An isac system security transmission method for illegal ris scene

Abstract

The application discloses an ISAC system security transmission method for an illegal RIS scene, and the method constructs an ISAC system based on RIS assistance to simultaneously complete communication and sensing tasks. The sensing signal can effectively interfere with the IRIS-assisted eavesdropping link while completing the sensing task. In order to more effectively cope with the system security performance deterioration challenge brought by IRIS, a beamforming matrix of the radar signal is designed. Considering the uncertainty of the accurate position of the eavesdropper (Eve), the ergodic security rate is used to represent the security performance of the system, and an approximate ergodic eavesdropping rate is derived. In order to efficiently solve the two non-convex optimization problems, the ergodic objective function is appropriately converted, and then an efficient alternating optimization algorithm is proposed to solve each optimization variable. It is verified that the joint beamforming and RIS reflection design of the application is more effective in coping with the eavesdropping process of IRIS, and it is revealed that the radar beamforming design plays a key role in alleviating the influence of IRIS.

Description

TECHNICAL FIELD

[0001] The application relates to an ISAC system security transmission method for an illegal RIS scene and belongs to the technical field of wireless communication. BACKGROUND

[0002] In recent years, driven by the demand for high-precision applications, the industry has begun to explore the construction of integrated sensing and communication networks. This has led to the emergence and development of a new concept: Integrated Sensing and Communication (ISAC). According to the main design focus, ISAC systems can be divided into three types: Communication-Centric (C&C) design, Radar-Centric (R&C) design, and Joint Waveform design. In the C&C design, sensing is integrated into the communication platform, with communication performance being prioritized. In contrast, the R&C design prioritizes sensing, embedding information into the sensing waveform to achieve communication without significantly affecting sensing performance. However, neither of these two designs achieves a balance between sensing and communication, so the development of the third design approach has been accelerated. This method focuses on tailoring new ISAC waveforms that are not dependent on existing communication or radar signals. Therefore, it provides higher degrees of freedom to simultaneously achieve communication and sensing functions. In this ISAC system, communication and radar systems share the same spectrum resources and can assist each other. This approach significantly improves the system's spectrum, energy, and hardware utilization, while providing additional integration and assistance gains.

[0003] Although ISAC brings performance improvements, it still faces challenges such as data transmission security. These challenges are mainly caused by the broadcast and superposition characteristics of wireless communication systems. Specifically, the security of wireless transmission systems depends on the difference in channel gain between the Base Station (BS) and the legitimate user / Eavesdropper (Eve). Current technologies enhance Physical Layer Security (PLS) by amplifying this difference, such as transmit beamforming, cooperative jamming, and relaying. However, the use of a large number of active interference and relays can result in significant energy consumption and high hardware costs. In addition, under unfavorable wireless propagation conditions, these techniques are not effective in enhancing PLS. Fortunately, the recently emerging Reconfigurable Intelligent Surface (RIS) technology provides a potential groundbreaking solution. Essentially, RIS is a planar array composed of many passive reflecting elements that can control the direction of reflected signals by adjusting the phase shift of incident signals. Therefore, reflected signals from RIS can be constructively or destructively combined at the desired receivers.

[0004] Recently, many studies have introduced RIS into ISAC systems to improve their security performance. However, while RIS has advantages such as low cost and reconfigurability, it also poses a serious potential risk. Specifically, a potential attacker can reduce the security performance of a legitimate communication link or enhance the eavesdropping effect of an eavesdropping link by correctly deploying RIS, thereby posing a serious security threat to wireless communication systems. Thus, the concept of illegal RIS (IRIS) has gradually emerged. Many recent studies have further explored the role of IRIS in facilitating interference attacks. In addition to interference attacks, IRIS can also be used to facilitate eavesdropping activities. Notably, when the direct link of BS-Eve is deteriorated or blocked, the channel reconfiguration capability of IRIS becomes crucial to assist the eavesdropping process. These studies highlight the security challenges currently faced by wireless communication systems. Although some scholars have proposed an artificial noise (AN) based countermeasure to address the eavesdropping threat caused by IRIS, this scheme significantly limits the potential performance of the system in power-limited situations. Therefore, an AN-based interference scheme is needed that can effectively interfere with IRIS while minimizing the "waste" of AN. The current ISAC system is well suited to complete this task. It can use radar signals to locate IRIS and simultaneously interfere with potential attackers, thereby enhancing the security performance of the system while improving the utilization of radar signals. However, there is currently no related research addressing the signal leakage problem caused by IRIS in the ISAC system. SUMMARY

[0005] The present application aims to overcome the defects and shortcomings of the prior art and provides an ISAC system security transmission method for illegal RIS scenarios. This method can use sensing signals to detect targets while interfering with IRIS and its auxiliary devices. The method can also analyze echo signals from around the IRIS to locate the IRIS and effectively counter potential eavesdroppers assisted by IRIS.

[0006] The technical scheme adopted by the present application to solve its technical problems is as follows: an ISAC system security transmission method for illegal RIS scenarios, which comprises the following steps:

[0007] Step 1: Establish a system model including BS, RIS, IRIS, users, and sensing targets, where the BS is equipped with M root transmitting and receiving antennas of the BS, and each antenna is a uniform linear array with a half-wavelength spacing. The BS completes the communication and sensing tasks simultaneously with the help of RIS.

[0008] Step 2: From the perspective of the legitimate device, construct an associated BS transmit / receive beamforming matrix and an associated optimization problem of RIS phase shift to maximize the system security rate, while meeting the system target detection requirement, BS transmit power budget and the unit constant modulus constraint of each RIS phase shift. From the perspective of eavesdropping, construct an optimization problem of maximizing the eavesdropping rate by adjusting the IRIS reflection phase shift, and meet the unit constant modulus constraint of IRIS phase shift.

[0009] Step 3: According to the range of possible positions of Eve, derive the ergodic security rate of the system. And use the closed-form fractional programming (CFFP) to convert the optimization problem of the legitimate angle into a form convenient for solving. The same method can be used to process the optimization problem of the eavesdropping angle.

[0010] Step 4: Use an alternating optimization method of joint Taylor approximation, minimax algorithm (MM) and alternating optimization multiplier method (ADMM) to solve the BS transmit / receive beamforming matrix and RIS phase shift.

[0011] Step 5: According to the algorithm proposed in step 4, simulate and analyze how to design the ISAC system to obtain better security performance.

[0012] Further, the above step 1 of the present application establishes an ISAC system composed of an ISAC BS with M transceiver antennas, a single N-element RIS, a single Ne-element IRIS, a point target, a single antenna Eve and a group of K single antenna users. The ISAC BS simultaneously completes the tasks of communication with users and sensing a point target. To avoid being monitored, Eve hides behind an obstacle and only uses the cascaded link BS-IRIS-Eve constructed by IRIS to complete the eavesdropping activity. The dual-function signal transmitted by the BS in the lth time slot of the system model of the present application is:

[0013] x[l]=W c s c [l]+W r s r [l]=Ws[l]

[0014] Wherein contains the communication symbols of K users, which satisfies contains M independent radar waveforms, and has It is assumed that the communication and sensing symbols do not interfere with each other, and respectively represent the beamforming matrices of communication and radar. In particular, in order to facilitate the representation of the transmitted signal matrix, we define a joint signal vector and a joint beamforming matrix W=[W c Wr ], where w i Let be the i-th column of the joint beamforming matrix W.

[0015] We assume the channel gain G,H of the communication link. d,i ,h i,e ,h d,k ,h r,k for All are subject to Rice decay, such as h i,e The expression is in κ and h represent path loss and Rice factor, respectively. (i,e)LoS It is the Loss-of-Stake (LoS) fading component that depends on the problem geometry, h (i,e)NLoS This represents the NLOS fading component with zero mean and unit variance. Specifically, Where ζ0 and These represent the path loss coefficient and the path loss exponent, respectively. Where θ IE The guide vector represents the angle of arrival between IRIS and Eve. For this communication system, we assume that BS, under existing advanced channel estimation methods, except for h i,e Apart from that, all other channels are known. Radar sensing channels are all considered line-of-sight links. For example, h d,t =α dt a M (θ1),h r,t =α rt a N (θ2), where θ1 and θ2 are the angles of arrival of the sensed target relative to the BS and RIS, respectively. Under this model, the system's safety rate can be expressed as follows:

[0016]

[0017] Where h k (φ)=h d,k +G T Φ T h r,k The phase shift matrix Φ = diag(φ) represents the channel from the base station to the user, where φ = [φ1, φ2, ..., φ]. N ] T ,and Similarly, It is the channel gain between IRIS and Eve. It is the channel gain between BS and IRIS, and the phase shift matrix Φ of the IRIS reflection. e =diag(φ e ), where φ e= [phi1, phi2,..., phiN] (1) Ne ] T , and Scalar n k [l] and n e [l] are additive white Gaussian noise generated in the communication and eavesdropping processes, respectively.

[0018] The BS analyzes the received target reflection echo signal, and the target detection probability P D and the relationship between each variable:

[0019]

[0020] wherein is the additive white Gaussian noise generated in the sensing process, denotes the receive beamforming matrix of the BS.

[0021] Further, the expression of the above step 2 of the present application for establishing the maximum system security rate problem from the legal angle is as follows:

[0022]

[0023] wherein the objective function represents the security rate when traversing the possible positions of Eve; constraint C1 represents the constraint on the target detection probability; constraint C2 represents the unit constant modulus constraint on the RIS phase shift; and constraint C3 represents the constraint on the total power of the BS transmission.

[0024] Further, the optimization problem of the above step 2 of the present application for establishing the maximum eavesdropping rate from the eavesdropping angle is as follows:

[0025] wherein constraint C1 represents the unit constant modulus constraint on the IRIS phase shift.

[0026] Further, the above step 3 of the present application for deriving the conversion problem P1 is as follows:

[0027] An approximation of the traversing security rate is made, and CFFP is used to convert the problem P1 into a form that has been solved. Since only the eavesdropping rate part exists in the traversal, according to

[0028] Proposition 1: Let and be the sum terms of non-negative random variables x i and y i , then the following approximation holds

[0029]

[0030] The accuracy of this approximation increases with K x and Ky increases. So there is

[0031]

[0032] When Eve is uniformly distributed in the plane region , we can get the approximate eavesdropping rate as

[0033]

[0034] Similarly, we can get the approximate eavesdropping rate for other probability distribution.

[0035] We use CFFP transformation to convert the objective function into a form that is easy to solve:

[0036] where γ k and y k are two auxiliary variables introduced in the CFFP transformation process, and const(1) and const(2) are constant terms independent of variables. Similarly, we can get the CFFP transformation of problem P2.

[0037] The eavesdropping rate can be converted from non-convex to convex by introducing two exponential auxiliary variables C e = [C e,1 , C e,2 ,..., C e,K ] T and D e = [D e,1 , D e,2 ,..., D e,K ] T To complete the conversion of the eavesdropping rate function from non-convex to convex, we first define:

[0038]

[0039] Therefore, problem P1 can be converted to

[0040]

[0041]

[0042] Further, the optimization process of problem P1 in step 4 of the above invention is as follows: we use the alternating optimization algorithm to optimize the converted problem P1:

[0043] Update auxiliary variables γ and y: we only need to solve the partial derivative.

[0044]

[0045] Updating beamforming matrix W: Since we have obtained the explicit objective function for the beamforming matrix W in the above CFFP transformation, we only need to convert the constraint term to convex. We need to first perform a first-order Taylor expansion on the equation on the right side of the constraint C1 at where is the optimal value at the last iteration and convert it into a second-order cone constraint.

[0046]

[0047] where Again, perform a first-order Taylor expansion on the left side of the constraint C2 at w i to get

[0048] So far, the problem of solving W can be converted into a convex problem, which can be solved by standard convex optimization techniques.

[0049] Updating BS receive beamforming matrix u: The solution of this matrix is an unconstrained optimization problem. In order to accelerate the convergence of the algorithm, we solve the following unconstrained problem to update u:

[0050]

[0051] It can be found that this is a typical Rayleigh quotient problem, and its optimal solution is

[0052]

[0053] Updating RIS phase shift φ: Since the RIS phase shift φ is not included in the eavesdropping rate of the objective function, and the constraints C1, C2, C5 are all irrelevant to the variable φ. And the above CFFP transformation has obtained the explicit objective function for the phase shift φ, so we only need to convert the constraint term to convex. We first use the MM algorithm to find an approximate convex constraint for the radar perception constraint C3, and then use the ADMM algorithm to construct the Lagrangian function to solve the unit constant modulus constraint C4.

[0054] Specifically, first convert the radar perception constraint:

[0055]

[0056] where Then use the MM algorithm to find a linear lower bound for the non-convex term u H Lvec(φφ H ) on the left side of the inequality. Let L = (u H L) T , Again, perform a first-order Taylor expansion on A first-order Taylor expansion is performed at the point phi, to obtain the following relationship:

[0057]

[0058] where lambda is the maximum eigenvalue of the matrix , U = [I N jI N ] H At this time, the original radar constraint can be written as:

[0059]

[0060] where

[0061] Next, the ADMM algorithm is used to construct a Lagrangian function to solve the unit constant modulus constraint C4, and the planning problem is solved as follows:

[0062] Obviously, it is a convex problem.

[0063] Further, the above step 5 of the present application can obtain better security performance according to the proposed algorithm, joint BS transmit / receive beamforming matrix, and RIS phase shift design. Meanwhile, the key role of the perception beamforming matrix design in mitigating the impact of IRIS eavesdropping is disclosed.

[0064] Advantages:

[0065] 1. The present application makes full use of the characteristics of the ISAC system, so that the perception signal can complete the perception task while interfering with the potential devices assisted by IRIS, and at the same time, the security performance of the system is improved by deploying a legal RIS.

[0066] 2. The present application finds a suitable solution for the communication and perception beamforming matrix. In order to simulate the real IRIS reflection phase shift, the present application also completes the phase shift design of IRIS from the perspective of eavesdropping to maximize the eavesdropping rate. Through simulation verification, the present application can effectively improve the security performance of the system in the power limited scene. BRIEF DESCRIPTION OF DRAWINGS

[0067] Figure 1 A security ISAC system model under an illegal RIS scenario is provided.

[0068] Figure 2 A whole beam diagram under normal operation of the system is provided.

[0069] Figure 3 A convergence diagram of the security rate that the system can achieve under various condition parameters is provided.

[0070] Figure 4 A safety rate diagram that each optimization scheme can obtain under different numbers of IRIS reflecting units is provided for the application. DETAILED DESCRIPTION

[0071] In order to make the purposes, technical solutions and advantages of the application clearer, the application will be described in detail below with reference to the drawings and specific embodiments.

[0072] As Figure 1 shown, the application establishes an ISAC system composed of an ISAC BS with M transceiving antennas, a single N-element RIS, a single N-element IRIS, a point-like sensing target, a single-antenna Eve and a K-user group of single antennas. The ISAC BS simultaneously completes the tasks of communication with the users and sensing a point-like target. In order to avoid being monitored, Eve hides himself behind an obstacle and only completes the eavesdropping activity through the cascaded link BS-IRIS-Eve constructed by IRIS. e

[0073] The dual-function signal transmitted by the system model BS in the lth time slot is:

[0074] x[l]=W c s c [l]+W r s r [l]=Ws[l]

[0075] where contains the communication symbols of K users, satisfying contains M independent radar waveforms, and has Assuming that the communication and sensing symbols do not interfere with each other, there are respectively represent the beamforming matrices of the communication and radar. In particular, in order to facilitate the representation of the transmitted signal matrix, we define a joint signal vector and a joint beamforming matrix W=[W c W r ], where w i is the i-th column of the joint beamforming matrix W.

[0076] We assume that the channel gains G, H d,i ,h i,e ,h d,k ,h r,k of the communication link satisfy are subject to Rayleigh fading, as h i,e is expressed as where and κ represent the path loss and the Rayleigh factor, respectively, and h (i,e)LoS ​It is the Loss-of-Stake (LoS) fading component that depends on the problem geometry, h (i,e)NLoS This represents the NLOS fading component with zero mean and unit variance. Specifically, Where ζ0 and These represent the path loss coefficient and the path loss exponent, respectively. Where θ IE The guide vector represents the angle of arrival between IRIS and Eve. For this communication system, we assume that BS, under existing advanced channel estimation methods, except for h i,e Apart from that, all other channels are known. Radar sensing channels are all considered line-of-sight links. For example, h d,t =α dt a M (θ1),h r,t =α rt a N (θ2), where θ1 and θ2 are the angles of arrival of the perceived target relative to BS and RIS, respectively.

[0077] Under this model, the system's safe rate can be expressed in the following form:

[0078]

[0079] Where h k (φ)=h d,k +G T Φ T h r,k The phase shift matrix Φ = diag(φ) represents the channel from the base station to the user, where φ = [φ1, φ2, ..., φ]. N ] T ,and Similarly, It is the channel gain between IRIS and Eve. It is the channel gain between BS and IRIS, and the phase shift matrix Φ of the IRIS reflection. e =diag(φ e ), where φ e =[φ1,φ2,...,φ Ne ] T ,and scalar n k [l] and n e [l] represents the additive white Gaussian noise generated during communication and eavesdropping.

[0080] By analyzing the received target reflection echo signal, the BS can obtain the target detection probability P under this ISAC system. D Relationship with each variable

[0081]

[0082] where is the additive white Gaussian noise generated by the sensing process, denotes the receive beamforming matrix of the BS.

[0083] Step 2 The expression of the maximization system security rate problem in the legitimate angle is as follows

[0084]

[0085] where the objective function represents the security rate when traversing the possible positions of Eve; constraint C1 represents the constraint on the target detection probability; constraint C2 represents the unit constant modulus constraint of the RIS phase shift; and constraint C3 represents the constraint of the total power of the BS transmission.

[0086] Step 2 The optimization problem of maximizing the eavesdropping rate from the eavesdropping angle is as follows:

[0087]

[0088] where constraint C1 represents the unit constant modulus constraint of the IRIS phase shift.

[0089] Step 3 The process of deriving the conversion problem P1 is as follows:

[0090] An approximation of the traversing security rate is made, and the CFFP is used to convert the problem P1 into a form that has been solved. Since there is only a traversing part of the eavesdropping rate, according to

[0091] Proposition 1: Let and be the summation terms of non-negative random variables x i and y i , then the following approximation is obtained

[0092]

[0093] The accuracy of this approximation increases with the increase of K x and K y . Therefore, there is

[0094]

[0095] When Eve is uniformly distributed in the planar region , the probability distribution can be obtained according to

[0096]

[0097] Similarly, the approximate traversing eavesdropping rate under other probability distribution conditions can be obtained.

[0098] The CFFP transformation is used to convert the objective function into a form that is easy to solve:

[0099] where γ k and y k are two auxiliary variables introduced in the CFFP transformation, and const(1) and const(2) are constant terms independent of the variables. Similarly, the CFFP transformation for problem P2 can be obtained.

[0100] The eavesdropping rate can be converted from non-convex to convex by introducing two exponential auxiliary variables C e = [C e,1 , C e,2 ,..., C e,K ] T and D e = [D e,1 , D e,2 ,..., D e,K ] T To complete the conversion of the eavesdropping rate function from non-convex to convex, we first define:

[0101]

[0102] Therefore, problem P1 can be converted to

[0103]

[0104]

[0105] The optimization process for problem P1 in Step 4 is as follows. We use the alternating optimization algorithm to optimize the converted problem P1:

[0106] Update auxiliary variables γ and y: We only need to solve the partial derivatives.

[0107]

[0108] Update the beamforming matrix W: Since we have obtained an explicit objective function for the beamforming matrix W in the above CFFP transformation, we only need to convert the constraint term to convexity. We need to first perform a first-order Taylor expansion of the expression on the right side of the constraint C1 at , where is the optimal value at the previous iteration, and convert it into a second-order cone constraint form.

[0109]

[0110] where Then perform a first-order Taylor expansion of the left side of the constraint C2 at w i , to obtain

[0111] At this point, the problem of solving W can be transformed into a convex problem, which can then be solved using the CVX toolbox through standard convex optimization techniques.

[0112] Updating the BS receiving beamforming matrix u: Solving this is an unconstrained optimization problem. To accelerate algorithm convergence, we solve the following unconstrained problem to update u:

[0113]

[0114] This can be seen as a typical Ruili quotient problem, and its optimal solution is:

[0115]

[0116] Update the RIS phase shift φ: Since the RIS phase shift φ is not included in the eavesdropping rate of the objective function, and constraints C1, C2, and C5 are all independent of the variable φ, and the explicit objective function regarding the phase shift φ has already been obtained in the above CFFP transformation, we only need to make the constraint terms convex. We first use the MM algorithm to find an approximate convex constraint for the radar sensing constraint C3, and then use the ADMM algorithm to construct the Lagrangian function to solve the unit constant modulus constraint C4.

[0117] Specifically, the radar perception constraints are first transformed:

[0118]

[0119] in Then, the MM algorithm is used to find the non-convex term u on the left side of the inequality. H Lvec(φφ H Find a linear lower bound. Let L = (u H L) T , and Again Performing a first-order Taylor expansion at point φ yields the following relationship:

[0120]

[0121] Where λ is a matrix The largest eigenvalue, U = [I N jI N ] H At this point, the original radar constraint can be written as:

[0122]

[0123] in

[0124] Next, the Lagrangian function is constructed by the ADMM algorithm to solve the unit constant modulus constraint C4, and the planning problem is solved as follows:

[0125] Obviously, it is a convex problem.

[0126] Step 5 According to the proposed algorithm, the joint BS transmit / receive beamforming matrix and the RIS phase shift design can obtain better security performance. The key role of the sensing beamforming matrix design in mitigating the IRIS impact is revealed.

[0127] The preferred embodiments of the present application have been specifically described above, but the present application is not limited to the embodiments described, and those skilled in the art can make various equivalent modifications or replacements without departing from the spirit of the present application. These equivalent modifications or replacements are all included in the scope defined by the claims of the present application.

Claims

1. A secure transmission method for an ISAC system in an illegal RIS scenario, characterized in that, The method includes the following steps: Step 1: Establish a system model including BS, RIS, IRIS, users and sensing targets. The BS is equipped with M transmit and receive antennas, and each antenna is a uniform linear array with half-wavelength spacing. The BS completes communication and sensing tasks simultaneously with the help of RIS. Step 2: From the perspective of legitimate equipment, construct an optimization problem related to the transmit / receive beamforming matrix of the associated BS and the phase shift of the RIS to maximize the system security rate, while simultaneously satisfying the system target detection requirements, the BS transmit power budget, and the unit constant mode constraint of each RIS phase shift. From the perspective of eavesdropping, construct an optimization problem to maximize the eavesdropping rate by adjusting the IRIS reflection phase shift, while satisfying the unit constant mode constraint of the IRIS phase shift. Step 3: Based on the possible location range of Eve, derive the system's traversal safety rate, and use closed fractional programming (CFFP) to transform the optimization problem of legal angles into a form that is easy to solve. The same method can be used to handle the optimization problem of eavesdropping angles. Step 4: Solve the transmit / receive beamforming matrix of BS and the RIS phase shift using an alternating optimization method combining Taylor approximation, minimization-maximization algorithm (MM), and alternating optimization multiplier method (ADMM); Step 5: Based on the algorithm proposed in Step 4, simulate and analyze how to design the ISAC system to achieve better security performance.

Images