Ris-assisted swipt-isac system security beamforming method

Abstract

The application discloses a kind of RIS auxiliary-based SWIPT-ISAC system security beam shaping methods, it is communicated with full-duplex legitimate user by RIS auxiliary base station and perception target, and full-duplex legitimate user uses SWIPT technology to shunt signal, collects energy and emits artificial noise to resist eavesdropping;The optimization objective of the optimization problem constructed is to minimize base station transmitting power, constraint condition includes the communication and energy collection efficiency of user and the eavesdropping rate of potential eavesdropper, and radar signal-to-noise ratio, by transforming constraint and alternating iterative optimization algorithm, the optimization of beam shaping matrix is realized, and communication security is enhanced;The advantage is that user uses the energy collected to produce artificial noise to confuse potential eavesdropper, and can use self-interference cancellation technology to eliminate the interference of own signal, to improve communication signal-to-noise ratio, solve the physical layer security problem in system.

Description

TECHNICAL FIELD

[0001] The present application relates to a reconfigurable intelligent surface (RIS) assisted wireless simultaneous wireless information and power transfer (SWIPT) and integrated sensing and communication (ISAC) combined system, in particular to a RIS assisted SWIPT-ISAC system security beamforming method, which aims at the physical layer security problem that the target to be sensed in the system will also eavesdrop the information of the full-duplex legitimate user, and takes the minimization of the base station's transmit power as the target, the full-duplex legitimate user uses the SWIPT technology to collect energy and transmits artificial noise signal to interfere potential eavesdroppers, under the constraints of the communication achievable rate of the full-duplex legitimate user, the energy collected by the full-duplex legitimate user, the eavesdropping rate of the potential eavesdropper, and the corresponding radar signal-to-noise ratio of the radar echo signal received by the base station, the active beamforming matrix of the base station and the passive beamforming matrix of the RIS are jointly optimized to realize the dual task of user communication and target sensing, and ensure the security of the system. BACKGROUND

[0002] The sixth generation (6G) wireless communication network needs a large number of low-power communication devices to sense the environment and transmit massive data to support the application of new Internet of Things technologies such as smart cities and smart homes, but the explosive growth of communication devices also brings the problems of reduced hardware utilization efficiency and spectrum resource shortage. ISAC technology is an innovative technology that integrates wireless communication and target sensing functions together, and is one of the important scenarios of 6G wireless communication. ISAC specifically refers to sharing software and hardware resources in the same system to provide high-quality communication and high-precision sensing functions, reducing costs and improving system performance, where the communication function refers to the traditional transmission of data information, and the sensing function includes ranging, speed measurement, angle measurement, imaging, detection, etc. Through the fusion in time domain, space domain and frequency domain, sensing and communication gradually develop from coexistence and cooperation to complete integration, and truly realize the overall improvement of system performance, providing more potential applications for the industrial upgrading of 6G.

[0003] In addition to solving the problem of spectrum resource shortage, green, economic and security are also the basic requirements of the next generation of wireless communication technology. SWIPT technology can obtain energy and information from radio frequency signals at the same time, thereby prolonging the service life of communication equipment, and is considered as a key technology to solve these requirements. However, in the non-line-of-sight environment with shielding, neither the energy beam nor the information beam can directly point to the target, which is considered as the performance bottleneck of the SWIPT and ISAC system in the actual scene. RIS technology has the potential to solve this challenge, and as an effective solution to improve the coverage of wireless communication networks, it has received extensive attention. RIS is usually composed of many low-cost passive reflection elements, each of which can independently control the amplitude and phase of the reflected signal without consuming energy radio frequency chain and power amplifier. By introducing RIS into the SWIPT and ISAC system, the reflected signal can be superimposed with the signal of other paths to enhance the signal receiving power and improve the overall performance of the SWIPT and ISAC system.

[0004] Due to the broadcast nature of wireless communication signals, the target to be perceived may steal the information of the legitimate system, and the security of the legitimate system will be greatly threatened, so the physical layer security problem of the RIS-assisted SWIPT and ISAC system is also worth attention. In the traditional relay-assisted wireless communication network, using the relay to generate artificial noise to interfere with the eavesdropper is an optional solution to improve the physical layer security performance of the system, however, RIS as a passive device cannot generate artificial noise, and the traditional relay artificial noise generation scheme is not applicable. In addition, some studies also use base stations or friendly jammers to generate artificial noise to interfere with eavesdroppers, but such artificial noise will actually interfere with the legitimate users and reduce the signal-to-noise ratio of the legitimate users, which cannot meet the higher communication rate requirement. SUMMARY

[0005] The technical problem to be solved by the present application is to provide a RIS-assisted SWIPT-ISAC system security beamforming method in the case where the target to be perceived is a potential eavesdropper, which splits the received signal into two parts by using the SWIPT technology of the full-duplex legitimate user, collects energy by using the energy part, and then generates an artificial noise signal to confuse the potential eavesdropper, while the full-duplex legitimate user can eliminate the interference of its own signal by using the self-interference cancellation technology, thereby improving the signal-to-noise ratio of communication and solving the physical layer security problem in the system.

[0006] The technical solution adopted by the present application to solve the above technical problem is: a RIS-assisted SWIPT-ISAC system security beamforming method, characterized by comprising the following steps:

[0007] Step 1: constructing a RIS-aided SWIPT-ISAC system model, in which there is a base station with the dual functions of sensing and communication, a RIS, a full-duplex legitimate user, and a target which acts as a potential eavesdropper; the base station communicates with the downlink full-duplex legitimate user and senses the target with the aid of the RIS; the full-duplex legitimate user splits its received signal into two parts by using the SWIPT technology, one part is the information part for information decoding, and the other part is the energy part for energy collection; the full-duplex legitimate user collects energy according to the energy part of its received signal, and then transmits artificial noise signals to destroy the eavesdropping of the potential eavesdropper;

[0008] Step 2: establishing the communication performance index and the sensing performance index of the RIS-aided SWIPT-ISAC system model; the communication performance index includes the achievable rate of the full-duplex legitimate user, the energy collected by the full-duplex legitimate user, and the eavesdropping rate of the potential eavesdropper, wherein the achievable rate of the full-duplex legitimate user is obtained according to the information part of the received signal of the full-duplex legitimate user, and the energy collected by the full-duplex legitimate user is obtained according to the energy part of the received signal of the full-duplex legitimate user; the sensing performance index includes the radar signal-to-noise ratio corresponding to the radar echo signal received by the base station;

[0009] Step 3: taking the minimum of the transmit power of the base station as the target, taking the achievable rate of the full-duplex legitimate user, the energy collected by the full-duplex legitimate user, the eavesdropping rate of the potential eavesdropper, and the radar signal-to-noise ratio corresponding to the radar echo signal received by the base station as constraints, jointly optimizing the active beamforming matrix of the base station and the passive beamforming matrix of the RIS, and constructing an optimization problem;

[0010] Step 4: converting the constraint related to the achievable rate of the full-duplex legitimate user in the optimization problem into an easily handled form to obtain an easily handled optimization problem, wherein the converted constraint contains a linear decoding matrix and an introduced auxiliary matrix;

[0011] Step 5: splitting the easily handled optimization problem into two sub-problems, the first sub-problem is a sub-problem of optimizing the active beamforming matrix of the base station by fixing the linear decoding matrix, the introduced auxiliary matrix, and the passive beamforming matrix of the RIS, and the second sub-problem is a sub-problem of optimizing the passive beamforming matrix of the RIS by fixing the linear decoding matrix, the introduced auxiliary matrix, and the active beamforming matrix of the base station; then, the two sub-problems are iteratively solved by using an alternating iterative optimization algorithm to obtain the respective optimization solutions of the active beamforming matrix of the base station and the passive beamforming matrix of the RIS.

[0012] In the step 1, the base station is equipped with N t transmit antennas and N rThe root receives an antenna; the RIS is composed of M passive reflecting elements; the full-duplex legitimate user is equipped with N u The root antenna.

[0013] In step 2, the communication reachable rate of the full-duplex legitimate user and the energy acquisition process collected by the full-duplex legitimate user are:

[0014] Step 2.a1: the received signal y u is expressed as wherein, H (s) represents the channel from the base station to the full-duplex legitimate user, H (s) represents the channel from the RIS to the full-duplex legitimate user, and Φ represents the passive beamforming matrix of the RIS, diag(·) is used to extract or construct the diagonal elements of a matrix, e is a natural constant, j is a virtual number representation form, m = 1, 2, …, M, and φ m represents the phase of the mth passive reflecting element of the RIS, φ m ∈ [0, 2π], F represents the channel from the base station to the RIS, x represents the transmitted signal of the base station, x = Ws, W represents the active beamforming matrix of the base station, s represents the data symbol vector, and satisfies H (s) represents the expectation operation, and the superscript "H" represents the conjugate transpose operation, I s represents the d s dimensional unit matrix, d SI represents the number of data streams, n SI represents the artificial noise interference remaining after the full-duplex legitimate user uses the self-interference cancellation technology, whose mean is 0 and whose covariance matrix is I Nu represents the N u dimensional unit matrix, n u represents the received noise at the full-duplex legitimate user, whose mean is 0 and whose covariance matrix is

[0015] Step 2.a2: the full-duplex legitimate user uses the SWIPT technology to split y u into an information part y ID and an energy part y EH , wherein, ρ represents the power allocation ratio of the SWIPT technology, n ID represents the process noise generated during information decoding, whose mean is 0 and whose covariance matrix is

[0016] Step 2.a3: according to y ID , the signal-to-interference-and-noise ratio (SINR) of the full-duplex legitimate user receiving information is obtainedu , Further, the communication reachable rate R of the full-duplex legitimate user is obtained u , wherein “||” is a modulus operator, I represents a unit matrix,

[0017] Step 2.a4: According to y EH , the energy E collected by the full-duplex legitimate user is obtained u ,

[0018] In the step 2, the obtaining process of the eavesdropping rate of the potential eavesdropper is as follows:

[0019] Step 2.b1: The received signal y e of the potential eavesdropper is expressed as wherein h t represents a channel from the base station to the potential eavesdropper, h e represents a channel from the full-duplex legitimate user to the potential eavesdropper, z represents an artificial noise signal transmitted by the full-duplex legitimate user using the collected energy, and n e represents the received noise at the potential eavesdropper, which has a mean of 0 and a variance of

[0020] Step 2.b2: According to y e , the signal-to-interference-and-noise ratio (SINR) of the eavesdropping of the potential eavesdropper is obtained e , Further, the eavesdropping rate R of the potential eavesdropper is obtained e , R e = log2(1 + SINR e ).

[0021] In the step 2, the obtaining process of the radar signal-to-noise ratio (SNR) corresponding to the radar echo signal received by the base station is as follows:

[0022] Step 2.c1: The received signal y r of the radar echo signal received by the base station is expressed as wherein a represents a complex reflection coefficient of the target, n r represents the received noise in the radar echo signal received by the base station, which has a mean of 0 and a covariance matrix of represents an N r -order unit matrix, θ0 represents an azimuth angle of the base station to the target, a t (θ0) represents a transmission steering vector at the azimuth angle θ0, a r (θ0) represents a reception steering vector at the azimuth angle θ0, the superscript “T” represents a transpose operation;

[0023] Step 2.c2: y r through a receive filter or receive beamformer where,

[0024] Step 2.c3: solve the optimization problem to obtain the optimal solution

[0025] Step 2.c4: obtain the radar signal-to-noise ratio (SNR) corresponding to the radar echo signal received by the base station according to r , wherein Tr(·) represents the trace of a matrix.

[0026] In the step 3, the optimization problem is described as:

[0027]

[0028] s.t.C1: R u ≥ γ u ,

[0029] C2: R e ≤ γ e ,

[0030] C3: E u ≥ E min ,

[0031] C4: SNR r ≥ γ r ,

[0032] C5: |φ m | = 1, m = 1,..., M.

[0033] wherein Tr(WW H ) represents the transmit power of the base station, γ u represents the lower limit threshold of the achievable rate, γ e represents the upper limit threshold of the eavesdropping rate, E min represents the lower limit threshold of the energy, γ r represents the lower limit threshold of the radar signal-to-noise ratio.

[0034] In the step 4, the process for obtaining the easy-to-handle optimization problem is:

[0035] The information estimated by the full-duplex legitimate user is represented as ​​​​Where U represents the linear decoding matrix; then the information of full-duplex legal user estimation is calculated The mean square error with the data symbol vector s Then obtain the communication achievable rate R of full-duplex legal users u The conversion expression is described as: Where V represents the auxiliary matrix introduced. Then, the constraint C1 in the optimization problem in step 3 is transformed into a form that is easy to handle as constraint C1'. The constraint C1' is described as: Finally, we obtain a tractable optimization problem, which is described as:

[0036] In step 5, the process of obtaining the first sub-problem is:

[0037] Step 5.a1: When U, V, and Φ are considered constants, the first subproblem is initially described as:

[0038]

[0039] Step 5.a2: In the initial description of the first subproblem, for constraint C1', change The optimal solution Substitution In, get Update to get constraint C1", where Convert constraint C2 to constraint C2', which is described as: Using the first-order Taylor expansion and SCA method, constraint C3 is transformed into constraint C3', which is described as: Convert constraint C4 to constraint C4', which is described as: Among them, Re{·} means taking the real part, W0 represents the value of W obtained in the previous iteration during the internal iteration process when using the first-order Taylor expansion and SCA method;

[0040] Step 5.a3: The final description of the first subproblem is:

[0041] In step 5, the process of obtaining the second sub-problem is:

[0042] Step 5.b1: When U, V, and W are considered constants, the second subproblem can be preliminarily described as follows:

[0043]

[0044]

[0045] Step 5.b2: In the initial description of the second subproblem, ignore the objective function Without the term containing variable Φ, the objective function is transformed into Then is substituted to obtain

[0046] and where, C = FWW H F H , Const1 and Const2 are constant terms; without the term containing variable Φ, the objective function is transformed into Tr(Φ H BΦC) + Tr(Φ H P H ) + Tr(ΦP), where, After that, the objective function is finally transformed into where φ represents the phase vector of RIS, φ = [φ1, …, φ M ] T “⊙” is Hadamard product operator, and superscript “*” represents conjugate operation, [P] 1,1 represents the element in the 1st row and 1st column of P, [P] M,M represents the element in the Mth row and Mth column of P;

[0047] In the preliminary description of the 2nd sub-problem, the constraint C3 is changed to which is further transformed into φ H (B⊙C T )φ + 2Re{φ H h *} ≥ E min / (1-ρ) - Tr(G r WW H ), where, Then, φ H (B⊙C T )φ + 2Re{φ H h *} ≥ E min / (1-ρ) - Tr(G2WW H ) is transformed into the constraint C3” by using the first-order Taylor expansion and SCA method, which is described as: where, represents the value of φ obtained in the previous iteration in the internal iteration process when the first-order Taylor expansion and SCA method are used;

[0048] Step 5.b3: the preliminary description of the 2nd sub-problem is transformed into

[0049] Step 5.b4: in In the step 5, the alternating direction multiplier method is adopted, and an auxiliary variable is introduced The constraint C5 is converted into constraint C5', which is described as: |φ m ≤ 1, m = 1, …, M, and constraint C6 is added: and constraint C7:

[0050] Step 5.b5: converting into

[0051] Step 5.b6: converting into an augmented Lagrangian function form as the final description of the second sub-problem, wherein η represents a preset penalty parameter, η ≥ 0, and μ represents a dual variable.

[0052] In the step 5, the alternating iterative optimization algorithm is adopted to iteratively solve the two sub-problems, and the process of obtaining the optimization solutions of the active beamforming matrix of the base station and the passive beamforming matrix of the RIS is as follows:

[0053] Step 5.c1: let l represent the iteration number, and let l max represent the maximum iteration number; initialize W and Φ;

[0054] Step 5.c2: given the values of W and Φ, let the first-order partial derivative of be zero, to obtain the optimal solution U opt of U Then, the U opt is substituted into to obtain the optimal solution of and further obtain the optimal solution V opt of V wherein the values of W and Φ given when l = 1 are initialization values, and the values of W and Φ given when l > 1 are the values of W and Φ obtained in the (l-1)th iteration;

[0055] Step 5.c3: given the values of U, V and Φ, a convex optimization solver is used to solve the final description of the first sub-problem to obtain the value of W in the lth iteration; wherein the values of U and V given correspond to the U opt and V opt obtained in the lth iteration, and the value of Φ given when l = 1 is an initialization value, and the value of Φ given when l > 1 is the value of Φ obtained in the (l-1)th iteration;

[0056] Step 5.c4: given the values of U, V and W, a convex optimization solver is used to solve the final description of the second sub-problem to obtain the value of Φ in the lth iteration; wherein the values of U and V given correspond to the Uopt and V opt the value of W given when l = 1 is the initial value, and the value of W given when l > 1 is the value of W obtained in the (l-1)th iteration;

[0057] Step 5.c5: judge whether l is less than or equal to l max If yes, let l = l + 1, and then return to step 5.c2 for continuous execution, otherwise, take the value of W obtained in the lth iteration as the optimized solution of the active beamforming matrix of the base station, and take the value of Φ obtained in the lth iteration as the optimized solution of the passive beamforming matrix of the RIS, where “=” in l = l + 1 is an assignment symbol. max max

[0058] Compared with the prior art, the present application has the following advantages:

[0059] 1) The method of the present application jointly designs the active beamforming matrix of the base station and the passive beamforming matrix of the RIS, so that the system can achieve high-speed communication and high-precision perception functions while minimizing the transmission power, meeting the green and economic needs of future wireless communication systems. Compared with the traditional method, the present application constructs an RIS-assisted SWIPT-ISAC system model, effectively solving the problems of blocked communication link and limited communication and perception integrated signal coverage range in practical application scenarios, and significantly improving the overall performance and reliability of the system.

[0060] 2) The method of the present application considers the communication and perception integrated application scenario when the target is a potential eavesdropper, in which the target will eavesdrop the information sent by the base station to the full-duplex legitimate user, threatening the legitimate communication. Based on this, the present application ingeniously uses the SWIPT technology, i.e. the full-duplex legitimate user splits the received signal into an information part and an energy part, and simultaneously sends artificial noise to interfere with the potential eavesdropper, reducing the eavesdropping rate. The advantage of this is that the full-duplex legitimate user will use the self-interference cancellation technology to eliminate the interference of artificial noise on itself, thereby significantly improving the communication rate.

[0061] 3) The method of the present application can achieve good convergence effect, complete the tasks of secure communication and target perception, and minimize the power consumption of the system, meeting the safety and energy saving needs of the communication and perception integrated system. BRIEF DESCRIPTION OF DRAWINGS

[0062] Figure 1 is the overall implementation flowchart of the method of the present application;

[0063] Figure 2 is a simple schematic diagram of the RIS-assisted SWIPT-ISAC system model;

[0064] ​​Figure 3 Fig. 1 is a schematic diagram of convergence curves of the method of the present application under different numbers of passive reflecting elements of RIS and different numbers of transmitting and receiving antennas of base station;

[0065] Figure 4 Fig. 4 is a schematic diagram of comparison curves of changes of transmitting power of base station under different numbers of transmitting and receiving antennas of base station for the method of the present application and two reference methods. DETAILED DESCRIPTION

[0066] The present application will be further described in detail below with reference to the embodiments of the drawings.

[0067] The present application proposes a RIS-aided SWIPT-ISAC system security beamforming method, which, in the case of a potential eavesdropper, uses SWIPT technology to split the received signal into two parts by the full-duplex legitimate user, collects energy according to the energy part, and transmits artificial noise signals using the collected energy to destroy the eavesdropping of the potential eavesdropper; it jointly optimizes two key indicators: the active beamforming matrix of the base station and the passive beamforming matrix of the RIS, to ensure that the base station can communicate securely with the downlink full-duplex legitimate user and complete the target sensing task, realizing dual functions; the initial optimization problem established is non-convex, which is very difficult to handle, so the minimum mean square error (MMSE) method, as well as the convex optimization algorithms of alternating iterative optimization, successive convex approximation, and alternating direction multiplier, are used to solve the two key indicators.

[0068] The present application proposes a RIS-aided SWIPT-ISAC system security beamforming method, which, in the case of a potential eavesdropper, uses SWIPT technology to split the received signal into two parts by the full-duplex legitimate user, collects energy according to the energy part, and transmits artificial noise signals using the collected energy to destroy the eavesdropping of the potential eavesdropper; it jointly optimizes two key indicators: the active beamforming matrix of the base station and the passive beamforming matrix of the RIS, to ensure that the base station can communicate securely with the downlink full-duplex legitimate user and complete the target sensing task, realizing dual functions; the initial optimization problem established is non-convex, which is very difficult to handle, so the minimum mean square error (MMSE) method, as well as the convex optimization algorithms of alternating iterative optimization, successive convex approximation, and alternating direction multiplier, are used to solve the two key indicators. Figure 1 As shown in Fig. 1, it includes the following steps:

[0069] Step 1: Construct an RIS-aided SWIPT-ISAC system model, as shown in Fig. 2. Figure 2As shown in the figure, the model consists of a base station (ISAC-BS) with dual sensing and communication functions, a RIS, a full-duplex legitimate user (User), and a target (Target), which also acts as a potential eavesdropper (Eve). The base station communicates with the downlink full-duplex legitimate user with the assistance of the RIS and simultaneously senses the target. The base station's transmitted signal reaches the full-duplex legitimate user via a direct link and a RIS reflection link. The full-duplex legitimate user uses SWIPT technology to split its received signal into two parts: one for information decoding and the other for energy harvesting. The full-duplex legitimate user collects energy based on the energy portion of its received signal and then uses the collected energy to emit artificial noise signals to defeat the potential eavesdropper's eavesdropping. The benefit of this is that the full-duplex legitimate user can use self-interference cancellation technology to eliminate the interference of artificial noise on itself, thereby improving the signal-to-interference-noise ratio and communication rate of legitimate communication.

[0070] Specifically, the base station is equipped with N t Transmitting antennas and N r RIS consists of M passive reflective elements; full-duplex legal users are equipped with N u In this embodiment, N t =16, N r =16, M=30, N u =2.

[0071] Step 2: Establish the communication performance indicators and perception performance indicators of the RIS-assisted SWIPT-ISAC system model. The communication performance indicators include the communication achievable rate of full-duplex legitimate users, the energy collected by full-duplex legitimate users, and the eavesdropping rate of potential eavesdroppers. The communication achievable rate of full-duplex legitimate users is obtained based on the information portion of the full-duplex legitimate users' received signals, and the energy collected by full-duplex legitimate users is obtained based on the energy portion of the full-duplex legitimate users' received signals. The perception performance indicator includes the radar signal-to-noise ratio (SNR) of the radar echo signal received by the base station.

[0072] As a preference, in step 2, the process of obtaining the communication achievable rate of full-duplex legal users and the energy collected by full-duplex legal users is as follows:

[0073] Step 2.a1: Transform the received signal y of the full-duplex legal user into u Expressed as in, Indicates the channel from the base station to full-duplex legal users, The dimension is N u ×N t , Indicates the channel from RIS to full-duplex legal users. The dimension is N uX M, Φ represents the passive beamforming matrix of the RIS, For the definition of symbols, diag(·) is used to extract or construct the diagonal elements of a matrix, e is a natural constant, e = 2.7…, j is a virtual number representation form, m = 1, 2, …, M, φ m represents the phase of the m-th passive reflecting element of the RIS, φ m ∈ [0, 2π], F represents the channel from the base station to the RIS, the dimension of F is M × N t , x represents the transmit signal of the base station, x = Ws, W represents the active beamforming matrix of the base station, the dimension of W is N t × d s , satisfies 1 ≤ d s ≤ min{N t ,N u}, min represents the minimum value function, s represents the data symbol vector, satisfies represents the expectation operation, the superscript “H” represents the conjugate transpose operation, the dimension of s is d s × 1, I ds represents the unit matrix of d s order, d s represents the number of data streams, n SI represents the artificial noise interference remaining after the full-duplex legitimate user utilizes the self-interference cancellation technology, the mean is 0 and the covariance matrix is represents the unit matrix of N u order, n u represents the received noise at the full-duplex legitimate user, the mean is 0 and the covariance matrix is In this embodiment

[0074] Step 2.a2: The full-duplex legitimate user splits y u into an information part y ID and an energy part y EH , where ρ represents the power allocation ratio of the SWIPT technology, n ID represents the process noise generated during information decoding, the mean is 0 and the covariance matrix is In this embodiment

[0075] Step 2.a3: According to y ID , the signal-to-interference-and-noise ratio (SINR) of the full-duplex legitimate user receiving information is obtained u , Further, the communication reachable rate R of the full-duplex legitimate user is obtained u , wherein “||” is a modulus operator, I represents a unit matrix,

[0076] Step 2.a4: according to y EH , the energy E collected by the full-duplex legitimate user is obtained u , In order to ensure the physical layer security of the system, the full-duplex legitimate user transmits an artificial noise signal z by using the collected energy, and therefore the power of the artificial noise signal z should satisfy

[0077] As a preferred, in step 2, the obtaining process of the eavesdropping rate of the potential eavesdropper is as follows:

[0078] Step 2.b1: the received signal y e of the potential eavesdropper is expressed as wherein h t represents the channel from the base station to the potential eavesdropper, h t has a dimension of N t ×1, h e represents the channel from the full-duplex legitimate user to the potential eavesdropper, h e has a dimension of N u ×1, z represents the artificial noise signal transmitted by the full-duplex legitimate user by using the collected energy, and n e represents the received noise at the potential eavesdropper, which has a mean of 0 and a variance of In the present embodiment

[0079] Step 2.b2: according to y e , the signal-to-interference-and-noise ratio (SINR) eavesdropped by the potential eavesdropper is obtained e , Further, the eavesdropping rate R of the potential eavesdropper is obtained e , R e = log2(1+SINR e ).

[0080] As a preferred, in step 2, the obtaining process of the radar signal-to-noise ratio corresponding to the radar echo signal received by the base station is as follows:

[0081] Step 2.c1: the target considered by the present application is an aerial target or a long-distance target far away from the full-duplex legitimate user, in which case the radar echo signal reflected by the RIS is extremely weak and has almost no contribution to the signal-to-noise ratio for detecting the target, and this part of signal can be ignored in the system optimization process. The radar echo signal y r received by the base station is expressed as where a denotes the complex reflection coefficient of the target, a depends on the radar cross section of the target and the path loss coefficient of the link from the base station to the target, and satisfies denotes the variance of a, n r denotes the received noise in the radar echo signal received by the base station, whose mean is 0 and whose covariance matrix is I Nr denotes the N r order identity matrix, in the embodiment denotes the azimuth angle from the base station to the target, in the embodiment t denotes the transmit steering vector at the azimuth angle a r denotes the receive steering vector at the azimuth angle the superscript "T" denotes the transpose operation.

[0082] Step 2.c2: In order to achieve satisfactory radar detection performance, make y r pass through a receive filter or receive beamformer get: wherein, the dimension of is N r x 1,

[0083] Step 2.c3: In order to maximize the radar signal-to-noise ratio, the minimum variance distortionless response (MVDR) beamforming algorithm is used to solve get the optimal solution of

[0084] Step 2.c4: According to obtain the corresponding radar signal-to-noise ratio SNR r of the radar echo signal received by the base station wherein, Tr(·) denotes the trace of the matrix.

[0085] Step 3: In order to minimize the transmit power of the base station, the communication achievable rate of the full-duplex legitimate user, the energy collected by the full-duplex legitimate user, the eavesdropping rate of the potential eavesdropper, and the corresponding radar signal-to-noise ratio of the radar echo signal received by the base station are used as constraints, and the active beamforming matrix of the base station and the passive beamforming matrix of the RIS are jointly optimized, and an optimization problem is constructed.

[0086] Further limited, the optimization problem is described as:

[0087] ​

[0088] s.t.C1:R u ≥γ u ,

[0089] C2:R e ≤γ e ,

[0090] C3:E u ≥E min ,

[0091] C4:SNR r ≥γ r ,

[0092] C5:φ m |=1,m=1,...,M.

[0093] where min is the min function, s.t. means “subject to”, Tr(WW H ) is the transmit power of the base station, γ u is the lower bound of the achievable rate threshold, γ e is the upper bound of the eavesdropping rate threshold, E min is the lower bound of the energy threshold, γ r is the lower bound of the radar signal-to-noise ratio threshold. In this embodiment, γ u = 5 bps / Hz, γ e = 1 bps / Hz, E min = 2 x 10 -4 , γ r = 15 dB. Constraint C1 means that the achievable rate R u of the full-duplex legitimate user is not less than γ u ; constraint C2 means that the eavesdropping rate R e of the potential eavesdropper is not greater than γ e ; constraint C3 means that the energy collected by the full-duplex legitimate user is not less than E min ; constraint C4 means that the radar signal-to-noise ratio SNR r of the radar echo signal received by the base station is not less than γ r ; and constraint C5 means that the modulus 1 constraint of the phase of each passive reflecting element of the RIS is satisfied.

[0094] Step 4: The optimization problem in step 3 is a non-convex variable-coupled problem, which is difficult to solve. First, the minimum mean square error (MMSE) method is used to simplify the constraint on the achievable rate of the full-duplex legitimate user in the optimization problem, and the constraint on the achievable rate of the full-duplex legitimate user in the optimization problem is converted into an easily handled form to obtain an easily handled optimization problem, wherein the converted constraint contains a linear decoding matrix and an introduced auxiliary matrix.

[0095] Further, in step 4, the obtaining process of the tractable optimization problem is: taking the information of the full-duplex legitimate user estimate as wherein U represents a linear decoding matrix, and the dimension of U is N u ×d s ; then calculating the mean square error of the full-duplex legitimate user estimate and the data symbol vector s Then, the communication reachable rate R u of the full-duplex legitimate user is obtained, and the conversion expression is described as: wherein V represents an introduced auxiliary matrix; and the constraint C1 in the optimization problem in step 3 is converted into a tractable form as a constraint C1', which is described as: Finally, the tractable optimization problem is obtained, which is described as:

[0096] Step 5: Since the tractable optimization problem is still non-convex, the tractable optimization problem is split into two sub-problems, the first sub-problem is a sub-problem of optimizing the active beamforming matrix of the base station while fixing the linear decoding matrix, the introduced auxiliary matrix and the passive beamforming matrix of the RIS, and the second sub-problem is a sub-problem of optimizing the passive beamforming matrix of the RIS while fixing the linear decoding matrix, the introduced auxiliary matrix and the active beamforming matrix of the base station; then an alternating iterative optimization algorithm is used to iteratively solve the two sub-problems to obtain the respective optimization solutions of the active beamforming matrix of the base station and the passive beamforming matrix of the RIS.

[0097] As a preferred, in step 5, the obtaining process of the first sub-problem is:

[0098] Step 5.a1: when U, V and Φ are regarded as constants, the first sub-problem is initially described as:

[0099]

[0100] Step 5.a2: in the initial description of the first sub-problem, for the constraint C1', the optimal solution of is substituted into to obtain , and the constraint C1” is updated, wherein The constraint C2 is converted into a constraint C2', which is described as: Due to the existence of non-convex constraints C3, C4, the preliminary description of the 1st sub-problem is still non-convex, thus the constraint C3 is transformed into constraint C3' by using the first-order Taylor expansion and SCA (Successive Convex Approximation) method, described as: The constraint C4 is transformed into constraint C4', described as: where Re{·} represents taking the real part, W0 represents the value of W obtained in the previous iteration in the internal iteration process when the first-order Taylor expansion and SCA method are used.

[0101] Step 5.a3: the final description of the 1st sub-problem is: The final description of the 1st sub-problem is convex in the objective function and constraints, which can be solved by a standard convex optimization solver.

[0102] As preferred, in step 5, the obtaining process of the 2nd sub-problem is:

[0103] Step 5.b1: when U, V and W are regarded as constants, the 2nd sub-problem is preliminarily described as: where max is the maximum function,

[0104]

[0105] Step 5.b2: in the preliminary description of the 2nd sub-problem, the term in the objective function which does not contain variable Φ is ignored, and the objective function is transformed into Then is substituted to obtain

[0106] and where C = FWW H F H , Const1 and Const2 are constant terms; the term which does not contain variable Φ is ignored, and the objective function is transformed into Tr(Φ H BΦC) + Tr(Φ H P H ) + Tr(ΦP), where After that, the objective function is finally transformed into where φ represents the phase vector of RIS, φ = [φ1, …, φ M ] T “⊙” is the Hadamard product operator, and the superscript “*” represents the conjugate operation, [P] 1,1 represents the element in the 1st row and 1st column of P, [P] M,MRepresents the element in row M and column M of P.

[0107] In the preliminary description of the second subproblem, constraint C3 is changed to Then converted into φ H (B⊙C T )φ+2Re{φ H h *}≥E min / (1-ρ)-Tr(G r WW H ),in, Then, the first-order Taylor expansion and SCA method are used to convert

[0108] φ H (B⊙C T )φ+2Re{φ H h *}≥E min / (1-ρ)-Tr(G2WW H ) is converted into constraint C3", which is described as: in, represents the value of φ obtained in the previous iteration during the internal iteration process when using the first-order Taylor expansion and SCA method.

[0109] Step 5.b3: Transform the initial description of the second subproblem into

[0110] Step 5.b4: In the algorithm, the Alternating Direction Method of Multipliers (ADMM) is used and auxiliary variables are introduced. Convert constraint C5 to constraint C5', which is described as: m |≤1,m=1,...,M, and add constraint C6: and constraint C7:

[0111] Step 5.b5: Convert to

[0112] Step 5.b6: Transformed into the augmented Lagrangian function form as the final description of the second subproblem, Among them, η represents the preset penalty parameter, η ≥ 0, μ represents the dual variable, and the final description of the second sub-problem is a multivariable optimization problem, which can be solved by iteratively updating each variable.

[0113] Preferably, in step 5, the alternating iterative optimization algorithm is used to iteratively solve the two sub-problems to obtain the optimized solutions of the active beamforming matrix of the base station and the passive beamforming matrix of the RIS respectively. The process is:

[0114] Step 5.c1: Let l denote the number of iterations, let l max Indicates the maximum number of iterations; initialize W and Φ.

[0115] Step 5.c2: Given the values ​​of W and Φ, let The first-order partial derivative of is zero, and the optimal solution U is obtained opt , Then U opt Substitution In, get The optimal solution Then we get the optimal solution V opt , The values ​​of W and Φ given when l=1 are initialization values, and the values ​​of W and Φ given when l>1 are the values ​​of W and Φ obtained in the l-1th iteration.

[0116] Step 5.c3: Given the values ​​of U, V, and Φ, use the convex optimization solver to solve the final description of the first subproblem and obtain the value of W at the lth iteration; where the given values ​​of U and V correspond to the U obtained at the lth iteration. opt and V opt When l=1, the value of Φ given is the initialization value, and when l>1, the value of Φ given is the value of Φ obtained in the l-1th iteration.

[0117] Step 5.c4: Given the values ​​of U, V, and W, use the convex optimization solver to solve the final description of the second subproblem and obtain the value of Φ at the lth iteration; where the given values ​​of U and V correspond to the U obtained at the lth iteration. opt and V opt , when l=1, the value of W given is the initialization value, and when l>1, the value of W given is the value of W obtained in the l-1th iteration.

[0118] Step 5.c5: Determine whether l is less than or equal to l max If yes, set l = l + 1, and then return to step 5.c2 to continue execution, otherwise, set the lth max The value of W obtained in the first iteration is used as the optimized solution of the active beamforming matrix of the base station. max The value of Φ obtained by the iteration is used as the optimized solution of the passive beamforming matrix of RIS, where the "=" in l=l+1 is an assignment symbol.

[0119] The above, F are modeled as a Rician channel model, which is determined by both large-scale fading and small-scale fading, where β denotes the path loss, which is modeled as C0 denotes the path loss at a reference distance d0, d denotes the path distance, ξ denotes the path loss exponent, κ is the Rician factor, and denotes the line-of-sight link from the base station to the full-duplex legitimate user, denotes the non-line-of-sight link from the base station to the full-duplex legitimate user, denotes the line-of-sight link from the RIS to the full-duplex legitimate user, denotes the non-line-of-sight link from the RIS to the full-duplex legitimate user, LoS denotes the line-of-sight link from the base station to the RIS, NLoS denotes the non-line-of-sight link from the base station to the RIS, F NLoS all follow a complex Gaussian distribution with mean 0 and variance 1.

[0120] The feasibility and effectiveness of the method of the present application are further illustrated through the following simulation.

[0121] Figure 3 The method of the present application is demonstrated in the number of passive reflecting elements M of the RIS and the number of transmitting and receiving antennas N = N t = N r of the base station. The convergence curves in different cases are shown in FIG. 2. Figure 3 It can be observed from FIG. 2 that the transmitting power of the base station reaches a good convergence effect after 20-30 iterations of the alternating iterative optimization algorithm. When the number of transmitting and receiving antennas N = N t = N r of the base station is fixed, the larger the number of passive reflecting elements M of the RIS is, the smaller the transmitting power required by the base station is, which also reflects the obvious role of the RIS in reducing the total power consumption of the system. Similarly, when the number of passive reflecting elements M of the RIS is fixed, the larger the number of transmitting and receiving antennas N = N t = N r of the base station is, the smaller the transmitting power required by the base station is, because the increase of the number of transmitting and receiving antennas of the base station improves the directivity of the signal and improves the transmission efficiency and signal-to-noise ratio, which can effectively reduce the total power consumption of the system while maintaining the same performance requirements.

[0122] Figure 4 The method of the present application and two benchmark methods are demonstrated in the change of the transmitting power of the base station with the change of the number of transmitting and receiving antennas N = N t = N r of the base station. The two benchmark methods are the method of random RIS phase shift and the method of not deploying RIS. From FIG. 3, it can be observed that the method of the present application has a smaller transmitting power than the two benchmark methods, which also reflects the obvious role of the RIS in reducing the total power consumption of the system.Figure 4 It can be clearly seen that the transmission power of the base station increases with the number of transmitting and receiving antennas of the base station N=N t =N r This is because as the number of transmitting and receiving antennas of the base station N=N t =N r The increase in the system can more effectively utilize spatial resources and improve the directionality of the signal, thereby reducing the transmission power of the base station. Within the range of antenna number (8 to 12), the transmission power of the base station obtained by the method of the present invention is lower than that of the two benchmark methods, which shows that the method of the present invention has obvious advantages in power saving. Although the random RIS phase shift method also uses RIS, the system power consumption is lower than the method without RIS deployment, but it does not fully play the role of RIS in improving system performance, and its power consumption is always higher than that of the method of the present invention. This also shows that the superiority of the method of the present invention lies in the optimized design of RIS phase shift, which can reduce system power consumption and improve energy efficiency in actual systems.

Claims

1. A RIS-assisted SWIPT-ISAC system security beamforming method, characterized by The following steps are involved: Step 1: Construct a RIS-assisted SWIPT-ISAC system model, which includes a base station with dual sensing and communication functions, a RIS, a full-duplex legitimate user, and a target, which also acts as a potential eavesdropper. With the assistance of RIS, the base station communicates with legitimate full-duplex users in the downlink and simultaneously senses the target. Legitimate full-duplex users use SWIPT technology to split their received signals into two parts: an information portion for information decoding and an energy portion for energy harvesting. Legitimate full-duplex users harvest energy from the energy portion of their received signals and then use the harvested energy to emit artificial noise signals to disrupt potential eavesdroppers. Step 2: Establish communication performance indicators and perception performance indicators for the RIS-assisted SWIPT-ISAC system model. The communication performance indicators include the achievable communication rate of full-duplex legitimate users, the energy collected by full-duplex legitimate users, and the eavesdropping rate of potential eavesdroppers. The achievable communication rate of full-duplex legitimate users is obtained based on the information portion of the full-duplex legitimate users' received signals, and the energy collected by full-duplex legitimate users is obtained based on the energy portion of the full-duplex legitimate users' received signals. The perception performance indicator includes the radar signal-to-noise ratio (SNR) of the radar echo signal received by the base station. Step 3: With the goal of minimizing the base station's transmit power, and with the achievable communication rate of full-duplex legitimate users, the energy collected by full-duplex legitimate users, the eavesdropping rate of potential eavesdroppers, and the radar signal-to-noise ratio of the radar echo signal received by the base station as constraints, the active beamforming matrix of the base station and the passive beamforming matrix of the RIS are jointly optimized to construct an optimization problem. Step 4: Transform the constraints on the achievable communication rate of full-duplex legitimate users in the optimization problem into a tractable form, thereby obtaining a tractable optimization problem. The transformed constraints include the linear decoding matrix and the introduced auxiliary matrix. Step 5: Split the tractable optimization problem into two subproblems. The first subproblem involves fixing the linear decoding matrix, the introduced auxiliary matrix, and the passive beamforming matrix of the RIS, and optimizing the active beamforming matrix of the base station. The second subproblem involves fixing the linear decoding matrix, the introduced auxiliary matrix, and the active beamforming matrix of the base station, and optimizing the passive beamforming matrix of the RIS. An alternating iterative optimization algorithm is then used to iteratively solve the two subproblems to obtain the optimal solutions for the active beamforming matrix of the base station and the passive beamforming matrix of the RIS.

2. The RIS-assisted SWIPT-ISAC system security beamforming method according to claim 1 is characterized in that In step 1, the base station is equipped with N t Transmitting antennas and N r RIS consists of M passive reflective elements; full-duplex legal users are equipped with N u Antenna.

3. The RIS-assisted SWIPT-ISAC system security beamforming method according to claim 2 is characterized in that In step 2, the communication achievable rate of full-duplex legal users and the energy collected by full-duplex legal users are obtained as follows: Step 2.a1: Transform the received signal y of the full-duplex legal user into u Expressed as in, Indicates the channel from the base station to full-duplex legal users, represents the channel from RIS to full-duplex legal users, Φ represents the passive beamforming matrix of RIS, diag(·) is used to extract or construct the diagonal elements of a matrix, where e is a natural constant, j is an imaginary number, m = 1, 2, …, M, and φ m represents the phase of the mth passive reflector element of RIS, φ m ∈[0,2π], F represents the channel from the base station to RIS, x represents the base station's transmitted signal, x=Ws, W represents the base station's active beamforming matrix, s represents the data symbol vector, satisfying represents the expectation operation, and the superscript "H" represents the conjugate transpose operation. Indicates d s The unit matrix of order d s Indicates the number of data streams, n SI It represents the artificial noise interference residual after the full-duplex legal user uses the self-interference cancellation technology, whose mean is 0 and the covariance matrix is Indicates N u The unit matrix of order, n u represents the received noise at the full-duplex legal user, whose mean is 0 and the covariance matrix is Step 2.a2: Full-duplex legitimate users use SWIPT technology to u Divided into information part y ID and the energy part y EH , Where ρ represents the power allocation ratio of SWIPT technology, n ID represents the process noise generated when decoding information, its mean is 0 and its covariance matrix is Step 2.a3: According to y ID , obtain the signal-to-interference-and-noise ratio (SINR) of the full-duplex legitimate user receiving information u , Then we can get the communication rate R of full-duplex legal users. u , Among them, "||" is the modulo operator, I represents the unit matrix, Step 2.a4: According to y EH , obtain the energy E collected by full-duplex legal users u , 4. The RIS-assisted SWIPT-ISAC system security beamforming method according to claim 3 is characterized in that In step 2, the process of obtaining the eavesdropping rate of the potential eavesdropper is as follows: Step 2.b1: Transform the potential eavesdropper’s received signal y e Expressed as Among them, h t represents the channel from the base station to the potential eavesdropper, h e represents the channel from the full-duplex legitimate user to the potential eavesdropper, z represents the artificial noise signal emitted by the full-duplex legitimate user using the collected energy, and n e represents the received noise at the potential eavesdropper, which has a mean of 0 and a variance of Step 2.b2: According to y e , obtain the signal-to-interference-and-noise ratio (SINR) of the potential eavesdropper e , Then we can get the potential eavesdropper’s eavesdropping rate R e , R e =log2(1+SINR e ).

5. The RIS-assisted SWIPT-ISAC system security beamforming method according to claim 4 is characterized in that In step 2, the process of obtaining the radar signal-to-noise ratio corresponding to the radar echo signal received by the base station is as follows: Step 2.c1: The radar echo signal y received by the base station r Expressed as Where α represents the complex reflection coefficient of the target, n r represents the receiving noise in the radar echo signal received by the base station, whose mean is 0 and the covariance matrix is Indicates N r The unit matrix of order, θ0 represents the azimuth angle from the base station to the target, a t (θ0) represents the launch steering vector at azimuth angle θ0, a r (θ0) represents the receiving steering vector at azimuth angle θ0, The superscript "T" indicates the transpose operation; Step 2.c2: Make y r Through a receive filter or receive beamformer get: in, Step 2.c3: Use the minimum variance distortion-free response beamforming algorithm to Solve it and get The optimal solution Step 2.c4: According to Get the radar signal-to-noise ratio (SNR) of the radar echo signal received by the base station r , Where Tr(·) represents the trace of the matrix.

6. The RIS-assisted SWIPT-ISAC system security beamforming method according to claim 5 is characterized in that In step 3, the optimization problem is described as: s.t.C1:R u ≥γ u , C2:R e ≤γ e , C3:E u ≥E min , C4:SNR r ≥γ r , C5:φ m |=1,m=1,...,M. Among them, Tr(WW H ) represents the base station’s transmit power, γ u represents the lower limit threshold of the achievable rate, γ e Indicates the upper threshold of the eavesdropping rate, E min represents the lower energy threshold, γ r Indicates the radar signal-to-noise ratio lower threshold.

7. The RIS-assisted SWIPT-ISAC system security beamforming method according to claim 6 is characterized in that In step 4, the process of obtaining a tractable optimization problem is as follows: Information estimated by full-duplex legitimate users Expressed as Where U represents the linear decoding matrix; then the information of full-duplex legal user estimation is calculated The mean square error with the data symbol vector s Then obtain the communication achievable rate R of full-duplex legal users u The conversion expression is described as: Where V represents the auxiliary matrix introduced. Then, the constraint C1 in the optimization problem in step 3 is transformed into a form that is easy to handle as constraint C1'. The constraint C1' is described as: Finally, we obtain a tractable optimization problem, which is described as:

8. The RIS-assisted SWIPT-ISAC system security beamforming method according to claim 7 is characterized in that In step 5, the process of obtaining the first sub-problem is: Step 5.a1: When U, V, and Φ are considered constants, the first subproblem is initially described as: Step 5.a2: In the initial description of the first subproblem, for constraint C1', change The optimal solution Substitution In, get Update to get constraint C1", where Convert constraint C2 to constraint C2', which is described as: Using the first-order Taylor expansion and SCA method, constraint C3 is transformed into constraint C3', which is described as: Convert constraint C4 to constraint C4', which is described as: Among them, Re{·} means taking the real part, W0 represents the value of W obtained in the previous iteration during the internal iteration process when using the first-order Taylor expansion and SCA method; Step 5.a3: The final description of the first subproblem is:

9. The RIS-assisted SWIPT-ISAC system security beamforming method according to claim 8 is characterized in that In step 5, the process of obtaining the second sub-problem is: Step 5.b1: When U, V, and W are considered constants, the second subproblem can be preliminarily described as follows: Step 5.b2: In the initial description of the second subproblem, ignore the objective function The term without variable Φ in the function transforms the objective function into Then Substituting into and in, C=FWW H F H , Const1 and Const2 are both constant terms; ignore the terms without variable Φ and transform the objective function into Tr(Φ H BΦC)+Tr(Φ H P H )+Tr(ΦP), where Then the objective function is finally transformed into Where φ represents the phase vector of RIS, φ=[φ1,…,φ M ] T , "⊙" is the Hadamard product operator, and the superscript "*" represents the conjugate operation. [P] 1,1 represents the first row and first column element in P, [P] M,M Represents the element in row M and column M of P; In the preliminary description of the second subproblem, constraint C3 is changed to Then converted into φ H (B⊙C T )φ+2Re{φ H h * }≥E min / (1-ρ)-Tr(G r WW H ),in, Then, the first-order Taylor expansion and SCA method are used to convert φ H (B⊙C T )φ+2Re{φ H h * }≥E min / (1-ρ)-Tr(G2WW H ) is converted into constraint C3", which is described as: in, represents the value of φ obtained in the previous iteration during the internal iteration process when using the first-order Taylor expansion and SCA method; Step 5.b3: Transform the initial description of the second subproblem into Step 5.b4: In this paper, the alternating direction multiplier method is used and auxiliary variables are introduced Convert constraint C5 to constraint C5', which is described as: m |≤1,m=1,...,M, and add constraint C6: and constraint C7: Step 5.b5: Convert to Step 5.b6: Transformed into the augmented Lagrangian function form as the final description of the second subproblem, Where η represents the preset penalty parameter, η ≥ 0, and μ represents the dual variable.

10. The RIS-assisted SWIPT-ISAC system security beamforming method according to claim 9 is characterized in that In step 5, the alternating iterative optimization algorithm is used to iteratively solve the two sub-problems to obtain the optimized solutions of the active beamforming matrix of the base station and the passive beamforming matrix of the RIS. The process is: Step 5.c1: Let l denote the number of iterations, let l max Indicates the maximum number of iterations; initialize W and Φ; Step 5.c2: Given the values ​​of W and Φ, let The first-order partial derivative of is zero, and the optimal solution U is obtained opt , Then U opt Substitution In, get The optimal solution Then we get the optimal solution V opt , Among them, the values ​​of W and Φ given when l = 1 are the initialization values, and the values ​​of W and Φ given when l > 1 are the values ​​of W and Φ obtained in the l-1th iteration; Step 5.c3: Given the values ​​of U, V, and Φ, use the convex optimization solver to solve the final description of the first subproblem and obtain the value of W at the lth iteration; where the given values ​​of U and V correspond to the U obtained at the lth iteration. opt and V opt , when l=1, the value of Φ given is the initialization value, and when l>1, the value of Φ given is the value of Φ obtained in the l-1th iteration; Step 5.c4: Given the values ​​of U, V, and W, use the convex optimization solver to solve the final description of the second subproblem and obtain the value of Φ at the lth iteration; where the given values ​​of U and V correspond to the U obtained at the lth iteration. opt and V opt , when l=1, the value of W given is the initialization value, and when l>1, the value of W given is the value of W obtained in the l-1th iteration; Step 5.c5: Determine whether l is less than or equal to l max If yes, set l = l + 1, and then return to step 5.c2 to continue execution, otherwise, set the lth max The value of W obtained in the first iteration is used as the optimized solution of the active beamforming matrix of the base station. max The value of Φ obtained from the iteration is used as the optimized solution of the passive beamforming matrix of RIS, where "=" in l=l+1 is an assignment symbol.

Images