Sensitivity integrated waveform design method for realizing physical layer security and target perception of RSMA system

Abstract

The invention discloses a general inductance integrated waveform design method for realizing physical layer security and target perception of an RSMA system. According to the invention, the base station mainly adopts the RSMA public signal and the artificial noise signal to cooperatively serve as the sensing signal so as to realize target sensing. In the aspect of perception, estimation of azimuth angles and reflection coefficients is mainly carried out, and Cramer-Rao bound is adopted for representation. For a communication link of a known eavesdropper, optimization design is carried out on an emission matrix by taking maximization of a safety rate and safety capacity as a target, and for the condition that an eavesdropper channel is unknown, optimization design is carried out on an emission vector by taking minimization of energy required by communication and perception as a target. Firstly, an optimization problem conforming to a scene is constructed, then, an original problem is converted into a convex problem based on an SCA method, rank 1 constraint is converted into a penalty function in an optimization target, finally, an iterative method is adopted for solving, and eigenvalue decomposition is performed on an optimization solution conforming to conditions to obtain emission vectors. Compared with a mode of sensing by adopting a proprietary sensing signal and an RSMA public signal, the method provided by the invention has better safety energy efficiency on the premise of ensuring safe transmission.

Description

Technical Field

[0001] The present invention relates to the field of physical layer security of an RSMA system under communication-sensing integration, specifically, the secure transmission of information and the estimation of target sensing parameters using an RSMA technique by a communication-sensing integrated base station. The communication-sensing integrated base station transmits communication signals to users. The communication signals are divided into public signals and private signals, and at the same time, artificial noise is sent to interfere with eavesdroppers. The public signal and the artificial noise cooperate as sensing signals, and the azimuth angle and reflection coefficient of the target are estimated through the echo of the signals. By jointly designing the base station transmission waveform, physical layer security and target sensing parameter estimation under the RSMA system are achieved. Background Art

[0002] The statements in this section merely provide background technical information related to the present disclosure, and these statements may constitute prior art. In the process of implementing the present invention, the inventors found that there are at least the following problems in the prior art.

[0003] With the increasing scarcity of the spectrum, the emergence of communication-sensing integration can effectively alleviate this trend. The RSMA technique divides information into public and private streams for transmission respectively, which improves the security performance and energy efficiency to a certain extent. Currently, many related studies on RSMA information transmission are underway. In the literature [H. Xia, S. Han, and C. Li, "Max-min fair optimization in RSMA-assisted secure communications with artificial noise," IEEE Commun. Lett., vol. 27, no. 12, pp. 3181 - 3184, Dec. 2023.], artificial noise is used in the RSMA system to improve the confidentiality of information transmission. Considering the fairness among users, the minimum user rate is maximized under the receiving SNR constraint of the eavesdropper.

[0004] The literature [C. Xu, B. Clerckx, S. Chen, Y. Mao, and J. Zhang, "Rate-splitting multiple access for multi-antenna joint radar and communications," IEEE J. Sel. Topics Signal Process., vol. 15, no. 6, pp. 1332 - 1347, Nov. 2021.] designed message splitting and precoders for communication (private and public) streams and radar sequences to jointly maximize the weighted sum rate (WSR) and minimize the mean square error (MSE) of the radar beam pattern approximation.

[0005] In [K. Chen, Y. Mao, L. Yin, C. Xu, and Y. Huang, "Rate-Splitting Multiple Access for Simultaneous Multi-User Communication and Multi-Target Sensing," IEEE Trans. Veh. Technol., vol. 73, no. 9, pp. 13909-13914, Sep. 2024.], the authors optimized the ISAC waveform to jointly maximize the max-min fairness (MMF) rate of communication users and minimize the maximum eigenvalue of the Cramér-Rao bound (CRB) matrix for unbiased estimation. The CRB matrix is related to the estimation of the angular directions, complex reflection coefficients, and Doppler frequencies of multiple moving targets.

[0006] The above solutions have the following problems:

[0007] 1. The research on physical layer security of traditional RSMA systems has not been extended to the field of communication-sensing integration;

[0008] 2. The research on communication-sensing integration of RSMA has not paid attention to the research on security energy efficiency;

[0009] 3. The cooperative sensing of common signals and artificial noise in communication signals has not been considered. Summary of the Invention

[0010] Aiming at the above problems, the purpose of the present invention is to solve a part of the problems in the prior art, or at least alleviate these problems.

[0011] The purpose of this application is to provide a communication-sensing integrated waveform design method for realizing physical layer security and target sensing of the RSMA system, jointly optimizing the transmission waveform of the base station, and realizing physical layer security and target sensing of the RSMA system by using the cooperative sensing of RSMA common signals and artificial noise.

[0012] The technical solution adopted by the present invention is: a communication-sensing integrated waveform design method for realizing physical layer security and target sensing of the RSMA system, including the following steps:

[0013] Construct a communication system model: The communication system model includes a communication-sensing integrated base station E. This base station uses the RSMA technology to serve K users simultaneously, while there are M external eavesdroppers eavesdropping on the legitimate information, and the base station needs to sense T targets at the same time. Determine the positions and channels of the K legitimate users, determine the positions and channels of the M eavesdroppers, and according to the forward experience information, know the azimuth and reflection coefficient of the targets. The base station E transmits artificial noise to interfere with the eavesdroppers to achieve physical layer security, and at the same time uses artificial noise and the RSMA common signal to cooperate as sensing signals to sense the T targets.

[0014] Construct an optimization model: When adopting the security performance index that the eavesdroppers' common signal eavesdropping rate and private signal eavesdropping rate are not higher than the threshold, with the goal of maximizing the minimum secure transmission energy efficiency of the system users, construct optimization model I with the common rate and private rate constraints as the constraints; when adopting the secrecy capacity as the security performance index, with the goal of maximizing the minimum secure transmission secrecy capacity energy efficiency of the system users, construct optimization model II with the rate constraints of the eavesdroppers and legitimate users as the constraints.

[0015] For the optimization model I and optimization model II, respectively use the methods of successive convex optimization and semidefinite relaxation, discard the rank-1 constraint to transform the unsolvable non-convex problem into a convex problem, and use the CVX toolbox to solve to obtain the optimal solution.

[0016] Starting from the optimal solution, relax the rank-1 constraint and turn it into a penalty function in the optimization objective; relax the rank-1 constraint to the difference between the trace of the matrix and the product of the maximum eigenvector of the matrix and the matrix, and transform it into a penalty function of the optimization objective, and iterate to solve until convergence.

[0017] Use the iterative algorithm to solve the base station transmit covariance matrix, and then use eigenvalue decomposition to obtain the transmit vector.

[0018] The present invention also provides a communication-sensing integrated system for realizing the physical layer security and target sensing of the RSMA system, including

[0019] Initialization module: Used to determine the positions and channels of the K legitimate users, determine the positions and channels of the M eavesdroppers, know the azimuth and reflection coefficient of the targets according to the forward experience information, and determine the security strategy, that is, adopt the security performance index that the eavesdroppers' common signal eavesdropping rate and private signal eavesdropping rate are not higher than the threshold, or adopt the secrecy capacity as the security performance index.

[0020] Mathematical model construction module: It is used to construct an optimization model according to the determined security policy. When adopting the security performance index that the eavesdropping rates of the public signal and the private signal of the eavesdropper are not higher than the threshold, aiming at maximizing the minimum secure transmission energy efficiency of the system users, an optimization model Ⅰ with the public rate and private rate constraints as the constraints is constructed; when adopting the secrecy capacity as the security performance index, aiming at maximizing the minimum secure transmission secrecy capacity energy efficiency of the system users, an optimization model Ⅱ with the rate constraints of the eavesdropper and legitimate users is constructed.

[0021] Optimization model solving module: It is used to respectively adopt the methods of successive convex optimization and semidefinite relaxation according to the optimization model Ⅰ and optimization model Ⅱ, discard the rank-1 constraint to transform the unsolvable non-convex problem into a convex problem, and use the CVX toolbox to solve to obtain the optimal solution.

[0022] Iterative solving module: It is used to take the above-mentioned optimal solution as the initial point, relax the rank-1 constraint to become the penalty function in the optimization objective, and continuously perform iterative solving, and the final result tends to the convergence value.

[0023] Transmit vector obtaining module: It is used to solve the base station transmit covariance matrix by using an iterative algorithm, and then perform eigenvalue decomposition to obtain the transmit vector.

[0024] Furthermore, the transmit vector obtaining module performs eigenvalue decomposition according to the iteratively obtained covariance matrices of the public signal, private signal, and artificial noise, and the eigenvector corresponding to the largest eigenvalue is the beam transmit vector.

[0025] The present invention mainly realizes the RSMA physical layer security and target perception under communication and sensing integration. Aiming at optimizing the two performances of RSMA physical layer security and target perception by multiplexing artificial noise and public signal for collaborative perception, a joint design of the beam transmit vector is carried out.

[0026] (1) The present invention first determines the channel model and sensing index, and according to different physical layer security policies, determines the optimization objective and constraint conditions.

[0027] (2) According to the established communication model, a mathematical model is established for the determined optimization objective. The SCA and SDR methods are used to transform the obtained non-convex problem into a convex problem, and CVX is used for solving.

[0028] (3) Based on the iterative idea, taking the optimal solution obtained in step (2) as the starting point, the rank-1 constraint is relaxed into a penalty function and continuously repeated for iteration, so that the final optimization value is stable, and the joint design of the RSMA physical layer security and target perception beam under communication and sensing integration is realized.

[0029] (4) The covariance matrix solved in step (3) is subjected to eigenvalue decomposition to obtain the optimized beamforming vector. Description of the Drawings

[0030] Figure 1 is the communication system model of the present invention;

[0031] Figure 2 is the simulation diagram of optimization algorithm 1;

[0032] Figure 3 is the simulation diagram of optimization algorithm 2. Detailed Implementation Manner

[0033] Next, the preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings in the embodiments of the present invention.

[0034] 1. Construction of the communication model

[0035] Construct a communication system model as shown in Figure 1 : Set up a communication-sensing integrated base station E. This base station uses RSMA technology to serve K users simultaneously. At the same time, there are M external eavesdroppers eavesdropping on the legitimate information, and the base station needs to sense T targets simultaneously. Taking the position of the base station E as the coordinate origin, the position azimuth angles of the K users and M eavesdroppers relative to the base station E are known, and the channels between E and the users and eavesdroppers are fully mastered. The M users are independent of each other. The base station E transmits artificial noise to interfere with the eavesdroppers to improve the security performance, and at the same time uses the artificial noise and the RSMA common signal to cooperate as the sensing signal to estimate the target parameters.

[0036] The channel between the communication-sensing integrated base station antenna and the ground legitimate user antenna follows the Rice distribution and can be expressed as:

[0037]

[0038] where β0 represents the channel gain at 1m, represents the distance between the base station and the legitimate user, α u represents the attenuation coefficient, ρ0 represents the adjustment coefficient, represents the component of the line-of-sight link, represents the component of the non-line-of-sight link, which follows a complex Gaussian distribution with a mean of 0 and a variance of 1.

[0039] The transmitted signal of the base station can be expressed as

[0040]

[0041] where w c represents the transmission vector of the common stream, w k represents the transmission vector of the private stream of the kth user, v represents the transmission vector of the noise signal, s c represents the unit power symbol of the common stream, sk Unit power symbol for the private stream, s v Unit power symbol for the artificial noise. The signal received by the legitimate user can be expressed as

[0042]

[0043] where Channel gain between the base station and the legitimate user, n k Represents additive white Gaussian noise H Represents the conjugate transpose operation of a matrix. After receiving the signal stream, the legitimate user k decodes in the order of public signal first and then private signal. The legitimate user k first decodes the common signal and extracts its own information. When receiving the common signal stream, all received private signals and artificial noise are regarded as interference. The signal-to-interference-plus-noise ratio (SINR) for receiving the public signal can be expressed as

[0044]

[0045] where V = vv H , Represents the power of additive white Gaussian noise of user k's channel, and tr(·) represents the operation of the trace of a matrix. After the legitimate user k finishes decoding the common signal, it uses the SIC technique to remove the common signal from the received signal, and then starts to decode the private signal. When decoding the private signal, all other signals except its own private signal are regarded as interference. At this time, the SINR can be expressed as

[0046]

[0047] According to the above SINR, the public rate and private rate of the legitimate user k can be expressed as

[0048] R c,k = log2(1 + γ c,k )

[0049] R p,k = log2(1 + γ p,k )

[0050] While the base station is communicating with the legitimate user, the illegal user is also eavesdropping. The antenna channel between the base station and the eavesdropping user can be expressed as

[0051]

[0052] Represents the component of the line-of-sight link of the eavesdropping user m Denote the components of the non-line-of-sight link for eavesdropping user m, which follows a complex Gaussian distribution with a mean of 0 and a variance of 1.

[0053] The signal received by the illegal user m can be expressed as

[0054]

[0055] where denotes the channel gain between the base station and the illegal user, and n m denotes the Gaussian white noise. The signal-to-interference-plus-noise ratio of the common signal received by the illegal user m can be expressed as

[0056]

[0057] denotes the Gaussian white noise power of the channel of the illegal user m. To ensure the secure communication between the base station and the legitimate user is not eavesdropped, artificial noise is used to reduce the signal-to-noise ratio of the eavesdropping user's signal reception, so that the reception rate of the common signal by the eavesdropping user is lower than the decoding threshold and cannot be decoded. At the same time, when the eavesdropping user receives the private signal, the non-successfully decoded common signal will also be regarded as interference. At this time, the signal-to-interference-plus-noise ratio of the private signal k received by the eavesdropping user m is

[0058]

[0059] The common rate of the illegal user m and the rate of eavesdropping on the private channel of the legitimate user k can be expressed as

[0060] R c,m = log2(1 + γ c,m )

[0061] R p,k,m = log2(1 + γ p,k,m )

[0062] Radar model:

[0063] According to the literature description, the performance of the dedicated radar signal and the common common stream signal of RSMA is comparable.

[0064] Here we plan to use two types of signals, artificial noise and common stream, as radar signals. At this time, the radar signal can be expressed as

[0065] X r = w c s c + vs v

[0066] Assume that the ISAC uses a linear array. At this time, the radar transmitting array steering vector can be expressed as:

[0067]

[0068] Let λ denote the wavelength, d denote the adjacent antenna spacing, with the leftmost first antenna as the reference point, and θ denote the emission angle. N t denotes the number of transmitting antennas. Similarly, at this time, the receiving array steering vector can be expressed as

[0069]

[0070] N r denotes the number of receiving antennas.

[0071] If there are multiple above target directions, the formula can be rewritten as: where the radar receiving steering vector A r is expressed as: A r = [a r (θ1),..., a r (θ t ),..., a r (θ T )], the radar transmitting steering vector A t is expressed as A t = [a t (θ1),..., a t (θ t ),..., a t (θ T )], B = diag(β1,..., β t ,..., β T ), θ = [θ1,..., θ t ,..., θ T , β t , θ t respectively represent the reflection coefficient and azimuth angle of target t. The covariance matrix Q of Z -1 .

[0072] The azimuth angles θ t of T targets and the real and imaginary parts of the target reflection coefficient β t , and the estimator combined by these parameters is The fisher information matrix about the estimator can be expressed as

[0073]

[0074] ζ i represents any one of the three estimated parameters in ζ, Re represents the operation of taking the real part, and tr represents the operation of the matrix trace.

[0075] Therefore, the fisher matrix F can be expressed as

[0076]

[0077] Among them, the internal elements can be expressed as

[0078]

[0079] Denote the target receiving matrix A r The derivative with respect to the azimuth angle, Denote the target transmitting matrix A t The derivative with respect to the azimuth angle, B * Denote the conjugate matrix of matrix B, R * Denote the conjugate matrix of matrix R, Denote The conjugate transpose matrix of, Denote the target transmitting matrix A t The conjugate transpose matrix of, B denotes the diagonal matrix composed of target reflection vectors.

[0080] The relationship between the estimated CRB and the fisher matrix is

[0081] φ = F -1 , φ represents the inverse of the fisher matrix, and F represents the CRB matrix.

[0082] 2 Construct an optimized mathematical model

[0083] 2.1 When adopting the strategy of restricting the receiving rate of the eavesdropper, the integrated communication and sensing base station uses beamforming technology and artificial noise to interfere with the eavesdropper to achieve physical layer security, and at the same time uses the public signal and artificial noise signal in RSMA for collaborative sensing.

[0084] 2.1.1 After we adopt the scheme of restricting the receiving rate of the eavesdropper, maximizing the minimum user receiving rate energy efficiency can model the problem as:

[0085]

[0086] s.t.(C1):

[0087] (C2): C k ≥0

[0088] (C3):

[0089] (C4):

[0090] (C5):

[0091] (C6):

[0092] (C7):

[0093] (C8):

[0094] (C9):

[0095] (C10):

[0096] (C11): rank(W c ) = 1, rank(W k ) = 1, rank(V) = 1

[0097] wherein, R c,k represents the public rate received by user k from the base station, R p,k represents the private rate received by user k, C k represents the information belonging to user k in the public signal, represents the total rate belonging to user k received by user k from the base station, R c,m represents the eavesdropping rate of the public signal by eavesdropper m, R p,k,m represents the eavesdropping rate of the private signal of user k by eavesdropper m, represents the threshold of the public rate, represents the threshold of the private rate, R th represents the user communication rate threshold, |φ| represents the determinant of the target sensing parameter CRB, represents the threshold of the CRB, W c represents the covariance matrix of the public signal transmission vector, W k represents the private rate transmission covariance matrix of user k, V represents the artificial noise covariance matrix, and P represents the transmission power constraint of the base station.

[0098] Among them, C1 represents the synthesis of the public information of all users into the public signal stream, C2 constrains that the information of each user in the public signal stream is non - negative, C3 and C4 indicate that the public rate and private rate of user k are greater than the public threshold and private threshold respectively, C5 and C6 indicate that the eavesdropping rate of the public signal by eavesdropper m is lower than the public threshold and the eavesdropping rate of the private signal of user k by eavesdropper m is lower than the private threshold, C7 indicates that the communication rate of user k is greater than the communication rate threshold, C8 represents the constraint on the determinant of the parameter estimation CRB for target sensing, C9 represents the transmission power constraint, and C10 and C11 indicate that the transmission covariance matrix is positive semi - definite and of rank 1.

[0099] The above optimization objective is a fractional structure and cannot be directly solved. The Dinkelbach algorithm is used to convert the fraction into an integer. The new problem after conversion can be expressed as

[0100]

[0101] such that (C1):

[0102] (C2): C k ≥ 0

[0103] (C3):

[0104] (C4):

[0105] (C5):

[0106] (C6):

[0107] (C7):

[0108] (C8):

[0109] (C9):

[0110] (C10):

[0111] (C11): rank(W c ) = 1, rank(W k ) = 1, rank(V) = 1

[0112] where f(W c , W k , V) = min(C k + R p,k ), λ1 is a slack variable and is updated in the iteration according to the following formula

[0113]

[0114] Since the optimization objective and the constraints C1, C3 - C7 and C11 are non - convex and cannot be solved directly. To solve the problem, first decompose R p,k to obtain

[0115]

[0116] and are both standard concave functions, and the SCA technique is used to relax , and at this time it can be expressed as

[0117]

[0118] (·) (j) Represents the value of the j-th iteration.

[0119] For R c,k Splitting it gives

[0120]

[0121] Where

[0122]

[0123] In the above formula and are concave functions with respect to W c , W k , V. Subtracting two concave functions results in the convexity and concavity of R c,k being undetermined. Therefore, it is necessary to perform a convex approximation on and take the upper bound using the SCA technique to obtain

[0124]

[0125] Splitting R c,m gives

[0126]

[0127] The above are respectively equal to

[0128]

[0129] and Both are concave functions with respect to W c , W i , V. Using a relaxation scheme for gives

[0130]

[0131] Similarly, R p,k,m can be disassembled and expressed as

[0132]

[0133] For performing relaxation, it can be expressed as

[0134]

[0135] After the above relaxation, the original problem can be rewritten as:

[0136]

[0137] s.t.(C1):

[0138] (C2):C k ≥0

[0139] (C3):

[0140] (C4):

[0141] (C5):

[0142] (C6):

[0143] (C7):

[0144] (C8):

[0145] (C9):

[0146] (C10):

[0147] (C11):

[0148] (C12):rank(W c )=1,rank(W k )=1,rank(V)=1

[0149] where, represents the upper bound of the first-order Taylor expansion, represents the upper bound of the first-order Taylor expansion, represents the upper bound of the first-order Taylor expansion, represents the upper bound of the first-order Taylor expansion;

[0150]

[0151]

[0152] The above problem can be solved using the SDR technique in the CVX toolbox after discarding the constraint C11.

[0153] 2.2 When the secrecy capacity is used as the physical layer security performance metric, the optimization problem is to maximize the energy efficiency of the minimum user secrecy capacity, and this problem can be expressed as

[0154]

[0155] such that (C1):

[0156] (C2):

[0157] (C3):

[0158] (C4):

[0159] (C5):

[0160] (C6):

[0161] (C7):

[0162] (C8):

[0163] (C9): rank(W c ) = 1, rank(W k ) = 1, rank(V) = 1

[0164] wherein, represents the secrecy capacity of the information occupied by user k in the public signal, represents the secrecy capacity of the private information of user k, R c,k represents the public information reception rate of user k, R p,k represents the private information reception rate of user k, R c,m represents the public information reception rate of eavesdropper m, R p,k,m the reception rate of eavesdropper m for wiretapping the private information of user k, |φ| represents the determinant of the target sensing parameter CRB, represents the threshold of CRB, W c represents the covariance matrix of the common signal transmission vector, W k represents the covariance matrix of the private rate transmission of user k, V represents the artificial noise covariance matrix, and P represents the maximum transmit power constraint of the base station. represents the threshold of the secrecy capacity.

[0165] Among them, C1 represents that the sum of the secrecy rates of the public information of all users is less than the secrecy rate of the public information, C2 represents the non - negative constraint on the secrecy capacity of the public information of user k, and the information of each user in the public signal flow is non - negative. C3 represents the limit on the secrecy capacity of the private information of user k, C4 represents the non - negative constraint on the secrecy rate of the public information of user k, C5 represents the total constraint on the secrecy capacity of user k, C6 represents the constraint on the determinant of the CRB of the parameter estimation for the target perception, C7 represents the transmit power constraint, and C8 and C9 represent that the transmit covariance matrix is positive semi - definite and of rank 1.

[0166] The above - mentioned problem is non - convex itself and is difficult to solve directly. By using fractional transformation, we can obtain

[0167]

[0168] s.t.(C1):

[0169] (C2):

[0170] (C3):

[0171] (C4):

[0172] (C5):

[0173] (C6):

[0174] (C7):

[0175] (C8):

[0176] (C9):rank(W c )=1,rank(W k )=1,rank(V)=1

[0177] Here λ2 is an iterative variable, and the update mode is:

[0178] SR (j) represents the SR value of the j - th iteration.

[0179] After the previous relaxation, we can solve and obtain

[0180]

[0181] s.t.(C1):SR - λ2g2(W c ,W k,V)≥t2

[0182] (C2):

[0183] (C3):

[0184] (C4):

[0185] (C5):

[0186] (C6):

[0187] (C7):

[0188] (C8):

[0189] (C9):

[0190] (C10):rank(W c )=1,rank(W k )=1,rank(V)=1

[0191] At this time, this problem discards C9 through the SDR method and is solved using the CVX toolbox.

[0192] The above solutions all discard the rank-1 constraint and do not conform to the actual situation. Starting from the solution obtained in step 2, after relaxing the rank-1 constraint, the penalty function method is used to iteratively gradually obtain the solution that meets the requirements. If the matrix rank is 1, then only one of its eigenvalues is non-zero, and the rest of the eigenvalues are zero. Therefore, the trace of the rank-1 matrix must be equal to its eigenvalue.

[0193]

[0194] The above χ c , χ k , χ V respectively represent the largest eigenvalues of the covariance matrices W c , W k , V. u c , u k , u V respectively represent the eigenvectors corresponding to the largest eigenvalues of the covariance matrices W c , W k , V.

[0195] When the rank of the solved covariance matrix is greater than 1, there must be tr(W c )>(u c ) H W c uc , tr(W k ) > (u k ) H W k u k , tr(V) > (u V ) H Vu V , set the penalty function factor, take their difference as the penalty function in the objective function. At this time, the rank-1 constraint is also transformed into the penalty function in the objective function, and then an iterative solution method is adopted.

[0196] 3.1 The optimization problem in the scenario where the eavesdropper's receiving rate is limited can be rewritten as:

[0197]

[0198] s.t. s.t. (C1):

[0199] (C2): C k ≥0

[0200] (C3):

[0201] (C4):

[0202] (C5):

[0203] (C6):

[0204] (C7):

[0205] (C8):

[0206] (C9):

[0207] (C10):

[0208] (C11):

[0209] where () (j) represents the j-th iteration.

[0210] 3.2 When the secrecy capacity is used as the security performance, the optimization problem can be rewritten as:

[0211]

[0212] s.t. (C1):

[0213] (C2):

[0214] (C3):

[0215] (C4):

[0216] (C5):

[0217] (C6):

[0218] (C7):

[0219] (C8):

[0220] (C9):

[0221] (C10): rank(W c ) = 1, rank(W k ) = 1, rank(V) = 1

[0222]

[0223] Table 1

[0224] 4. For the solved covariance matrix value, by performing eigenvalue decomposition on it

[0225] R = DΣD -1

[0226] Σ represents the diagonal matrix composed of eigenvalues, and D represents the matrix composed of eigenvectors corresponding to the eigenvalues.

[0227] The beamforming vector is the eigenvector D(λ max ) corresponding to the largest eigenvalue λ max .

[0228] RSMA-WV is the collaborative sensing scheme of artificial noise and RSMA common signal proposed by us. RSMA-V adopts a proprietary radar signal sensing scheme. RSMA-W adopts the RSMA common signal sensing and uses the artificial noise scheme at the same time. RSMA-NV represents the sensing using only the RSMA common signal without the artificial noise scheme. SDMA-V represents that the SDMA system adopts a proprietary radar signal sensing scheme.

[0229] Figure 2 is the simulation diagram of Optimization Algorithm 1. In the figure, as the CRB threshold increases, the energy efficiency of the maximum and minimum user security rates gradually increases. Comparing several schemes, it is obvious that the scheme we proposed has better performance.

[0230] Figure 3 It is the simulation diagram of the optimization algorithm 2. In the figure, as the CRB threshold increases, the energy efficiency of the maximum and minimum user secrecy capacities gradually increases. Compared with the above several schemes, the performance of the scheme we proposed is better. Among them, the optimization algorithms 1 and 2 respectively refer to using the P1 problem and the P2 problem in the method of the present invention.

Claims

1. A communication-sensing integrated waveform design method for realizing physical layer security and target awareness in an RSMA system, characterized in that It includes the following steps: Construct a communication system model: The communication system model includes a joint communication and sensing base station E, which uses RSMA technology to serve K users simultaneously. At the same time, there are M external eavesdroppers eavesdropping on the legitimate information, and the base station needs to sense T targets simultaneously; determine the positions and channels of the K legitimate users, determine the positions and channels of the M eavesdroppers, and know the azimuth and reflection coefficient of the targets according to the forward experience information; the base station E transmits artificial noise to interfere with the eavesdroppers, and at the same time uses artificial noise and the RSMA common signal to cooperate as sensing signals to realize the sensing of T targets; Construct an optimization model: When adopting the security performance index that the eavesdroppers' common signal eavesdropping rate and private signal eavesdropping rate are not higher than the threshold, with the goal of maximizing the minimum secure transmission energy efficiency of the system users, construct optimization model Ⅰ with the common rate and private rate constraints as the constraints; When adopting the secrecy capacity as the security performance index, with the goal of maximizing the minimum secure transmission secrecy capacity energy efficiency of the system users, construct optimization model Ⅱ with the rate constraints of the eavesdroppers and legitimate users as the constraints; For the above optimization model Ⅰ and optimization model Ⅱ, respectively use the methods of successive convex optimization and semidefinite relaxation, discard the rank-1 constraint to transform the unsolvable non-convex problem into a convex problem, and use the CVX toolbox to solve to obtain the optimal solution; Starting from the above optimal solution, relax the rank-1 constraint and change it to a penalty function in the optimization objective; Use the iterative algorithm to solve the base station transmission covariance matrix, and then use eigenvalue decomposition to obtain the transmission vector.

2. The integrated communication and sensing waveform design method for realizing physical layer security and target awareness in an RSMA system according to claim 1, wherein: The above optimization model Ⅰ is: (C2):C k ≥0 (C8):|φ|≤θ (C11): rank(W c ) = 1, rank(W k ) = 1, rank(V) = 1 Where, R c,k represents the public rate received by user k from the base station, R p,k represents the private rate received by user k, C k represents the information of user k in the public signal, represents the total rate of user k received by user k from the base station, R c,m represents the eavesdropping rate of the public signal by eavesdropper m, R p,k,m represents the eavesdropping rate of the private signal of user k by eavesdropper m, represents the threshold of the public rate, represents the threshold of the private rate, R th represents the threshold of the user communication rate, |φ| represents the determinant of the target sensing parameter CRB, θ represents the threshold of the CRB, W c represents the covariance matrix of the public signal transmission vector, W k represents the covariance matrix of the private rate transmission of user k, V represents the covariance matrix of the artificial noise, P represents the transmission power constraint of the base station; Among them, C1 represents the synthesis of the common information of all users into the common signal stream, C2 constrains that the information of each user in the common signal stream is non-negative, C3 and C4 represent that the common rate and private rate of user k are greater than the common threshold and private threshold respectively, C5 and C6 represent that the eavesdropping rate of the common signal by eavesdropper m is lower than the common threshold and the eavesdropping rate of the private rate of user k is lower than the private threshold, C7 represents that the communication rate of user k is greater than the communication rate threshold, C8 represents the constraint on the determinant of the parameter estimation CRB for target sensing, C9 represents the transmission power constraint, and C10 and C11 represent that the transmission covariance matrix is positive semidefinite and rank-1.

3. The integrated communication and sensing waveform design method for realizing physical layer security and target awareness of the RSMA system according to claim 2, characterized in that: The above optimization model Ⅰ uses the Dinkelbach algorithm to transform the fraction into an integer, and the transformed new problem is expressed as: (C2):C k ≥0 (C8):|φ|≤θ (C11): rank(W c ) = 1, rank(W k ) = 1, rank(V) = 1 Among them λ1 is an iterative update variable, (·) (j) represents the value of the j-th iteration.

4. The integrated communication and sensing waveform design method for realizing physical layer security and target awareness of the RSMA system according to claim 3, characterized in that: After relaxation processing, introducing the relaxation variable t, problem (P1.1) is transformed into (P1.2): (C3):C k ≥0 (C8):|φ|≤θ (C11): rank(W c ) = 1, rank(W k ) = 1, rank(V) = 1 where denotes the upper bound of the first-order Taylor expansion, denotes the upper bound of the first-order Taylor expansion, denotes the upper bound of the first-order Taylor expansion, denotes the upper bound of the first-order Taylor expansion, The above problem is a standard convex problem after discarding C11 and is solved using CVX.

5. The integrated communication and sensing waveform design method for realizing physical layer security and target awareness of the RSMA system according to claim 1, characterized in that: The above optimization model Ⅱ is: (C6):|φ|≤θ (C9): rank(W c ) = 1, rank(W k ) = 1, rank(V) = 1 In the formula, represents the secrecy capacity of the information occupied by user k in the public signal, represents the secrecy capacity of the private information of user k, R c,k represents the public information reception rate of user k, R p,k represents the private information reception rate of user k, R c,m represents the public information reception rate of eavesdropper m, R p,k,m the reception rate at which eavesdropper m eavesdrops on the private information of user k, |φ| represents the determinant of the target perception parameter CRB, θ represents the threshold of CRB, W c represents the covariance matrix of the public signal transmission vector, W k represents the covariance matrix of the private rate transmission of user k, V represents the artificial noise covariance matrix, and P represents the maximum transmit power constraint of the base station; represents the threshold of the secrecy capacity; Among them, C1 represents that the sum of the secrecy rates of the public information of all users is less than the secrecy rate of the public information, C2 represents the non - negative constraint on the secrecy capacity of the public information of user k, and the information of each user in the public signal stream is non - negative. C3 represents the secrecy capacity limit of the private information of user k, C4 represents the non - negative constraint on the secrecy rate of the public information of user k, C5 represents the total constraint on the secrecy capacity of user k, C6 represents the constraint on the determinant of the CRB of the parameter estimation of the target perception, C7 represents the transmit power constraint, and C8 and C9 represent that the transmit covariance matrix is positive semi - definite and of rank 1.

6. The integrated communication and sensing waveform design method for achieving physical layer security and target awareness in an RSMA system according to claim 5, characterized in that: The optimization model II can be obtained by fractional transformation as follows: (C6): |φ| ≤ θ (C9): rank(W c ) = 1, rank(W k ) = 1, rank(V) = 1 where λ2 is an iterative variable, and the update mode is: SR (j) represents the SR value at the j-th iteration.

7. A method for integrated communication and sensing waveform design to achieve physical layer security and target awareness in an RSMA system, characterized in that: By introducing a slack variable t2, problem (P2.1) is transformed into (P2.2): s.t.(C1):SR-λ2g2(W c ,W k ,V)≥t2 (C7): |φ| ≤ θ (C9):W c > 0, W k > 0, V > 0 (C10): rank(W c ) = 1, rank(W k ) = 1, rank(V) = 1 denote the upper bound of the first-order Taylor expansion denote the lower bound of the first-order Taylor expansion denote the upper bound of the first-order Taylor expansion denote the upper bound of the first-order Taylor expansion 8. The integrated communication and sensing waveform design method for realizing physical layer security and target awareness of the RSMA system according to claim 1, characterized in that: The relaxation of the rank - 1 constraint and its transformation into a penalty function in the optimization objective is specifically to relax the rank - 1 constraint into the difference between the trace of the matrix and the product of the largest eigenvector of the matrix and the matrix, and then transform it into a penalty function in the optimization objective.

9. An integrated communication and sensing system for realizing physical layer security and target awareness of an RSMA system using the method according to any one of claims 1-8, characterized in that: including Initialization module: used to determine the positions and channels of K legitimate users, determine the positions and channels of M eavesdroppers, and according to the prior experience information, determine the azimuth and reflection coefficient of the target, and determine the security strategy, that is, whether to adopt the security performance index that the eavesdropping rates of the public and private signals of the eavesdroppers are not higher than the threshold, or to adopt the secrecy capacity as the security performance index; Mathematical model construction module: used to construct an optimization model according to the determined security strategy. When adopting the security performance index that the eavesdropping rates of the public and private signals of the eavesdroppers are not higher than the threshold, aiming at maximizing the minimum secure transmission energy efficiency of the system users, construct optimization model I with the public rate and private rate constraints; When adopting the secrecy capacity as the security performance index, aiming at maximizing the minimum secure transmission secrecy capacity energy efficiency of the system users, construct optimization model II with the rate constraints of the eavesdroppers and legitimate users; Optimization model solving module: used to adopt the methods of successive convex optimization and semi - definite relaxation according to the optimization model I and optimization model II respectively, discard the rank - 1 constraint to transform the intractable non - convex problem into a convex problem, and use the CVX toolbox to solve to obtain the optimal solution; Iterative solving module: used to take the above - mentioned optimal solution as the initial point, relax the rank - 1 constraint and turn it into a penalty function in the optimization objective, and continuously solve iteratively until the final result converges to a convergence value; Transmit vector acquisition module: used to solve the base - station transmit covariance matrix by using an iterative algorithm, and then perform eigenvalue decomposition to obtain the transmit vector.

10. The integrated communication and sensing system for realizing physical layer security and target awareness of the RSMA system according to claim 9, characterized in that: The transmit vector acquisition module performs eigenvalue decomposition according to the iteratively obtained covariance matrices of the public signal, private signal, and artificial noise, and the eigenvector corresponding to the largest eigenvalue is the beam transmit vector.

Images