMA network system.

[0034] As shown in Figure 2, an IRS-assisted NOMA network system is constructed. In a practical IRS-assisted NOMA network, due to issues such as power amplifier errors and oscillator noise, the high-frequency Wi-Fi (HWI) between the base station and user equipment (UE) is considered. Furthermore, due to factors such as residual SIC errors and wireless channel randomness, the system is built with consideration for imperfect CSI. The system model includes a base station equipped with N antennas, a smart reflector equipped with L reflective units, and K single-antenna users.

[0035] Assume that the channels from the base station to the smart reflector, from the smart reflector to the user, and from the base station to the user are: Phase shift matrix Indicates that The microcontroller is located between the smart reflector and the base station and is used to dynamically adjust the phase offset of the reflector elements to enhance the transmission performance of the non-orthogonal multiple access network.

[0036] S2: Construct a base station transmission power minimization model based on user service quality constraints, serial interference cancellation constraints, and reflection phase shift constraints.

[0037] User service quality constraints: C1:R k→k ≥R min R k→k =log2(1+SINR k→k )

[0038] Among them, R k→k represents the decoding rate of user k, R min Indicates the user minimum rate threshold, SINR k→k It is the signal-to-noise ratio of user k’s decoded signal; e = [θ 1] H represents the passive beamforming vector of IRS, θ l represents the phase shift matrix of the reflection array element l, and the superscript H represents the conjugate transposed matrix; represents the composite channel matrix from BS to user k, represents the channel matrix from BS to user k, represents the channel matrix from the smart reflection surface to user k, G represents the channel matrix from the BS to the smart reflection surface; w irepresents the active beamforming vector sent by BS to user i, w k represents the active beamforming vector sent by the BS to user k; n k =(e H H k )n t +n0+n a,k represents the total noise received by user k, Represented as n k The conjugate of n t represents the noise caused by HWI at the BS, and n0 represents the additive white noise at the receiving end of user k.

[0039] Successive interference cancellation constraints: C2:R j→k ≥R k→k ,Ω(j)>Ω(k) R j→k =log2(1+SINR j→k )

[0040] Among them, R j→k represents the decoding rate of user j decoding user k’s signal, Ω(j) represents the decoding order of user j, Ω(k) represents the decoding order of user k, and n j is the total noise received by user j.

[0041] Reflection phase shift constraint: C3:|e l | 2 =1,e l ∈e

[0042] Among them, e l Represents the lth element of matrix e.

[0043] The base station transmission power minimization model is expressed as:

[0044] Among them, C1 is the user service quality constraint, that is, the user service quality rate of the user is not less than R min C2 is the constraint that the user signal can be decoded smoothly, and C3 is the reflection phase shift constraint, that is, the reflection phase shift of the smart reflection surface has a modulo 1 constraint.

[0045] S3: Solve the base station transmission power minimization model to obtain the optimal transmission plan; the system transmits according to the optimal transmission plan.

[0046] The process of solving the base station transmission power minimization model includes:

[0047] S31: Construct a composite channel uncertainty model.

[0048] The present invention takes into account that channel uncertainty is inevitable in practical applications. Therefore, the present invention takes channel uncertainty into account in the process of establishing the transmission power minimization model and models the channel uncertainty as an additive model. The composite channel uncertainty model can be expressed as:

[0049] Among them, H k represents the composite channel from BS to user k, represents the estimated value of the composite channel, ΔH k represents the estimation error of the composite channel, ξ h,k Indicates the radius of the error area.

[0050] S32: Rewrite the base station transmission power minimization model according to the composite channel uncertainty model.

[0051] Considering channel uncertainty, the base station transmission power minimization model can be rewritten as:

[0052] Among them, η h =[η h,1 ,...,η h,K ] T ≥0 indicates the first slack variable, η h,K represents the first slack variable of user K, μ h =[μ h,1 ,...,μ h,K ] T ≥0 indicates the second slack variable, μ h,K represents the second slack variable of user K, I (L×N) represents the L×N order unit matrix, A k,k 、a k,k 、C k Represent the first, second and third intermediate parameters respectively, which are the designators in the linear approximation process and have no physical meaning.

[0053] α a,k represents the proportional coefficient of HWI at the receiving end of user k, λ k represents the interference plus noise power of user k, δ represents the variance of Gaussian white noise, e represents the passive beamforming vector of IRS, represents the estimated composite channel matrix of user k, ω k =[w k+1 ,...,w K ] represents the active beamforming matrix, 0 1×N Represents a 1×N order zero matrix, I (K-k)represents the identity matrix of order Kk, ξ h,k Indicates the radius of the error area, I N Represents the N×N identity matrix.

[0054] S33: The rewritten base station transmission power minimization model is divided into the active beamforming vector optimization subproblem and the passive beamforming vector optimization subproblem.

[0055] The variable coupling problem is solved by using the idea of ​​alternating optimization, and the rewritten base station transmission power minimization model is divided into two optimization sub-problems, namely the active beamforming vector {w k}Optimization subproblem and passive beamforming vector e Optimize subproblems.

[0056] Given a passive beamforming vector e , active beamforming vector {w k The optimization sub-problem can be expressed as: stC2,C4,C5,C6

[0057] Given the active beamforming vector {w k}, the optimization problem becomes the passive beamforming vector e Feasibility problem, passive beamforming vector e The optimization subproblem can be expressed as:

[0058] Where p=[p1,...,p K ] T represents the SINR residual matrix, p K represents the SINR residual vector of user K, H k represents the composite channel from BS to user k, Λ k represents the total noise power of user k, A j represents the total noise power of user j, H j represents the composite channel from BS to user j.

[0059] S34: Solve the active beamforming vector optimization subproblem and the passive beamforming vector optimization subproblem to obtain the optimal transmission solution.

[0060] In the active beamforming vector optimization subproblem, the non-convex term C2 is transformed through linear approximation and successive convex approximation methods, and the optimization subproblem is converted into a standard semidefinite programming problem, which can be expressed as:

[0061] The convex optimization toolbox is used to solve the above problem and obtain the corresponding active beamforming vector value {w k}.

[0062] In the passive beamforming vector optimization subproblem, the penalty convex-concave process algorithm is used to transform the non-convex term C2, resulting in a convex optimization problem, which can be expressed as:

[0063] Where q=[q1,...,q 2L ] T Expressed as the slack variable of the equivalent linear constraint, q l represents the slack variable of the lth equivalent linear constraint; represents the penalty term in the objective function; χ [m] is the regularization factor, which can be used to scale ||q||1 to control the feasibility of the constraint; represents the mth iteration of the lth passive beamforming vector element, represents the conjugate of the lth passive beamforming element, q L+l Represents the slack variable of the L+lth equivalent linear constraint.

[0064] In some preferred embodiments of the present invention, as shown in FIG3 , the process of solving the optimal transmission solution is as follows:

[0065] Step 1: Set system parameters: K, N, L, G, h d,k 、h r,k 、 ΔH k ,Φ,{w k}, e, ε, iterative initialization, beamforming vector e (0) 、{w k} (0) ; Set the convergence accuracy to ε = 10 -6 ;

[0066] Specifically: System parameters include the number of single-antenna users K, the number of base station antennas N, the number of smart reflector array elements L, the channel G from the base station to the smart reflector, and the channel h from the base station to user k. d,k , channel h from the smart reflector to user k r,k , the estimated value of the composite channel The estimation error of the composite channel ΔH k , the phase shift matrix Φ of the smart reflector, the active beamforming vector {w k}, passive beamforming vector e, convergence accuracy ε.

[0067] Step 2: Passive beamforming by given e (n) According to the active beamforming vector optimization sub-problem solving process, the active beamforming vector {w k} (n+1) .

[0068] Step 3: Through the {w k} (n+1) , the passive beamforming vector e is calculated according to the passive beamforming vector optimization subproblem solution process (n+1)

[0069] Step 4: Update {w k} (n+1) ={w k} (n) , e (n+1) =e (n) , n=n+1;

[0070] Step 5: Judgement If the conditions are met, then output {w k} (n) 、e (n) 、 Otherwise, go to step 2 to continue updating and iterating.

[0071] Evaluation of the present invention:

[0072] The following simulation results show the practical application of the present invention in detail.

[0073] 1) Simulation conditions

[0074] As shown in Figure 4, preferably, the number of base station antennas is set to 4, the number of smart reflector units is set to 12, and the number of users is designed to be 8. The path loss indexes from the base station to the smart reflector, the user, and the smart reflector to the user are set to 3.5, 2.2, and 3.6, respectively. The channel bandwidth B = 10MHz, the hardware damage coefficient α = 0.02, and the Gaussian noise power n0 = 10dBm. The coordinates of the base station antenna and the smart reflector in the system are located at (0m, 0m, 20m) and (60m, 20m, 20m), respectively, and the user equipment is randomly distributed within a range with a horizontal radius of 10m and a coordinate of (100m, 10m, 1m) as the center.

[0075] 2) Simulation results

[0076] Figure 5 describes the relationship between the feasibility rate and the HWI scaling factor. The feasibility rate is defined as the ratio of the number of channels that meet the user's QoS requirements to the total number of channels. It can effectively characterize the robustness of the system. As the HWI scaling factor increases, the feasibility rates of the algorithm of the present invention, the HWI robust algorithm, and the CSI robust algorithm decrease to varying degrees. The feasibility rate of the non-robust algorithm is always 0, but the feasibility rate of the algorithm of the present invention is always greater than that of the HWI robust algorithm and the CSI robust algorithm.

[0077] The above embodiments further illustrate the purpose, technical solutions and advantages of the present invention in detail. It should be understood that the above embodiments are only preferred implementation plans of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made to the present invention within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An IRS-assisted NOMA network transmission method based on hardware damage, characterized in that: include: S1: Build an IRS-assisted NOMA network system; S2: Construct a base station transmission power minimization model based on user service quality constraints, serial interference elimination constraints, and reflection phase shift constraints; S3: Solve the base station transmission power minimization model to obtain the optimal transmission plan; the system transmits according to the optimal transmission plan.

2. According to claim 1, the IRS-assisted NOMA network transmission method based on hardware damage is characterized in that: The IRS-assisted NOMA network system includes: a base station equipped with N antennas, an intelligent reflecting surface and K single-antenna users; the intelligent reflecting surface is equipped with L reflecting units.

3. According to claim 1, the IRS-assisted NOMA network transmission method based on hardware damage is characterized in that: The base station transmission power minimization model is expressed as: stC1:R k→k ≥R min C2:R j→k ≥R k→k ,Ω(j)>Ω(k) C3:|e l | 2 =1,e l ∈e Among them, w k represents the active beamforming vector sent by the base station to user k, K represents the total number of users, e represents the passive beamforming vector of IRS, R k→k represents the decoding rate of user k, R min Represents the user minimum rate threshold, R j→k represents the decoding rate of user j decoding user k’s signal, Ω(j) represents the decoding order of user j, Ω(k) represents the decoding order of user k, e l Represents the lth element of the passive beamforming vector of IRS.

4. According to claim 1, the IRS-assisted NOMA network transmission method based on hardware damage is characterized in that: The process of solving the base station transmission power minimization model includes: S31: Construct composite channel uncertainty model; S32: rewrite the base station transmission power minimization model according to the composite channel uncertainty model; S33: The rewritten base station transmission power minimization model is divided into an active beamforming vector optimization subproblem and a passive beamforming vector optimization subproblem; S34: Solve the active beamforming vector optimization subproblem and the passive beamforming vector optimization subproblem to obtain the optimal transmission solution.

5. According to claim 4, the IRS-assisted NOMA network transmission method based on hardware damage is characterized in that: The composite channel uncertainty model is expressed as: Among them, H k represents the composite channel matrix from the base station to user k, represents the estimated value of the composite channel, ΔH k represents the estimation error of the composite channel, ξ h,k Indicates the radius value of the error area.

6. According to claim 4, the IRS-assisted NOMA network transmission method based on hardware damage is characterized in that: The active beamforming vector optimization subproblem is expressed as: stC2:R j→k ≥R k→k ,Ω(j)>Ω(k) C6:η h ≥0,μ h ≥0 Among them, w k represents the active beamforming vector sent by the base station to user k, K represents the total number of users, R k→k represents the decoding rate of user k decoding its own signal, R j→k represents the decoding rate of user j decoding user k’s signal, Ω(j) represents the decoding order of user j, Ω(k) represents the decoding order of user k, I (L×N) represents the L×N-order identity matrix, A k,k 、a k,k , C k Represent the first, second and third intermediate parameters respectively, α a,k represents the proportionality coefficient of the HWI at the receiving end of user k, λ k represents the interference plus noise power of user k, δ represents the variance of Gaussian white noise, e represents the passive beamforming vector of IRS, represents the estimated composite channel matrix of user k, ω k represents the active beamforming matrix, 0 1×N represents a 1×N-order zero matrix, I (K-k) represents the identity matrix of order Kk, ξ h,k Indicates the radius value of the error area, I N represents the N×N unit matrix, η h =[η h,1 ,...,η h,K ] T ≥0 indicates the first slack variable, η h,K represents the first slack variable of user K, μ h =[μ h,1 ,...,μ h,K ] T ≥0 indicates the second relaxation Quantity, μ h,K represents the second slack variable of user K.

7. According to claim 4, the IRS-assisted NOMA network transmission method based on hardware damage is characterized in that: The passive beamforming vector optimization subproblem is expressed as: stC3:|e l | 2 =1,e l ∈e C6:η h ≥0,μ h ≥0 C8:p≥0 Among them, p k represents the SINR residual of user k, K represents the total number of users, e represents the passive beamforming vector of IRS, and e l The lth element of the IRS passive beamforming vector, I (L×N) represents the L×N-order identity matrix, A k,k 、a k,k , C k Represent the first, second and third intermediate parameters respectively, α a,k represents the proportionality coefficient of the HWI at the receiving end of user k, λ k represents the interference plus noise power of user k, δ represents the variance of Gaussian white noise, represents the estimated composite channel matrix of user k, ω k represents the active beamforming matrix, 0 1×N represents a 1×N-order zero matrix, I (K-k) represents the identity matrix of order Kk, ξ h,k Indicates the radius value of the error area, I N represents the N×N unit matrix, η h =[η h,1 ,...,η h,K ] T ≥0 indicates the first slack variable, η h,K represents the first slack variable of user K, μ h =[μ h,1 ,...,μ h,K ] T ≥0 indicates the second slack variable, μ h,K represents the second slack variable of user K, p represents the SINR residual matrix, H k represents the composite channel matrix from the base station to user k, w i represents the active beamforming vector sent by the base station to user i, Λ k represents the total noise power of user k, Λ j represents the total noise power of user j, H j represents the composite channel matrix from the base station to user j.

8. According to claim 4, an IRS-assisted NOMA network transmission method based on hardware damage is characterized in that: The process of solving the active beamforming vector optimization subproblem and the passive beamforming vector optimization subproblem includes: in the active beamforming vector optimization subproblem, the non-convex term C2 is transformed by linear approximation and successive convex approximation methods to obtain a standard semidefinite programming problem, and the corresponding active beamforming vector value is obtained using a convex optimization toolbox; in the passive beamforming vector optimization subproblem, the non-convex term C2 is transformed by a penalized convex-concave process algorithm to obtain a convex optimization problem, and the corresponding passive beamforming vector value is calculated using a convex optimization toolbox; in the alternating optimization framework, the two optimization subproblems are iteratively solved to obtain the optimal solution of the original problem.

Images