Reconfigurable intelligent surface-assisted communication and induction integrated system safety rate maximization method and electronic equipment

Abstract

The invention relates to a reconfigurable intelligent surface assisted ISAC system safety rate maximization method and electronic equipment. The method comprises the following steps: S1, a base station transmits a pre-coding signal to a receiving end; s2, taking the signal in the S1 as an input parameter, executing a safety rate maximization model construction method, outputting a problem model, and initializing a system parameter; s3, according to the problem model and the system parameters in the S2, executing a safety rate maximization transmitting beam forming design method, and outputting w * and R *; s4, inputting the w * and the R * obtained in the S3 into a safety rate maximization RIS reflection phase shift design method, and outputting theta *; s5, the w * and the R * obtained in the S3 and the theta * obtained in the S4 serve as input parameters, the system safety rate # imgabs0 # is calculated, whether the formula # imgabs1 # is met or not is judged, and if yes, the step S6 is executed; otherwise, executing the step S3 and letting l = l + 1; and S6, outputting the maximum value # imgabS2 of the system safety rate. According to the method, the system safety rate can be maximized, the radar detection performance is considered, and effective compromise can be performed between the system safety rate and the radar detection performance.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communication, and in particular relates to a method for maximizing the safety rate of an ISAC system assisted by a reconfigurable intelligent surface (RIS) and an electronic device. Background Art

[0002] Integrated Sensing and Communication (ISAC) is considered to be the core technology for realizing intelligent communication and high-precision environmental perception. By integrating communication and perception functions, ISAC significantly improves spectrum and energy efficiency. As an emerging metasurface technology, Reconfigurable Intelligent Surface (RIS) provides a new solution for optimizing the performance of ISAC systems. RIS is equipped with passive digitally controlled reflective elements that can intelligently adjust the phase shift of the signal that hits it to create a favorable propagation environment and provide additional degrees of freedom for optimization. RIS has shown great potential in improving the communication and perception capabilities of ISAC systems. However, while sharing spectrum resources, ISAC systems also face severe security risks. Physical Layer Security (PLS) technology has received widespread attention in recent years. By optimizing beamforming or introducing artificial interference signals, PLS provides security for ISAC systems. However, there are currently problems with secure communication by malicious eavesdroppers in the downlink communication link of the ISAC system. Summary of the invention

[0003] Aiming at the problem of secure communication in the presence of malicious eavesdroppers in the downlink communication link of the existing ISAC system, the present invention proposes a method and electronic equipment for maximizing the security rate of a synaesthesia integrated system assisted by a reconfigurable intelligent surface.

[0004] To solve the above problems, the present invention adopts the following technical solutions:

[0005] The method for maximizing the safety rate of the synaesthesia integrated system assisted by a reconfigurable intelligent surface has the following specific steps:

[0006] S1: The base station transmits a precoded signal x to the intelligent surface RIS, the communication user, and the detection target. The precoded signal x is an N×N dimensional matrix whose elements are all complex numbers;

[0007] S2: At the receiving end of the communication user and the detection target, the precoded signal of step S1 is received as an input parameter, the security rate maximization model construction method is executed, the problem model is output, and the system parameters are initialized, and the number of iterations l=1;

[0008] S3: Based on the problem model and system parameters obtained in step S2, the safety rate maximization transmit beamforming design method is executed, and the output elements are all complex N×N dimensional matrix w * With R * , the superscript * indicates the optimal value of the matrix;

[0009] S4: The matrix w obtained in step S3 * With R * , which is input into the safety rate maximization RIS reflection phase shift design method, and the output elements are all complex M×1-dimensional column vectors θ * , the superscript * indicates the optimal value of the vector;

[0010] S5: w obtained in step S3 * , R * and θ obtained in step S4 * As input parameter, calculate the system safety rate And judge whether it satisfies the formula in, is the system safety rate value of the lth iteration, ζ is a constant. If not satisfied, return to step S3 and set l=l+1. If satisfied, execute step S6;

[0011] S6: The maximum system safety rate obtained in step S5 As the final output result, the present invention can maximize the system safety rate and take into account the radar detection performance.

[0012] Preferably, in step S2, the method for constructing a safety rate maximization model is as follows:

[0013] The RIS-assisted synaesthesia integrated system includes a base station equipped with a uniform linear array and a positive integer N antenna number, a positive integer K single-antenna communication user, an eavesdropper, a positive integer T detection target, and a RIS with a positive integer M passive reflection units. RIS mainly aims the beamforming at the user to assist communication.

[0014] The communication channel coefficient matrix is ​​expressed as follows: is an N×1-dimensional vector, representing the channel coefficient matrix from the ISAC base station to the communication user k or the eavesdropper; It is an M×N dimensional matrix, representing the channel coefficient matrix from the ISAC base station to the RIS; is an M×1 dimensional vector, representing the channel coefficient matrix from RIS to communication user k or eavesdropper. Where z∈{k,e}, k is a positive integer, k=1,2,…,K. The RIS phase shift matrix is ​​defined as is an M×1 dimensional vector, satisfying |θ m |=1,m=1,2,…,M.

[0015] Therefore, the signal-to-noise ratio of communication user k is:

[0016]

[0017] The superscript H represents the conjugate transpose. is an N×1 dimensional intermediate variable vector, represents the communication beamforming vector of the received signal of communication user k in N×1 dimension, represents the N×N dimensional radar signal covariance matrix, is a constant and is the variance of the complex Gaussian white noise at user k. The signal-to-noise ratio of the eavesdropper is:

[0018]

[0019] in, is an N×1 dimensional intermediate variable vector, is a constant, which is the variance of the complex Gaussian white noise at the eavesdropper. Therefore, the secure rate of communication user k is:

[0020] C s,k =[C k -C e ] + (3)

[0021] Where [x] + =max(x,0),C k =log 2 (1+γ k ) and C e =log 2 (1+γ e ) are the transmission rates of communication user k and eavesdropper respectively. The system security rate is the sum of the security rates of all communication users, expressed as:

[0022] The transmitted signal of the RIS-assisted synaesthesia integrated system base station can be used for radar detection and communication functions. The covariance matrix of the transmitted dual-function waveform is defined as:

[0023]

[0024] in, Define all N×N dimensional matrices Therefore, the transmit beam pattern is expressed as:

[0025]

[0026] in, is the 1×N dimensional transmitting antenna array steering vector

[0027] Designing a near-ideal detection beam pattern is to design the transmit beam covariance matrix, which can be expressed as:

[0028]

[0029] in, To detect the target angle, it can typically be represented as a detection range covering -90° to 90° with an interval of 1°, and L=181 at this time. is the ideal detection beam pattern required by the target, and υ is the proportional coefficient. The similarity between the transmission beam and the near-ideal detection beam can be used as the radar detection performance indicator in the system. Therefore, the detection performance constraint is:

[0030]

[0031] in, It is obtained from formula (6) that ||·|| F is the F-norm, and Γ is the beam similarity coefficient.

[0032] Set the base station power budget and the base station signal transmission power constraint as:

[0033]

[0034] Among them, P max is the power budget and Tr(·) is the trace of the matrix.

[0035] According to the definition of RIS passive reflection unit, the RIS reflection phase shift unit mode constraint is obtained:

[0036]

[0037] The goal of the constructed problem model is to maximize the system safety rate. In order to take into account the radar detection performance, the formula (6) is Substituting into the communication beamforming problem, we get the detection performance constraint, and satisfy the base station transmit power constraint and the RIS phase shift unit mode constraint, and jointly optimize the communication beamforming vector w, radar covariance matrix R and RIS phase shift θ. Therefore, the problem model is:

[0038]

[0039] Among them, C1 is the transmit beam detection performance constraint, C2 is the base station maximum transmit power constraint, and C3 is the unit mode constraint of the RIS reflection phase shift. Set the number of iterations l = 1, and the base station maximum transmit power P max , convergence coefficient ζ, beam similarity coefficient Γ, initialize w according to the constraints 0 , R 0 ,θ 0 .

[0040] Preferably, in the safety rate maximization transmit beamforming design method of step S3,

[0041] According to the problem model and system parameters obtained in step S2, since the problem model has non-convex solution objectives and C3 constraints and multi-variable highly coupled, the present invention proposes an algorithm based on alternating optimization, decoupling problem P1 into two sub-problems, and using the existing semi-definite relaxation (SDR) and continuous convex approximation (SCA) to alternately solve w, R and θ. Specifically:

[0042] S3.1: Fix the arbitrary RIS reflection phase shift θ of the input and optimize the arbitrary w and R of the input, so problem P1 can be rewritten as:

[0043]

[0044] Since the objective function is a non-convex complex quadratic fraction, the present invention adopts the existing SDR method. Define an N×N dimensional matrix W=ww H And W≥0, Rank(W)=1, and an auxiliary variable constant α is introduced, so P2 is rewritten as:

[0045]

[0046] Among them, the rank 1 constraint in C4 is non-convex, which is temporarily ignored, so the relaxed form of P3 is obtained:

[0047]

[0048] S3.2: Defining the solution set Expand and simplify the C1 constraint to:

[0049] {X(Ω)+F(Ω)+G(Ω)+Y(Ω)}≥α k (14)

[0050] in,

[0051]

[0052] in, is the intermediate variable matrix, z∈{k,e}. Since the constraint C1 is still non-convex, the continuous asymptotic convex approximation (SCA) method of the prior art is used to perform first-order Taylor expansion on the intermediate terms F(Ω) and G(Ω), and the approximate replacement formula for its convex lower bound is obtained as follows:

[0053]

[0054] Among them, W n and R n are the nth iteration values ​​of W and R respectively.

[0055] S3.3: P4 is a prior art semidefinite programming (SDP) problem, which can be solved directly using the prior art convex optimization solver. Since the rank 1 constraint of the matrix W is ignored in S3.1, when the optimal W is obtained * Satisfy Rank(W * )=1, w can be obtained by eigenvalue decomposition of the prior art * Otherwise, the Gaussian randomization method of the prior art is used to recover w * Finally, the optimal output w * With the optimal R * .

[0056] Preferably, in step S4, the safety rate maximization RIS reflection phase shift design method is specifically as follows:

[0057] S4.1: Input and fix any w obtained in step S3 * , R * , optimize the RIS reflection phase shift θ. In order to transform the objective function in P1 into a more easily solvable form, define the intermediate variable matrix:

[0058]

[0059] Where z∈{k,e}. Introduce the matrix V=vv H , similar to S3.1, the relaxed form of P1 is obtained:

[0060]

[0061] in,

[0062]

[0063] S4.2: Obviously, problem P5 is still non-convex, so the SCA method of the prior art is used to perform first-order Taylor expansion on the intermediate variables Gu and Fe to obtain their lower bounds:

[0064]

[0065] Among them, V n Represents the nth iteration value of V.

[0066] S4.3: In summary, P5 is a semi-definite programming (SDP) problem in the prior art, which can be directly solved by the convex optimization solver in the prior art. * When v is obtained, the eigenvalue decomposition and Gaussian randomization of the existing technology can be used. * . Restore the reflected phase shift and output:

[0067]

[0068] Among them, satisfying |θm |=1,m=1,2,…,M.

[0069] The present invention also discloses an electronic device, comprising:

[0070] processor;

[0071] The memory is used to store a program. When the program is called and executed by the processor, the processor executes the method as described above.

[0072] Some prior art introductions of the present invention are as follows:

[0073] 1. Alternating Optimization Method

[0074] Alternating optimization is a commonly used optimization method, especially suitable for complex optimization problems that are decomposed into multiple sub-problems. In dealing with system design with non-convex objective functions or constraints, alternating optimization effectively reduces the difficulty of global optimization by optimizing each sub-problem step by step. In each step, the remaining variables are fixed and only the current variable is optimized, so as to gradually approach the global optimal solution. This method has been widely used in scenarios such as beamforming, resource allocation, and joint optimization of intelligent reflector-assisted systems in communication systems. For details, see: "Luo ZQ, Ma WK, So AMC, et al. Semidefinite relaxation of quadratic optimization problems[J]. IEEE Signal Processing Magazine, 2010, 27(3): 20-34.".

[0075] 2. Semi-definite relaxation method

[0076] Semidefinite relaxation (SDR) is an effective method for solving non-convex optimization problems and is widely used in signal processing, communication system design and other fields. For non-convex problems such as quadratically constrained quadratic programming (QCQP), it is usually difficult to solve them directly. SDR converts the original problem into a semidefinite programming (SDP) that is easier to solve by relaxing the non-convex constraints in the original problem into semidefinite constraints. Specifically, SDR upgrades the variables to matrix form and replaces their positive definiteness requirements with semidefinite conditions, thus forming a convex optimization problem. Although the relaxed solution to the problem may not satisfy the constraints of the original problem, a feasible solution close to the optimal solution of the original problem can be recovered from the SDP solution through randomization methods or other post-processing techniques. SDR has shown good performance in problems such as beamforming design and power allocation, and is widely used in the design of RIS-assisted systems. For details, see: "Luo Z Q, Ma WK, So AM C, et al. Semidefinite relaxation of quadratic optimization problems[J]. IEEE Signal Processing Magazine, 2010, 27(3): 20-34.".

[0077] 3. Continuous Asymptotically Convex Approximation Method

[0078] The continuous asymptotic convex approximation method (SCA) is an iterative method for dealing with non-convex optimization problems. By localizing the non-convex problem in each iteration, SCA replaces the original problem solution with an approximate solution to the convex problem, gradually approaching the global or local optimal solution. In the specific implementation, SCA replaces the objective function and constraints with convex functions, while ensuring that the solution of each step belongs to the feasible domain of the original problem. This method can theoretically guarantee iterative convergence and efficiently solve the problem in each step using existing convex optimization tools. For details, see: "Scutari G, Facchinei F, Song P. Decomposition by partiallinearization: Parallel optimization of multi-agent systems[J]. IEEE Transactions on Signal Processing, 2014, 62(3): 641-656.".

[0079] 4. Eigenvalue Decomposition Method

[0080] Eigenvalue decomposition (EVD) is a basic tool in matrix theory, which is used to decompose a square matrix into a combination of its eigenvalues ​​and eigenvectors. The core principle is that for an n×n square matrix A, if there exists a scalar λ and a non-zero vector v, satisfying Av=λv, then λ is called the eigenvalue of the matrix A, and v is the corresponding eigenvector. Through eigenvalue decomposition, the matrix A can be expressed as A=VΛV -1 , where Λ is a diagonal matrix containing all eigenvalues, and V is an eigenvector matrix. The essence of EVD is to decompose the matrix mapping into a set of linearly independent directions and scaling factors. This process not only reveals the geometric properties of the matrix, but also provides an important basis for matrix dimensionality reduction, data analysis, and solving optimization problems. For details, see: "Horn RA, Johnson CR. Matrix Analysis [M]. Cambridge University Press, 2012.".

[0081] 5. Gaussian randomization method

[0082] Gaussian randomization is a method for extracting approximate feasible solutions from semidefinite relaxation (SDR) solutions by random sampling. Its core principle is to generate random vectors using Gaussian distribution and ensure that the solution satisfies the constraints of the original problem through post-processing. Specifically, the semidefinite matrix obtained after SDR solution is decomposed into eigenvalues ​​and eigenvectors. Random vectors are generated by Gaussian distribution and combined with the eigencomponents of the matrix to form an initial solution. Subsequently, the random vectors are adjusted using normalization or projection methods to satisfy the constraints of the original problem. Finally, the one with the best objective function value is selected from the multiple candidate solutions generated as the final solution. Gaussian randomization is widely used in non-convex problems such as quadratically constrained quadratic programming (QCQP). For details, see: "Wu Q, Zhang R. Intelligent reflecting surface enhanced wireless network via joint active and passive beamforming [J]. IEEE Transactions on Wireless Communications, 2019, 18 (11): 5394-5409.".

[0083] 6. Convex Optimization Solver

[0084] Convex optimization solvers are a general term for algorithms and software tools specifically used to solve convex optimization problems (such as linear programming, quadratic programming, second-order cone programming, semidefinite programming, etc.). Most of them are based on efficient numerical algorithms such as the Interior Point Method, and can converge to the global optimal solution in polynomial time or give high-precision approximate solutions within the feasible domain. Common convex optimization solvers include SeDuMi, SDPT3, MOSEK, Gurobi, etc. In addition, there are toolkits based on modeling languages, such as CVX (MATLAB environment) or CVXPY (Python environment), which can greatly simplify the problem modeling and solution process. Convex optimization solvers are widely used in fields such as wireless communications, signal processing, and machine learning. By reasonably convexifying or relaxing the feasible domain and objective function, many originally complex non-convex problems can obtain feasible high-quality solutions with the assistance of convex optimization solvers. For details, see: "Luo ZQ, Ma WK, So AM C, et al. Semidefiniterelaxation of quadratic optimization problems [J]. IEEE SignalProcessingMagazine, 2010, 27(3): 20-34."

[0085] The beneficial effects of the present invention are:

[0086] The present invention provides a RIS-assisted ISAC system security rate maximization method for the presence of malicious eavesdroppers in the downlink communication link of the ISAC system. The present invention maximizes the radar detection performance under power constraints by designing the transmit beam covariance matrix and substituting it into the communication link optimization problem, and achieves an effective compromise between detection performance and security performance by adjusting the beam similarity coefficient; based on the continuous convex approximation (SCA) and semi-definite relaxation (SDR) methods, the communication beamforming vector w, the radar covariance matrix R and the RIS phase shift θ are alternately optimized to maximize the system security rate. Simulations show that the technical solution proposed by the present invention is superior to the ISAC system without RIS assistance and the random phase RIS-assisted ISAC system; in addition, the present invention can maximize the system security rate while taking into account the radar detection performance. The present invention can take into account both radar detection performance and system security rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] Figure 1 A model diagram of an ISAC system assisted by RIS according to a preferred embodiment of the present invention;

[0088] Figure 2 A flow chart of a RIS-assisted ISAC system safety rate maximization method according to a preferred embodiment of the present invention;

[0089] Figure 3 This is a comparison chart of algorithm convergence performance under different schemes;

[0090] Figure 4 It is a comparison chart of the relationship between the number of RIS reflection units and the system safety rate under different schemes;

[0091] Figure 5 The figure is a comparison chart of the relationship between the transmission beam detection performance and the system safety rate under different schemes. DETAILED DESCRIPTION

[0092] The present invention will be further described below in conjunction with specific embodiments. The specific implementation of the present invention can be described in detail through the following embodiment figures.

[0093] Figure 1 The RIS-assisted ISAC system model diagram involved in the embodiment of the present invention. The RIS-assisted ISAC system includes a base station equipped with a uniform linear array and N antennas, K single-antenna communication users, a single-antenna eavesdropper, T detection targets, and a RIS with M passive reflection units. The radar waveform shares the spectrum with the communication signal to form the transmission signal. The RIS mainly aims the beamforming at the user, so it only assists the communication and ignores the reflected echo signal.

[0094] Figure 2 This is a flow chart of a RIS-assisted ISAC system safety rate maximization method according to an embodiment of the present invention, which is mainly completed by the following steps:

[0095] Step S1: The base station transmits a precoded signal through free space to the RIS, the communication user, and the detection target;

[0096] Step S2: At the receiving end of the communication user and the detection target, the precoded signal of step S1 is received as the input parameter, the safety rate maximization model construction method is executed, the problem model is output, the number of iterations l=1, and the maximum transmission power of the base station P is set. max , convergence coefficient ζ, beam similarity coefficient Γ, initialize w according to the constraints 0 , R 0 ,θ 0 ;

[0097] Step S3: According to the problem model and system parameters obtained in step S2, fix θ l-1 , execute the safe rate maximization transmit beamforming design method, and output w l , R l ;

[0098] Step S4: Input and fix w obtained in step S3 l , R l, θ is obtained by the safe rate maximization RIS reflection phase shift design method l ;

[0099] Step S5: Based on w obtained in step S3 l , R l and θ obtained in step S4 l , calculate the system safety rate of the lth iteration And judge Is it true? If it is true, go to step S6. If it is not true, go back to step S3 and set l = l + 1;

[0100] Step S6: The maximum system safety rate obtained in step S5 as the final output result.

[0101] In step S2 of this embodiment, the method for constructing a safety rate maximization model is as follows:

[0102] The communication channel coefficient matrix is ​​expressed as follows: is an N×1-dimensional vector, representing the channel coefficient matrix from the ISAC base station to the communication user k or the eavesdropper; It is an M×N dimensional matrix, representing the channel coefficient matrix from the ISAC base station to the RIS; is an M×1 dimensional vector, representing the channel coefficient matrix from RIS to communication user k or eavesdropper. Where z∈{k,e}, k is a positive integer, k=1,2,…,K. The RIS phase shift matrix is ​​defined as is an M×1 dimensional vector, satisfying |θ m |=1,m=1,2,…,M.

[0103] Therefore, the signal-to-noise ratio of communication user k is:

[0104]

[0105] The superscript H represents the conjugate transpose. is an N×1 dimensional intermediate variable vector, represents the communication beamforming vector of the received signal of communication user k in N×1 dimension, represents the N×N dimensional radar signal covariance matrix, is a constant and is the variance of the complex Gaussian white noise at user k. The signal-to-noise ratio of the eavesdropper is:

[0106]

[0107] in, is an N×1 dimensional intermediate variable vector, is a constant, which is the variance of the complex Gaussian white noise at the eavesdropper. Therefore, the secure rate of communication user k is:

[0108] C s,k =[C k -C e ] + (3)

[0109] Where [x] + =max(x,0),C k =log 2 (1+γ k ) and C e =log 2 (1+γ e ) are the transmission rates of communication user k and eavesdropper respectively. The system security rate is the sum of the security rates of all communication users, expressed as:

[0110] The transmitted signal of the RIS-ISAC system base station can be used for radar detection and communication functions. The covariance matrix of the transmitted dual-function waveform is defined as:

[0111]

[0112] in, Define all N×N dimensional matrices Therefore, the transmit beam pattern is expressed as:

[0113]

[0114] in, is the 1×N dimensional transmitting antenna array steering vector

[0115] Designing a near-ideal detection beam pattern is to design the transmit beam covariance matrix, which can be expressed as:

[0116]

[0117] in, To detect the target angle, it can typically be represented as covering a detection range of -90° to 90° with an interval of 1°, in which case L=181. is the ideal detection beam pattern required by the target, and υ is the proportional coefficient. The similarity between the transmission beam and the near-ideal detection beam can be used as the radar detection performance indicator in the system. Therefore, the detection performance constraint is:

[0118]

[0119] in, It is obtained from formula (6) that ||·|| F is the F-norm, and Γ is the beam similarity coefficient.

[0120] Set the base station power budget and the base station signal transmission power constraint as:

[0121]

[0122] Among them, P max is the power budget and Tr(·) is the trace of the matrix.

[0123] According to the definition of RIS passive reflection unit, the RIS reflection phase shift unit mode constraint is obtained:

[0124]

[0125] The goal of the constructed problem model is to maximize the system safety rate. In order to take into account the radar detection performance, the formula (6) is Substituting into the communication beamforming problem, we get the detection performance constraint, and satisfy the base station transmit power constraint and the RIS phase shift unit mode constraint, and jointly optimize the communication beamforming vector w, radar covariance matrix R and RIS phase shift θ. Therefore, the problem model is:

[0126]

[0127] Among them, C1 is the transmit beam detection performance constraint, C2 is the base station maximum transmit power constraint, and C3 is the unit mode constraint of the RIS reflection phase shift. Set the number of iterations l = 1, and the base station maximum transmit power P max , convergence coefficient ζ, beam similarity coefficient Γ, initialize w according to the constraints 0 , R 0 ,θ 0 .

[0128] In the safety rate maximization transmit beamforming design method of step S3 of this embodiment, according to the problem model and system parameters obtained in step S2, since the problem model, the solution target and C3 constraint are non-convex and multi-variables are highly coupled, the present invention proposes an algorithm based on alternating optimization, decouples the problem P1 into two sub-problems, and uses the existing semi-definite relaxation (SDR) and continuous convex approximation (SCA) to alternately solve w, R and θ. Specifically:

[0129] S3.1: Fix the arbitrary RIS reflection phase shift θ of the input and optimize the arbitrary w and R of the input, so problem P1 can be rewritten as:

[0130]

[0131] Since the objective function is a non-convex complex quadratic fraction, the present invention adopts the existing SDR method. Define an N×N dimensional matrix W=ww H And W≥0, Rank(W)=1, and an auxiliary variable constant α is introduced, so P2 is rewritten as:

[0132]

[0133] Among them, the rank 1 constraint in C4 is non-convex, which is temporarily ignored, so the relaxed form of P3 is obtained:

[0134]

[0135] S3.2: Defining the solution set Expand and simplify the C1 constraint to:

[0136] {X(Ω)+F(Ω)+G(Ω)+Y(Ω)}≥α k (14)

[0137] in,

[0138]

[0139] in, is the intermediate variable matrix, z∈{k,e}. Since the intermediate variables F(Ω) and G(Ω) are still non-convex terms, the continuous asymptotic convex approximation (SCA) method of the prior art is used to perform a first-order Taylor expansion on the non-convex terms to obtain the approximate replacement formula of its convex lower bound as follows:

[0140]

[0141] Among them, W n and R n are the nth iteration values ​​of W and R respectively.

[0142] S3.3: P4 is a prior art semidefinite programming (SDP) problem, which can be solved directly using a convex optimization solver. Since the rank 1 constraint of the matrix W is ignored in S3.1, when the optimal W is obtained * Satisfy Rank(W * )=1, w can be obtained by eigenvalue decomposition of the prior art * Otherwise, the Gaussian randomization method of the prior art is used to recover w * Finally, the optimal output w * With the optimal R * .

[0143] In step S4 of this embodiment, the safety rate maximization RIS reflection phase shift design method is specifically as follows:

[0144] S4.1: Input and fix any w obtained in step S3 * , R * , optimize the RIS reflection phase shift θ. In order to transform the objective function in P1 into a more easily solvable form, define the intermediate variable matrix:

[0145]

[0146] Where z∈{k,e}. Introduce the matrix V=vv H , similar to S3.1, the relaxed form of P1 is obtained:

[0147]

[0148] in,

[0149]

[0150] S4.2: Obviously, the intermediate variables Gu and Fe are still non-convex terms, so the SCA method of the prior art is used to perform a first-order Taylor expansion on the non-convex terms to obtain their lower bounds:

[0151]

[0152] Among them, V n Represents the nth iteration value of V.

[0153] S4.3: In summary, P5 is a semi-definite programming (SDP) problem in the prior art, which can be solved directly using a convex optimization solver. * When v is obtained, the eigenvalue decomposition and Gaussian randomization of the existing technology can be used. * . Restore the reflected phase shift and output:

[0154]

[0155] Among them, satisfying |θ m |=1,m=1,2,…,M.

[0156] Figure 3 The following is a comparison chart of algorithm convergence performance under different schemes. The number of antennas is 8, the number of communication users is 4, the number of RIS reflection units is 20, and the base station transmission power is set to 20dB. Figure 3 It can be seen that the convergence point of each scheme can be reached in the fifth iteration, which proves the rapid convergence of the method of the present invention. The final safety rate of the method of the present invention is significantly higher than that of other comparative schemes, which proves the effectiveness of the method of the present invention in maximizing the safety rate in this system.

[0157] Figure 4The figure is a relationship diagram between the number of RIS reflection units and the system safety rate under different schemes. According to the definition of the RIS phase shift matrix, a larger number of RIS reflection elements means that a stronger beam gain can be formed by utilizing a higher spatial degree of freedom. Therefore, when N=8 and K=4, as the number of RIS reflection elements increases from M=20 to M=60, the system safety rate of the RIS-assisted system also increases accordingly. In addition, the system safety rate of the random phase RIS-assisted method does not increase significantly with the increase in the number of RIS reflection elements, and the gap with the method of the present invention increases significantly, reflecting the significance of the RIS phase shift designed by the method of the present invention for improving the system safety rate. The results show that deploying RIS in the ISAC system and designing its phase shift can improve the system's secure communication service quality.

[0158] Figure 5 It is a relationship diagram between the transmission beam detection performance and the system safety rate under different schemes. The smaller the beam similarity coefficient Γ, the closer the transmission beam is to the ideal detection beam, and the stronger the beam detection performance. Due to the transmission power constraint of the base station, regardless of the size of the beam similarity coefficient Γ, the energy concentration of the transmission beam in the target direction is basically consistent with the reference beam, which reflects the consideration of the signal detection performance under the goal of improving the system safety rate in the transmission beam design. As the beam similarity coefficient Γ increases, the system confidentiality rate is higher, and when the beam similarity coefficient Γ = 625, the system safety rate of the method of the present invention is the highest. The reason is that as the beam similarity coefficient Γ increases, the energy concentration of the transmission beam in the target direction becomes smaller and the difference from the ideal detection beam becomes larger. At this time, the power allocated to the communication part by the base station increases, and the system safety rate is naturally improved. In summary, the method of the present invention can maximize the system safety rate while taking into account the radar detection performance. The proposed method can achieve an effective compromise between radar detection performance and system safety rate by adjusting the beam similarity coefficient Γ.

[0159] Although the embodiments of the present invention have been clearly described, it is clear that those skilled in the art may make various changes, modifications, substitutions and variations to these embodiments without departing from the principle and spirit of the method of the present invention. The scope of the present invention is defined by the appended claims and their equivalents, and still falls within the scope of the method of the present invention and is still considered to be the scope of protection of the present invention.

Claims

1. A method for maximizing the safety rate of a synaesthesia integrated system assisted by a reconfigurable intelligent surface, characterized in that: The following steps are involved: S1: The base station transmits a precoded signal x to the intelligent surface RIS, the communication user, and the detection target; wherein the precoded signal x is an N×N dimensional matrix whose elements are all complex numbers; S2: At the receiving end of the communication user and the receiving end of the detection target, the precoded signal of step S1 is received as an input parameter, the safety rate maximization model construction method is executed, the problem model is output, and the system parameters are initialized, and the number of iterations l=1; S3: According to the problem model and system parameters of step S2, the safety rate maximization transmit beamforming design method is executed, and the output elements are all complex N×N dimensional matrix w * With R * , the superscript * indicates the optimal value of the matrix; S4: The matrix w obtained in step S3 * With R * , which is input into the safety rate maximization RIS reflection phase shift design method, and the output elements are all complex M×1-dimensional column vectors θ * , the superscript * indicates the optimal value of the vector; S5: w obtained in step S3 * , R * and θ obtained in step S4 * As input parameter, calculate the system safety rate And judge whether it satisfies the formula in, is the system safety rate value of the lth iteration, ζ is a constant; if not satisfied, return to step S3 and set l=l+1; if satisfied, execute step S6; S6: The system safety rate obtained in step S5 is The maximum value is taken as the final output result.

2. The method for maximizing the safety rate of the synaesthesia integrated system assisted by a reconfigurable intelligent surface as claimed in claim 1, characterized in that: In step S2, the safety rate maximization model construction method is as follows: The RIS-assisted synaesthesia integrated system includes a base station equipped with a uniform linear array and a positive integer N of antennas, a positive integer K of single-antenna communication users, an eavesdropper, a positive integer T of detection targets, and a RIS with a positive integer M of passive reflection units; The communication channel coefficient matrix is ​​expressed as follows: is an N×1-dimensional vector, representing the channel coefficient matrix from the ISAC base station to the communication user k or the eavesdropper; It is an M×N dimensional matrix, representing the channel coefficient matrix from the ISAC base station to the RIS; is an M×1-dimensional vector, representing the channel coefficient matrix from RIS to communication user k or eavesdropper; where z∈{k,e}, k is a positive integer, k=1,2,…,K; the RIS phase shift matrix is ​​defined as is an M×1 dimensional vector, satisfying |θ m |=1,m=1,2,…,M; The signal-to-noise ratio of communication user k is: The superscript H represents the conjugate transpose. is an N×1 dimensional intermediate variable vector, represents the communication beamforming vector of the received signal of communication user k in N×1 dimension, represents the N×N dimensional radar signal covariance matrix, is a constant, which is the variance of the complex Gaussian white noise at user k; the signal-to-noise ratio of the eavesdropper is: in, is an N×1 dimensional intermediate variable vector, is a constant, which is the variance of the complex Gaussian white noise at the eavesdropper; the secure rate of communication user k is: C s,k =[C k -C e ] + ; where [x] + =max(x,0),C k =log2(1+γ k ) and C e =log2(1+γ e ) are the transmission rates of communication user k and the eavesdropper respectively; the system security rate is the sum of the security rates of all communication users, expressed as: The base station's transmitted signal is used for radar detection and communication functions. The covariance matrix of the transmitted dual-function waveform is defined as: in, Define all N×N dimensional matrices The transmit beam pattern is expressed as: in, is the 1×N dimensional transmitting antenna array steering vector; The transmit beam covariance matrix is ​​expressed as: in, To detect the target angle; is the ideal detection beam pattern required by the target, υ is the proportional coefficient; the detection performance constraint is: in, It is obtained from formula (5) that ||·|| F is the F-norm, Γ is the beam similarity coefficient; Set the base station power budget and the base station signal transmission power constraint as: Among them, P max is the power budget, Tr(·) is the trace of the matrix; RIS reflection phase shift unit mode constraint: The covariance matrix of the transmitted dual-function waveform is combined with the R obtained by equation (5) x Matching, detection performance constraints are obtained, and the base station transmit power constraints and RIS phase shift unit mode constraints are met. The communication beamforming vector w, radar covariance matrix R and RIS phase shift θ are jointly optimized; the problem model is: Among them, C1 is the transmit beam detection performance constraint, C2 is the base station maximum transmit power constraint, and C3 is the unit mode constraint of RIS reflection phase shift; set the number of iterations l = 1, the base station maximum transmit power P max , convergence coefficient ζ, beam similarity coefficient Γ, initialize w according to the constraints 0 , R 0 ,θ 0 .

3. The method for maximizing the safety rate of the synaesthesia integrated system assisted by a reconfigurable intelligent surface as claimed in claim 2, characterized in that: In step S3, the safety rate maximization transmit beamforming design method is implemented by the following steps: S3.1: Fix the arbitrary RIS reflection phase shift θ of the input, optimize the arbitrary w and R of the input, and rewrite the problem P1 as: Define N×N dimensional matrix W=ww H And W≥0, Rank(W)=1, and introduce auxiliary variable constant α, rewrite P2: Among them, the rank 1 constraint in C4 is non-convex; S3.2: Defining the solution set Expand and simplify the C1 constraint to: {X(Ω)+F(Ω)+G(Ω)+Y(Ω)}≥a k (12) among them, in, is the intermediate variable matrix, z∈{k,e}; the intermediate terms F(Ω) and G(Ω) of the non-convex constraint C1 are subjected to first-order Taylor expansion using the continuous asymptotic convex approximation method, and the alternative formula for its convex lower bound is obtained as follows: Among them, W n and R n are the nth iteration values ​​of W and R respectively; S3.3: P3 is a semi-positive programming problem, which is solved by a convex optimization solver; when the optimal W is obtained * Satisfy Rank(W * )=1, the optimal w is obtained by eigenvalue decomposition * Otherwise, the Gaussian randomization method is used to restore the optimal w * ; Finally output the optimal w * With the optimal R * .

4. The method for maximizing the safety rate of the synaesthesia integrated system assisted by a reconfigurable intelligent surface as claimed in claim 3, characterized in that: In step S4, the safety rate maximization RIS reflection phase shift design method is implemented by the following steps: S4.1: Input and fix any w obtained in step S3 * , R * , optimize the RIS reflection phase shift θ; Define the intermediate variable matrix: Among them, z∈{k,e}; introduce the matrix V=vv H , relax the rank 1 constraint to obtain: in, S4.2: Using the continuous asymptotic convex approximation method, perform first-order Taylor expansion on Gu and Fe and obtain the lower bound: Among them, V n Represents the nth iteration value of V; S4.3: P4 is a semi-positive programming problem, which is solved by a convex optimization solver; when the optimal V * When , we can get v by eigenvalue decomposition and Gaussian randomization. * , the reflected phase shift is restored to θ * And output.

5. An electronic device, characterized in that: include: processor; The memory is used to store a program, and when the program is called and executed by the processor, the processor executes the method according to any one of claims 1 to 4.

Images