Sum rate optimization method of STAR-RIS auxiliary rate division multiple access system

Abstract

The invention relates to a sum rate optimization method of an STAR-RIS auxiliary rate division multiple access system, and belongs to the technical field of wireless communication. According to the method provided by the invention, the communication performance of a downlink rate division multiple access (RSMA) system assisted by an intelligent reconfigurable surface (STAR-RIS) capable of transmitting and reflecting at the same time is improved by combining a beam forming method and a rate allocation optimization method, and an effective technical reference is provided for improving the communication performance of a wireless communication system. According to the method, an optimization problem of maximization and rate is constructed, optimization parameters are a beam forming vector, a rate allocation vector and an STAR-RIS transmission and reflection beam forming matrix, then the optimization problem is decomposed into two sub-problems by using a block coordinate descent (BCD) method, the beam forming vector and the rate allocation vector are optimized by using continuous convex approximation (SCA), and the beam forming vector and the rate allocation vector are optimized by using the STAR-RIS transmission and reflection beam forming matrix. And based on fractional programming (FP) and quadratic constraint quadratic programming (QCQP) methods, STAR-RIS transmission and reflection beam forming matrixes are optimized, and finally, sum rate maximization under the power condition is realized through alternate iteration.

Description

Technical Field

[0001] The invention relates to a sum rate optimization method of a STAR-RIS assisted rate division multiple access system, belonging to the technical field of wireless communication. Background Art

[0002] In recent years, reconfigurable intelligent surfaces (RISs), also known as intelligent reflecting surfaces (IRS), have been widely regarded as a promising solution to improve network coverage and data rates to meet the needs of ubiquitous wireless connectivity. RISs are based on a planar array of metamaterials and contain multiple passive reflective elements, each of which can independently adjust the phase shift and amplitude through an intelligent controller to precisely manipulate the propagation path of the incident signal. As a planar metasurface with a large number of reconfigurable passive elements, RISs show broad development prospects. By adjusting the phase and amplitude of each element, RISs can effectively improve the propagation of incident signals and create a favorable transmission environment. However, traditional RISs can only reflect signals and can therefore only serve users within the 180° half-plane on the same side of the RIS. To overcome this limitation, a new type of RIS, called simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS), is proposed. This new type of RIS can simultaneously transmit and reflect incident signals to users on both sides of the RIS. Therefore, compared with the traditional RIS that can only reflect signals, STAR-RIS not only achieves full spatial coverage, but also introduces new degrees of freedom (DoF), thereby improving the overall system performance.

[0003] With the rapid development of advanced multimedia applications such as virtual reality, the next generation of wireless networks needs to have high spectral efficiency and large-scale connectivity. Non-orthogonal multiple access (NOMA) can provide services to multiple users simultaneously on the same frequency or time resources by grouping users in the power domain, thereby achieving higher spectral efficiency than traditional orthogonal multiple access (OMA). In the NOMA system, users need to decode all interference signals when receiving information, which significantly increases the computational complexity of signal processing. To solve this problem, the concept of rate splitting multiple access (RSMA) is proposed. The core idea of ​​RSMA is to divide the messages transmitted to users into public messages and private messages. Public messages are decoded by multiple users, while private messages are only received by specific users. When decoding public messages, users must first deal with the interference of other users; while when decoding private messages, the interference of other users' private messages is treated as noise. By flexibly adjusting the division ratio between public and private messages, a balance between computational complexity and data transmission rate can be achieved in RSMA. However, in wireless network environments, the practical application of RSMA faces multiple challenges, including the reasonable division of public and private messages, the efficient management of private message transmission resources, and the guarantee of message synchronization during transmission.

[0004] Deploying STAR-RIS has great potential in enhancing channel strength and achieving full spatial coverage, especially when the link quality between the base station and the user is poor. RSMA can achieve higher spectrum efficiency and system capacity than traditional multiple access schemes by splitting the message into multiple sub-messages, serving multiple users at the same time and flexibly managing interference. The combination of STAR-RIS and RSMA proposes a new paradigm for wireless communication architecture, effectively combining the advantages of both to improve system performance in Internet of Everything (IoE) networks. On the one hand, the deployment of large-scale STAR-RIS elements brings passive array gain, thereby reducing the need for a large number of antenna arrays in RSMA systems. On the other hand, the optimal amplitude and phase shift design of STAR-RIS effectively mitigates the impact of channel randomness and simplifies the complex design of Successive Interference Cancellation (SIC) receivers in RSMA systems.

[0005] In summary, although the above research contents have made some progress in the combination of STAR-RIS and RSMA, most of them are only analyzed for specific scenarios or conditions. In addition, the existing research still lacks consideration of issues such as resource allocation, power control and channel randomness. Summary of the invention

[0006] The purpose of the present invention is to provide a STAR-RIS assisted rate division multiple access system and rate optimization method, aiming to solve the technical problem that most of the existing researches only analyze specific scenarios or conditions, and still do not give enough consideration to issues such as resource allocation, power control and channel randomness.

[0007] To achieve the above object, the technical solution of the present invention is: a sum rate optimization method for a STAR-RIS assisted rate division multiple access system, the specific steps are:

[0008] Step 1: Establish a STAR-RIS-assisted RSMA downlink communication system model;

[0009] Step 2: Based on the established communication system model, the base station beamforming matrix, the STAR-RIS phase shift matrix and the rate allocation vector are taken as the optimization problem, and the problem statement of the optimization target is constructed with the goal of maximizing the sum rate;

[0010] Step 3: Given the STAR-RIS phase shift matrix, optimize the base station transmit beamforming matrix and rate allocation vector;

[0011] Step 4: Given the base station transmit beamforming matrix and rate allocation vector, optimize the STAR-RIS phase shift matrix;

[0012] Step 5: Jointly optimize the base station transmit beamforming matrix, rate allocation vector, and STAR-RIS phase shift matrix to maximize the system and rate.

[0013] The Step 1 is specifically as follows:

[0014] Step 1.1: The communication system consists of base stations, STAR-RIS and users. Assume that the base station BS consists of N t The STAR-RIS consists of M passive element units arranged evenly in a plane, and there are N users with a single antenna.

[0015] Step 1.2: All channels in the communication system adopt the Rice fading channel model. Assume that the channel gain from BS to STAR-RIS is expressed as The channel gain from STAR-RIS to the user is expressed as

[0016]

[0017] In the formula, ω B-S and is the Ricean factor of the BS-STAR-RIS link and the STAR-RIS-user link, G LoS and r i LoS represents the Rayleigh distribution line-of-sight component, G NLoS and r i NLoS represents the non-line-of-sight component;

[0018] Step 1.3: Assuming that the BS, STAR-RIS and users can obtain ideal channel state information, and the BS antenna adopts a uniform linear array, STAR-RIS uses an energy splitting protocol to divide the coverage area into a transmission area and a reflection area. The transmission and reflection coefficients are and According to the principle of conservation of energy, The refraction and reflection coefficient matrices are:

[0019]

[0020] Indicates whether the user is a transmission zone user or a reflection zone user;

[0021] Based on the basic idea of ​​RSMA, the signal sent by BS is:

[0022]

[0023] Among them, s0 represents a public message. Represents private messages for each user;

[0024] The basic idea of ​​RSMA is to send i The user's message is divided into two parts: a public part and a private part. The public parts of all users are combined and encoded into a single public message s0 using a standard codebook, on the other hand, each private part is encoded as a private message for each user

[0025] Step 1.4: Since there is no link between the BS and the user, when the BS sends a signal to the user, the signal is transmitted or reflected by STAR-RIS to the user. i The signal y received by the user i for:

[0026]

[0027] Among them, BS and mth i The large-scale path loss between users is expressed as α represents the path loss component, d B-S and They represent the distance between BS and STAR-RIS and the distance between STAR-RIS and the mth i The distance between users, H represents the transpose of the matrix, n i ~CN(0,σ 2 ) is expressed as the mth i The additive Gaussian white noise of the user, Θ i,p For m i The user's refraction and reflection coefficient matrices;

[0028] User m i The signal-to-interference-noise ratios of the public and private signals at are:

[0029]

[0030] Step 1.5: User m i The achievable rates (bps / Hz) for decoding the public stream and the expected private stream are expressed as:

[0031] c i =log2(1+γ c,i )

[0032]

[0033] To ensure that all users can successfully decode the public message, the rate of the public message is chosen to be A given public message min i∈N c i and assigned to user m i The rate a i , the data rate constraint for each user to receive public messages is:

[0034]

[0035] The SIC operation is implemented at the receiving end, and the transmission power of each user satisfies the constraint:

[0036]

[0037] Where θ is the minimum difference between the decoded signal power and the undecoded inter-user interference signal power plus the noise power, and the system sum rate is the sum of the public rate and the achievable private flow rate, expressed as:

[0038]

[0039] Among them, R i is the sum rate of the system.

[0040] The Step 2 is specifically as follows:

[0041] The beamforming matrix w and STAR-RIS phase shift matrix Θ at the BS in the joint optimization system network are i,p And the rate allocation vector, to maximize the system and rate, so the following optimization problem is constructed:

[0042]

[0043] in, R min represents the minimum rate required by all users, w = [w0; w1; ...; w N ] is represented as the beamforming matrix at the BS, a=[a1,a2,...,a N ] T It is represented as rate allocation vector constraint; C1 represents the minimum rate limit for all users, C2 represents ensuring that each user can decode the public information, C3 represents the SIC power constraint, P is the maximum transmit power of the base station, C4 represents the maximum transmit power constraint of the BS, C5 and C6 represent the refraction and reflection power constraints of STAR-RIS, and C7 is the constraint on the data rate for each user to receive the public message.

[0044] The Step 3 is specifically as follows:

[0045] Given the phase shift matrix Θ of the active STAR-RIS i,p , optimize the base station transmit beamforming matrix w and rate allocation vector a, use the SCA method to deal with non-convex constraints, and simplify the optimization problem as follows:

[0046]

[0047] Introducing the slack variable ψ i and η i , the optimization problem is reconstructed as:

[0048]

[0049] stC8:α i +log2(1+ψ i )≥R i

[0050]

[0051] C3, C4, C7

[0052] Where ψ=[ψ1,ψ2,...,ψ N ] and η=[η1,η2,...,η N], the objective function of the equivalent problem is convex, and equations C10, C11 and C3 make the problem as a whole non-convex, introducing a new non-negative slack variable α i , Formula C10 is reformulated as:

[0053]

[0054] In the formula By phase shifting the beamformed w i After any rotation of , the constraint can be equivalent to in Indicates plural The real part of the function is replaced by a first-order Taylor series. The formula can be restated as:

[0055]

[0056] The superscript (n-1) indicates the value of the variable at the (n-1)th iteration. Formula C11 can be reformulated as:

[0057]

[0058] The non-convexity in the formula is approximated by the difference of two convex functions DC, and the formula is approximated as:

[0059]

[0060] On the left The first-order Taylor series of C3 can be reformulated by using the DC approximation as:

[0061]

[0062] The non-convex problem is transformed into:

[0063]

[0064] stC4, C7, C8, C9, C12-C16

[0065]

[0066] It is concluded that the optimization problem is a convex problem, and the problem is solved using the CVX toolkit in Matlab.

[0067] The Step 4 is specifically as follows:

[0068] Step 4.1: Given the base station transmit beamforming vector w and rate allocation vector a, optimize Θ i,p, use FP changes to process the objective function and transform the problem into a standard QCQP quadratic constrained quadratic programming problem, let φ m represents the reflection or transmission coefficient of the mth unit of STAR-RIS, which is defined in complex form. φ represents the reflection or transmission coefficient vector of all units of STAR-RIS. Auxiliary variables are introduced. Using Lagrange dual transformation, the objective function in the formula is equivalently expressed as:

[0069]

[0070] The first term measures the logarithmic utility of the channel capacity, and the second term is the ratio of the channel gain to the noise. The optimal solution to obtain the elements in τ is:

[0071]

[0072] The formula can be restated as:

[0073]

[0074] stC1, C2, C5, C6

[0075] f2(φ,τ,υ) represents the objective function after optimization and adjustment of Lagrangian relaxation variables;

[0076] Step 4.2: Simplify the above problem. Since the above formula is non-convex, introduce A 1,c , A 1,p , B 1,p and B 1,c Auxiliary variables:

[0077]

[0078] According to the above formula, SINR and is equivalent to:

[0079]

[0080] make:

[0081]

[0082] A 2,p =A 1,p +C,

[0083]

[0084] B 2,p =B 1,p +D

[0085] Then the denominator of SINR is equivalent to:

[0086]

[0087] Step 4.3: Definition Then the constraints C1 and C2 are transformed into:

[0088]

[0089] After the above transformation, the optimization problem is transformed into:

[0090]

[0091] st(34),(35),C5,C6

[0092] in:

[0093]

[0094] Step 4.4: The optimal solutions of υ and τ can be obtained from the above respectively, and the optimal transmission and reflection coefficient matrices can be obtained using the CVX toolbox.

[0095] The Step 5 is specifically as follows:

[0096] Step 5.1: Initialize noise, path loss parameters, simulate channel information, coordinate positions of base stations, STAR-RIS, and users, number of iterations T, maximum number of iterations K, and set convergence accuracy ε;

[0097] Step 5.2: Obtain the optimal solution of the rate and the base station transmit beamforming matrix and the STAR-RIS phase shift matrix by solving w (t) , Θ (t) ;

[0098] Step 5.3: Until Stop iteration.

[0099] The beneficial effect of the present invention is that compared with the traditional RIS-assisted RSMA scheme and the STAR-RIS-assisted NOMA scheme, the present invention can significantly improve the communication performance of the system and provide an effective technical reference for improving the communication performance of the wireless communication system. BRIEF DESCRIPTION OF THE DRAWINGS

[0100] Figure 1 This is a model diagram of the STAR-RIS assisted RSMA communication system constructed by the present invention;

[0101] Figure 2 It is a graph showing the change of the sum rate with the number of iterations of the present invention;

[0102] Figure 3 It is a graph showing the variation of the sum rate of each solution of the present invention with the number of STAR-RIS elements;

[0103] Figure 4 It is a graph showing the variation of the sum rate of each solution of the present invention with the total power of the system;

[0104] Figure 5 This is a diagram showing how the sum rate of each solution of the present invention changes with the number of antennas. DETAILED DESCRIPTION

[0105] The present invention will be further described below in conjunction with the accompanying drawings and specific implementation methods.

[0106] Example 1: Figure 1 As shown in the figure, a communication system model of STAR-RIS assisted RSMA is constructed, and a downlink multiple-input single-output (MISO) communication system model is considered. The system consists of a base station, STAR-RIS and multiple users. Assume that the base station (BS) consists of N t The STAR-RIS system is composed of M uniform linearly arranged antennas, and the STAR-RIS system is composed of M uniform planar arrangement (UPA) reflection units, and each user has a single antenna. A STAR-RIS-assisted rate division multiple access system and rate optimization method, the specific steps are:

[0107] Step 1: Establish a STAR-RIS-assisted RSMA downlink communication system model;

[0108] Step 2: Based on the established communication system model, the base station beamforming matrix, the STAR-RIS phase shift matrix and the rate allocation vector are taken as the optimization problem, and the problem statement of the optimization target is constructed with the goal of maximizing the sum rate;

[0109] Step 3: Given the STAR-RIS phase shift matrix, optimize the base station transmit beamforming matrix and rate allocation vector;

[0110] Step 4: Given the base station transmit beamforming matrix and rate allocation vector, optimize the STAR-RIS phase shift matrix;

[0111] Step 5: Jointly optimize the base station transmit beamforming matrix, rate allocation vector, and STAR-RIS phase shift matrix to maximize the system and rate.

[0112] Specifically, all channels in the communication system adopt the Rice fading channel model, assuming that the channel gain from BS to STAR-RIS is expressed as The channel gain from STAR-RIS to the user is expressed as

[0113]

[0114] In the formula, ω B-S and is the Ricean factor of the BS-STAR-RIS link and the STAR-RIS-user link, G LoS and r i LoS represents the Rayleigh distribution line-of-sight component, G NLoS and r i NLoS Represents the non-line-of-sight component.

[0115] Assuming that the BS, STAR-RIS and users can obtain ideal channel state information, and the BS antenna adopts a uniform linear array, STAR-RIS uses an energy splitting protocol to divide the coverage area into a transmission area and a reflection area, and the transmission and reflection coefficients are and According to the principle of conservation of energy, The refraction and reflection coefficient matrices are:

[0116]

[0117] Indicates whether the user is a transmission area user (t) or a reflection area user (r). The signal sent by the BS is

[0118]

[0119] Among them, s0 represents a public message. Represents private messages for each user.

[0120] In the simulation system scenario, the distance from the base station to STAR-RIS is set to 100m, and the distance from the user to STAR-RIS is set to 50m. The number of base station antennas is set to N = 5, the user is a single antenna, and the number of STAR-RIS elements is set to 8. The noise power is set to 90dBm, and the convergence accuracy ε = 10 -3 .

[0121] In order to verify the performance difference of STAR-RIS assisted RSMA under different conditions, the number of BS antennas is fixed to 2, and the number of STAR-RIS elements and the total power of BS are changed. The results are shown in Figure 2 As shown. Figure 2 It can be seen that with the increase in the number of components and the increase in power, the algorithm can converge within the 8th iteration, which has good convergence performance. At the same time, it can be observed that in this algorithm, the convergence speed of the curve when the number of STAR-RIS components is 8 and the power is 30dBm is not as fast as when the number of STAR-RIS components is 4. When the number of components is small, the convergence rate is faster than when the number of components is large.

[0122] From the BS-STAR-RIS-user channel, we can see that as the number of STAR-RIS elements increases, the number of selectable phase adjustment schemes also increases. In order to illustrate the impact of the transmission and reflection coefficient matrix on the channel gain, assuming that the base station transmit power is 20dBm, the sum rate of different schemes under different numbers of elements is compared. The results are shown in the figure. Figure 3 As shown. It can be seen that as the number of components increases, the sum rates of various schemes show an upward trend. Specifically, when the number of STAR-RIS components increases from 4 to 64, the total rate of the STAR-RIS-assisted RSMA system increases from 5.4bps / Hz to 11.4bps / Hz, an increase of 111%. In addition, compared with OMA and NOMA, combining STAR-RIS with the RSMA system has a higher total sum rate gain. When the number of STAR-RIS components is small, it is still 32% higher than the total rate of NOMA, and much higher than OMA. When the number of STAR-RIS components is large, it is 28% higher than NOMA.

[0123] In order to demonstrate the power gain effect of the proposed method, the sum rate performance of each scheme under different base station transmit powers is compared. The results are as follows: Figure 4 As shown. Figure 4 It can be seen that with the increase of transmission power, the sum rate of each scheme has increased. Compared with the NOMA system, the scheme of the RSMA system can significantly improve the sum rate. It can be observed that with the increase of total power, the total rate of STAR-RIS is increased to 37.3% compared with the traditional RIS. This is because RSMA can enhance the channel capacity by decoding part of the multi-user interference and treating the rest of the interference as noise.

[0124] from Figure 5It can be seen that as the number of antennas increases, the spectrum efficiency of different schemes also increases. Specifically, the gain of the STAR-RIS-RSMA scheme is the most significant. When the number of antennas increases from 4 to 16, its spectrum efficiency increases from about 8bps / Hz to 12bps / Hz. The gains of the STAR-RIS-NOMA and traditional RIS-RSMA schemes also increase with the increase in the number of antennas, but not as significant as STAR-RIS-RSMA. In contrast, the gain of traditional RIS-NOMA is the smallest. It can be seen that the application of RSMA in STAR-RIS significantly improves the spectrum efficiency of the system, which is mainly due to its effective interference management and flexible message segmentation mechanism. RSMA enables users to better handle interference and improve transmission efficiency by segmenting transmission messages into public messages and private messages, rather than relying solely on the beamforming gain brought by the increase in the number of antennas. Compared with other schemes, STAR-RIS-RSMA takes advantage of RSMA and can flexibly manage interference when multiple users share resources, thereby achieving greater spectrum efficiency gains.

[0125] The specific implementation modes of the present invention are described in detail above in conjunction with the accompanying drawings, but the present invention is not limited to the above implementation modes, and various changes can be made within the knowledge scope of ordinary technicians in this field without departing from the purpose of the present invention.

Claims

1. A method for sum rate optimization of a STAR-RIS assisted rate division multiple access system, characterized in that: Step 1: Establish a STAR-RIS-assisted RSMA downlink communication system model; Step 2: Based on the established communication system model, the base station beamforming matrix, the STAR-RIS phase shift matrix and the rate allocation vector are taken as the optimization problem, and the problem statement of the optimization target is constructed with the goal of maximizing the sum rate; Step 3: Given the STAR-RIS phase shift matrix, optimize the base station transmit beamforming matrix and rate allocation vector; Step 4: Given the base station transmit beamforming matrix and rate allocation vector, optimize the STAR-RIS phase shift matrix; Step 5: Jointly optimize the base station transmit beamforming matrix, rate allocation vector, and STAR-RIS phase shift matrix to maximize the system and rate.

2. The method for sum rate optimization of a STAR-RIS assisted rate division multiple access system according to claim 1, characterized in that: The Step 1 is specifically as follows: Step 1.1: The communication system consists of base stations, STAR-RIS and users. Assume that the base station BS consists of N t The STAR-RIS consists of M passive element units arranged evenly in a plane, and there are N users with a single antenna. Step 1.2: All channels in the communication system adopt the Rice fading channel model. Assume that the channel gain from BS to STAR-RIS is expressed as The channel gain from STAR-RIS to the user is expressed as In the formula, ω B-S and is the Ricean factor of the BS-STAR-RIS link and the STAR-RIS-user link, G LoS and r i LoS represents the Rayleigh distribution line-of-sight component, G NLoS and r i NLoS represents the non-line-of-sight component; Step 1.3: Assuming that the BS, STAR-RIS and users can obtain ideal channel state information, and the BS antenna adopts a uniform linear array, STAR-RIS uses an energy splitting protocol to divide the coverage area into a transmission area and a reflection area. The transmission and reflection coefficients are and According to the principle of conservation of energy, The refraction and reflection coefficient matrices are: Indicates whether the user is a transmission zone user or a reflection zone user; Based on the basic idea of ​​RSMA, the signal sent by BS is: Among them, s0 represents a public message. Represents private messages for each user; Step 1.4: When the BS sends a signal to the user, the signal is transmitted or reflected by STAR-RIS to the user. i The signal y received by the user i for: Among them, BS and mth i The large-scale path loss between users is expressed as α represents the path loss component, d B-S and They represent the distance between BS and STAR-RIS and the distance between STAR-RIS and the mth i The distance between users, H represents the transpose of the matrix, n i ~CN(0,σ 2 ) is expressed as the mth i The additive Gaussian white noise of the user, Θ i,p For m i The user's refraction and reflection coefficient matrices; User m i The signal-to-interference-noise ratios of the public and private signals at are: Step 1.5: User m i The achievable rates for decoding the public stream and the expected private stream are expressed as: c i =log2(1+γ c,i ) p i =log2(1+γ p,i ), The rate of public messages is chosen to be A given public message min i∈N c i and assigned to user m i The rate a i , the data rate constraint for each user to receive public messages is: The SIC operation is implemented at the receiving end, and the transmission power of each user satisfies the constraint: Where θ is the minimum difference between the decoded signal power and the undecoded inter-user interference signal power plus the noise power, and the system sum rate is the sum of the public rate and the achievable private flow rate, expressed as: Among them, R i is the sum rate of the system.

3. The method for sum rate optimization of a STAR-RIS assisted rate division multiple access system according to claim 1, characterized in that: The Step 2 is specifically as follows: The beamforming matrix w and STAR-RIS phase shift matrix Θ at the BS in the joint optimization system network are i,p And the rate allocation vector, to maximize the system and rate, so the following optimization problem is constructed: in, R min represents the minimum rate required by all users, w = [w0; w1; ...; w N ] is represented as the beamforming matrix at the BS, a=[a1,a2,...,a N ] T It is represented as rate allocation vector constraint; C1 represents the minimum rate limit for all users, C2 represents ensuring that each user can decode the public information, C3 represents the SIC power constraint, P is the maximum transmit power of the base station, C4 represents the maximum transmit power constraint of the BS, C5 and C6 represent the refraction and reflection power constraints of STAR-RIS, and C7 is the constraint on the data rate for each user to receive the public message.

4. The method for sum rate optimization of a STAR-RIS assisted rate division multiple access system according to claim 1, characterized in that: The Step 3 is specifically as follows: Given the phase shift matrix Θ of the active STAR-RIS i,p , optimize the base station transmit beamforming matrix w and rate allocation vector a, use the SCA method to deal with non-convex constraints, and simplify the optimization problem as follows: Introducing the slack variable ψ i and η i , the optimization problem is reconstructed as: Where ψ=[ψ1,ψ2,...,ψ N ] and η=[η1,η2,...,η N ], the objective function of the equivalent problem is convex, and equations C10, C11 and C3 make the problem as a whole non-convex, introducing a new non-negative slack variable α i , Formula C10 is reformulated as: In the formula By phase shifting the beamformed w i After any rotation of , the constraint can be equivalent to in Indicates plural The real part of the function is replaced by a first-order Taylor series. The formula can be restated as: The superscript (n-1) indicates the value of the variable at the (n-1)th iteration. Formula C11 can be reformulated as: The difference between two convex functions DC is used to approximate the non-convexity in the formula, and the formula is approximated as: On the left The first-order Taylor series of C3 can be reformulated by using the DC approximation as: The non-convex problem is transformed into: stC4, C7, C8, C9, C12-C16 It is concluded that the optimization problem is a convex problem, and the problem is solved using the CVX toolkit in Matlab.

5. The method for sum rate optimization of a STAR-RIS assisted rate division multiple access system according to claim 1, characterized in that: The Step 4 is specifically as follows: Step 4.1: Given the base station transmit beamforming vector w and rate allocation vector a, optimize Θ i,p , use FP changes to process the objective function and transform the problem into a standard QCQP quadratic constrained quadratic programming problem, let φ m represents the reflection or transmission coefficient of the mth unit of STAR-RIS, which is defined in complex form. φ represents the reflection or transmission coefficient vector of all units of STAR-RIS. Auxiliary variables are introduced. Using Lagrange dual transformation, the objective function in the formula is equivalently expressed as: The first term measures the logarithmic utility of the channel capacity, and the second term is the ratio of the channel gain to the noise. The optimal solution to obtain the elements in τ is: The formula can be restated as: stC1, C2, C5, C6 f2(φ,τ,υ) represents the objective function after optimization and adjustment of Lagrangian relaxation variables; Step 4.2: Simplify the above problem. Since the above formula is non-convex, introduce A 1,c , A 1,p , B 1,p and B 1,c Auxiliary variables: According to the above formula, SINR and is equivalent to: make: A 2,p =A 1,p +C, B 2,p =B 1,p +D Then the denominator of SINR is equivalent to: Step 4.3: Definition Then the constraints C1 and C2 are transformed into: After the above transformation, the optimization problem is transformed into: st(34),(35),C5,C6 in: Step 4.4: The optimal solutions of υ and τ can be obtained from the above respectively, and the optimal transmission and reflection coefficient matrices can be obtained using the CVX toolbox.

6. The method for sum rate optimization of a STAR-RIS assisted rate division multiple access system according to claim 1, characterized in that: The Step 5 is specifically as follows: Step 5.1: Initialize noise, path loss parameters, simulate channel information, coordinate positions of base stations, STAR-RIS, and users, number of iterations T, maximum number of iterations K, and set convergence accuracy ε; Step 5.2: Obtain the optimal solution of the rate and the base station transmit beamforming matrix and the STAR-RIS phase shift matrix by solving w (t) , Θ (t) ; Step 5.3: Until Stop iteration.

Images