AN enhanced RIS assisted physical layer security communication robust beam forming method

Abstract

According to the robust beam forming method for the AN enhanced RIS assisted physical layer security communication, the transmitting power of a base station is minimized on the premise that the system service quality requirement and the security rate outage probability constraint are met. The method comprises the following steps: firstly, constructing a security rate interruption constraint based on a base station-RIS-eavesdropper cascade channel transmitting end imperfect CSI model, and then approaching a security rate interruption probability constraint by using a Bernstein inequality, thereby reexpressing an original problem as a processable convex constraint. Under an alternating iterative optimization framework, positive semi-definite relaxation is utilized to convert a sub-problem into a convex problem, a sub-optimization problem of a given beam forming vector optimization RIS phase shift matrix is rewritten, a slack variable is introduced to re-express an original feasibility verification problem, and finally a base station transmitting beam forming vector is iteratively optimized. The AN beam forming vector and the RIS reflection beam forming matrix improve the system safety rate and the system robustness.

Description

Technical Field

[0001] The invention relates to the technical field of wireless communications, in particular to an AN enhanced RIS assisted physical layer secure communication robust beamforming method. Background Art

[0002] 6G wireless communication networks provide higher transmission rates, lower latency, and more reliable transmission solutions, but the inherent broadcast nature of wireless channels will lead to huge eavesdropping risks in wireless communication systems. The design of secure communication solutions at the physical layer level to achieve reliable information transmission is generally based on the fact that the achievable rate of the legal channel is better than the achievable rate of the eavesdropping channel, that is, the system can obtain a non-zero secure transmission rate. The greater the quality advantage of the legal channel over the eavesdropping channel, the higher the confidentiality of the system.

[0003] In recent years, the application of smart reflective surfaces (RIS) in wireless communications has helped solve these problems and has shown great potential in improving network performance. RIS is a passive device that does not have a radio frequency unit and baseband processing circuit and only reflects wireless signals. Each passive reflective element can independently adjust the reflection coefficient of its reflective element through passive beamforming, thereby dynamically regulating the amplitude and phase shift of the incident signal, realizing intelligent control of the wireless propagation environment, and then changing the signal propagation environment, so that it can assist wireless communication systems in different scenarios and improve system performance. When the distance between the transmitter and the legitimate receiver is greater than the distance between the transmitter and the eavesdropper, or the transmission path between the transmitter and the legitimate receiver is blocked by obstacles, and the eavesdropper is hidden near the legitimate receiver, the physical layer security performance of the wireless communication system is usually poor. RIS can improve the security performance of the system at the physical layer by reconstructing the wireless propagation environment and increasing the channel difference between legitimate users and eavesdroppers. Specifically, the signal reflected by RIS can enhance the signal received by the legitimate receiver and weaken the signal received by the eavesdropper. Therefore, as an emerging wireless communication technology, RIS provides a new solution for improving physical layer security.

[0004] Although there are a lot of studies on RIS-assisted secure communications, there are some defects in the existing solutions. First, most of the literature usually assumes that both the base station (BS) and the RIS can obtain the perfect channel state information (CSI) of the eavesdropper to study the theoretical upper limit of system performance. However, due to the passive nature of RIS, it cannot send or receive signals, so it is difficult to obtain perfect CSI on the RIS side. Whether perfect CSI can be obtained or whether there is an error in CSI has a great impact on the design and performance of the security solution, and may even cause RIS to produce negative gain. Therefore, it is imperative to take imperfect CSI into consideration, and the robust transmission design of imperfect CSI at the transmitter end based on the BS-RIS-eavesdropper cascade channel is more universal than the previous research that only considers the imperfect RIS-eavesdropper channel. When the error is caused by noise in the channel estimation, the CSI error is more accurately described by the statistical CSI error model than the bounded error model. In addition, the use of artificial noise (AN) to reduce the channel quality of the eavesdropper in previous physical layer security schemes has been widely studied and has become a common solution to achieve physical layer security. Therefore, the current research on RIS-assisted MIMO system secure communication solutions should consider scenarios containing AN. In summary, the joint optimization scheme of inferring appropriate BS transmit beamforming vectors, AN beamforming vectors and RIS reflection beamforming matrices while ensuring the robustness of the imperfect CSI system has become a difficult problem to be solved in current research.

[0005] The technical differences between this application and the prior art are as follows: Technical comparison with the patent publication number CN118264295A "A robust and secure beamforming method for STAR-RIS-assisted wireless energy-carrying communication system"; Patent CN118264295A constructs a wireless energy-carrying communication system model assisted by STAR-RIS; minimizes the transmitter's transmit power by jointly designing the transmit beamforming vector and the reflection and transmission coefficients of STAR-RIS, and proposes a robust and secure beamforming design method under the uncertainty of channel state information. The alternating optimization framework is used to release the coupling variables, and the power consumption of the transmit power is significantly reduced while ensuring the user service quality and robustness. The present invention is aimed at a wireless communication system assisted by artificial noise and RIS, and through iterative optimization of the base station transmit beamforming vector, the AN beamforming vector and the RIS reflection beamforming matrix, effective interference with the channel quality of the eavesdropper under imperfect CSI conditions is achieved. In addition, the CSI error modeling in the present invention is based on a statistical CSI error model, which is more in line with the actual complex environment than the bounded uncertainty assumption of the channel state information error adopted, and is particularly more suitable for error scenarios caused by channel estimation noise. The present invention strengthens the robustness of the algorithm under imperfect CSI conditions, and at the same time has a significant improvement in the physical layer security performance. These innovations make the present invention have significant technical advantages and application potential in the classification field of wireless communication technology.

[0006] Technical comparison with patent publication number CN114268350A "Robust beamforming design method under non-ideal channels in intelligent reflector-assisted communication" Patent CN118264295A studies the weighted sum rate maximization problem in RIS-assisted multi-user millimeter wave systems. The optimization goal focuses on improving the spectrum efficiency and communication performance of the system. It performs iterative optimization by jointly optimizing the base station's transmit beamforming, RIS layout, and RIS reflection beamforming. An alternating optimization algorithm based on penalty and gradient projection is used to alleviate the performance loss caused by the impact of imperfect CSI. This patent focuses on improving the security of the system. While meeting the requirements of the safe rate quality of service, it minimizes the base station transmit power. Based on the imperfect CSI situation of the base station-RIS-eavesdropper cascade channel transmitter, a robustness optimization method is designed for physical layer security by introducing a statistical CSI error model, realizing robust communication under the constraint of safe rate interruption probability, which is particularly suitable for scenarios with high physical layer security requirements. In addition, the present invention combines artificial noise and RIS to realize robust secure communication under the constraint of safe rate interruption probability. The comprehensive optimization scheme for eavesdropper interference and imperfect CSI provides stronger protection for secure communication. Summary of the invention

[0007] To solve the above technical problems, the present invention provides an AN enhanced RIS assisted physical layer secure communication robust beamforming method, an AN beamforming vector and a RIS reflection beamforming matrix, thereby minimizing the transmission power of the system.

[0008] To achieve the above object, the technical solution adopted by the present invention is: The AN enhanced RIS-assisted physical layer secure communication robust beamforming method comprises the following steps: Step 1: construct a system model for secure communication of AN-enhanced RIS-assisted MIMO system, and construct an optimization problem with the goal of minimizing the system transmit power; Step 2, introduce the imperfect CSI situation, construct the security rate interruption constraint based on the imperfect CSI channel model of the BS-RIS-eavesdropper cascade channel transmitter, that is, the robustness constraint of the system, and approximate the security rate interruption probability constraint based on the Bernstein type inequality; Step 3: Under the AO framework, construct the given optimization Subproblems and constructs given optimization Sub-problem, iteratively optimize the BS transmit beamforming vector, AN beamforming vector and RIS reflection beamforming matrix.

[0009] As a further improvement of the present invention, in step 1, a system model of AN enhanced RIS assisted MIMO system secure communication is constructed, and the constructed system model is: ; in , , , , Respectively represent BS to RIS, BS to k users, BS to eavesdropper, RIS to k The channel from user RIS to eavesdropper, the channel model includes large-scale fading and small-scale fading. The large-scale fading model is expressed as ,in is the path loss exponent, is the link distance in meters. The small-scale fading is assumed to be Rayleigh fading distribution. The transmission signal is expressed as ,in represents the BS beamforming vector, represents the AN beamforming vector, represent information signal and AN respectively, represents the phase shift of the reflective element of the RIS, and are the noises at the user and the eavesdropper, respectively, with a mean of zero and a variance of , the achievable rates of the user and the eavesdropper are expressed as: .

[0010] As a further improvement of the present invention, in step 1, an optimization problem is constructed with the goal of minimizing the system transmission power, as follows: ; in, represents the lower bound of the user achievable rate required by the system, represents the upper bound of the eavesdropper's achievable rate required by the system, Indicates the interruption probability requirement coefficient that the system needs to meet.

[0011] As a further improvement of the present invention, in step 2, an imperfect CSI situation is introduced to construct a security rate interruption constraint based on an imperfect CSI channel model at the transmitter of the BS-RIS-eavesdropper cascade channel, wherein the error of the imperfect CSI channel model is described by a statistical CSI error model: ; in Indicates k The estimated channel of the reflected cascade channel of the eavesdropper is expressed as For the k The reflected cascade channel estimation error of each eavesdropper, each CSI error vector follows a circularly symmetric complex Gaussian distribution, that is: ; In the formula, represents the error covariance matrix.

[0012] As a further improvement of the present invention, in step 2, the safety rate interruption probability constraint is approximated based on the Bernstein type inequality, as follows: ; in , For convenience of expression, is the introduced slack variable, ,in , represents the beamforming matrix of other users except the kth user, and I represents the identity matrix.

[0013] As a further improvement of the present invention, in step 3, under the AO framework, construct a given optimization The sub-questions are as follows: ; in ,and , and use SDR to transform the subproblem into a convex problem.

[0014] As a further improvement of the present invention, in step 3, construct a given optimization Slack variables are introduced into the sub-problem, as follows, to reformulate the original feasibility verification problem: ; in is the slack variable introduced to transform the original feasible line verification problem, and .

[0015] Beneficial effects: 1. Enhanced robustness: By incorporating the probability of system safety rate interruption into constraints and selecting an appropriate modeling scheme for the channel estimation error, the robustness of the system is improved. Compared with the non-robust optimization scheme and the scheme based on the bounded CSI error model, this method has better robustness, especially when the degree of channel error uncertainty is higher, this method has obvious benefits.

[0016] 2. Improve the system safety rate: RIS and AN jointly participate in the design of system safety, which improves the system safety rate. Under the same power conditions, this method has a higher safety rate than only AN and only RIS.

[0017] 3. Reduce the impact of channel uncertainty on system performance: Compare the non-robust optimization scheme and the scheme based on the bounded CSI error model. Under the same channel uncertainty, this method requires a lower minimum transmit power to meet the system QoS requirements.

[0018] 4. Reduce the risk of RIS negative gain: By comparing the non-robust optimization scheme and the scheme based on the bounded CSI error model, this method can better suppress the impact of channel errors. When the channel error is too high, this method can better reduce the risk of RIS negative gain. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 Is a flow chart of the present invention; Figure 2 Is the communication system model diagram of the present invention; Figure 3 It is the process flow chart of P1 optimization problem; Figure 4 It is a simulation diagram comparing the BS transmit power change trend of this method with the AN-free scheme, the non-robust scheme and the scheme based on the bounded CSI error model as the probability of safe rate interruption changes; Figure 5 This method is compared with the AN-free scheme, the non-robust scheme and the scheme based on the bounded CSI error model as the channel uncertainty degree increases. Simulation diagram of the change of the minimum required BS transmission power; Figure 6It is a simulation diagram comparing the feasibility of the present method with the AN-free solution, the non-robust solution and the solution based on the bounded CSI error model as the threshold of the safety rate required to be met by the system changes; Figure 7 It is a simulation diagram comparing the transmission power change of the minimum BS with the change of the number of RIS elements among the present method, the scheme without AN, the scheme without robustness and the scheme based on bounded CSI error model. DETAILED DESCRIPTION

[0020] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments: This specific implementation provides a robust beamforming method for enhancing the communication security of an imperfect CSI RIS-assisted MIMO system using AN. Figure 1 As shown: Construct a secure communication system model of AN-enhanced RIS-assisted MIMO system and formulate an optimization problem with the goal of minimizing the system transmit power; A secure rate interruption constraint based on the imperfect CSI channel model at the transmitter end of the BS-RIS-eavesdropper cascade channel, namely the robustness constraint of the system, is constructed, and the secure rate interruption probability constraint is approximated by the Bernstein type inequality.

[0021] AO iteratively optimizes the BS transmit beamforming vector, AN beamforming vector and RIS reflection beamforming matrix to improve the system safety rate and system robustness.

[0022] Specifically, the present invention uses AN to enhance the robust beamforming method of imperfect CSI RIS-assisted MIMO system communication security, which includes the following steps: Step 1: Construct a system model for secure communication of AN-enhanced RIS-assisted MIMO system, and construct an optimization problem with the goal of minimizing the system transmit power; The system model of AN enhanced RIS assisted MIMO system secure communication is as follows: Figure 2 As shown; Define the user set as BS to RIS, BS to k users, BS to eavesdropper, RIS to k The channel coefficients from user to RIS and eavesdropper are respectively , , , , Indicates that The BS sends confidential information s with zero mean and unit variance to the user through beamforming. The BS beamforming vector is represented by , the AN beamforming vector is expressed as , and the transmission signal can be expressed as ,in Represent the information signal and AN respectively. Each unit of RIS reflects the received signal from BS with an adjustable phase shift. The phase shift of the reflective element of RIS is Indicates that , where the superscript T represents the transposition operation. And let For simplicity, we assume that the reflection coefficient of RIS is 1 to obtain the maximum reflection power gain, so q satisfy , considering the serious path loss, thus ignoring the signal reflected more than twice by RIS, the signals received by the user and the eavesdropper can be expressed as: ; and are the noises at the user and the eavesdropper, respectively, with a mean of zero and a variance of Therefore, the achievable rates of the user and the eavesdropper can be expressed as: ; The secure rate from BS to user (in bits / second / Hz) can be expressed as: .

[0023] The signal vectors of the reflected cascade channel BS-RIS-user and BS-RIS-eavesdropper at the user and eavesdropper ends can be expressed as .

[0024] Our goal is to minimize the BS transmit power by jointly optimizing BS beamforming, AN beamforming, and RIS phase shift while meeting the system QoS requirements and considering the robustness constraint of imperfect CSI. The optimization problem P1 can be expressed as follows: ; in, represents the lower bound of the user achievable rate required by the system, represents the upper bound of the eavesdropper's achievable rate required by the system, represents the interruption probability requirement coefficient that the system needs to meet. The constraint ensures that each user can The probability of successfully decoding its message at a data rate of no less than , and ensure that the eavesdropper's achievable rate is less than The probability is not less than ; Given a BS transmit beamforming vector, in order to make the confidentiality rate Maximize, RIS reflection beamforming vector The design of the reflection channel generally needs to achieve the following two goals. With direct channel Phase alignment to maximize signal power received at the user, thereby maximizing On the other hand, the reflection channel Direct channel with eavesdropper In opposite phase, to cancel the base station signal received by the eavesdropper, so that Minimum.

[0025] Step 2: Construct a secure rate interruption constraint based on the imperfect CSI channel model of the transmitter of the BS-RIS-eavesdropper cascade channel, and use the Bernstein type inequality to approximate the secure rate interruption probability constraint: The constructed imperfect CSI channel model is as follows: Assuming that the CSI of the BS-user channel and the BS-eavesdropper channel is accurate, it is a huge challenge to obtain accurate CSI at the RIS end due to the passive nature of RIS, so there are errors in the CSI of the BS-RIS-user channel and the BS-RIS-eavesdropper. Taking the eavesdropping channel as an example, its coefficient matrix (or vector) can be expressed as; ; in Indicates k The estimated channel of the reflected cascade channel of the eavesdropper is expressed as For the k The reflected cascade channel estimation error of each eavesdropper, and in general, the channel estimation error obeys Gaussian distribution. Therefore, this application considers the statistical CSI error model. Specifically, it is assumed that each CSI error vector follows a circularly symmetric complex Gaussian distribution, that is: ; In the formula, represents the error covariance matrix.

[0026] The Bernstein-type inequality is used to approximate the safety rate interruption probability constraint. Figure 3 As shown: First, constrain Rephrased, this is converted to the following: ; In the formula ,in , represents the beamforming matrix of other users except the kth user. For the convenience of derivation, we express the covariance matrix of the channel estimation error as , so the CSI error can be expressed as of the form, where , I represents the identity matrix, and the definition , the outage probability constraint can be restated as follows: ; in: ; Here we introduce the slack variable According to the characteristics of Bernstein-type inequality, the opportunity constraint of user k can be transformed into the following form: ; For the three formulas above Some simplified mathematical transformations are done: ; Through the above transformations, the original optimization problem can be transformed into the following form: ; Here ,and .

[0027] Step 3: Under the AO framework, iteratively optimize the BS transmit beamforming vector, AN beamforming vector and RIS reflection beamforming matrix to improve the system security rate and system robustness.

[0028] The details are as follows: Since the above non-convex constraints and The optimization problem is still difficult to solve due to the coupling and non-convex constraints. Therefore, we use the AO method to decouple and alternately optimize the BS transmit beamforming matrix, the AN beamforming matrix and the RIS reflection phase shift matrix.

[0029] For a given optimization , the optimization problem is transformed into the following form: ; constraint The SDR method can be used to process it, and the resulting semi-definite programming (SDP) convex problem can be effectively solved using the CVX tool.

[0030] For a given optimization Subproblem, the original problem is transformed into a feasibility check problem, in order to obtain The optimal solution of .

[0031] ; And will Slack variables are also introduced and transformed into the following form: ; Here .

[0032] In summary, the original optimization problem can be transformed into the following form: ; The SDR method is used to relax the rank-one constraint, and the CVX tool is used to solve the convex SDP problem. The Gaussian randomization technique can be used to obtain the optimal .

[0033] In the experimental scheme of the present invention, the parameters are set as K = 4 users, M = 6 antennas at the BS, N = 16 reflective elements at the RIS, and the coordinates of the BS, RIS and eavesdropper are (0, 0), (10, 50) and (70, 10) respectively. The coordinates of the four users are (50, 0), (50, 5), (50, 10) and (50, 15) respectively. The coordinate unit is meter (m), and it is assumed that the channel model contains large-scale fading and small-scale fading. The large-scale fading model is expressed as ,in is the path loss exponent, d is the link distance in meters, and the small-scale fading is assumed to be Rayleigh fading distribution. In the experiment, the eavesdropper’s eavesdropping channel statistical error distribution model for each user k is designed as ,in Indicates the uncertainty degree of imperfect CSI. In addition, the experiment in this paper also assumes that the system has the same security rate requirements for all users. The system simulation parameter table is shown in Table 1; Table 1

[0034] Figure 4 The variation trend of the BS transmission power required by the system in this paper with the probability of safe rate interruption is shown. Here, the channel uncertainty is set to , and set the target confidentiality rate to , as shown in the figure, with the safety rate interruption probability As the value of increases, the minimum BS transmission power of the system also increases. The experiment also compares the non-robust optimization scheme and the bounded CSI error model. The results show that the performance of this system is better, because the algorithm can provide users with a higher achievable rate, or provide eavesdroppers with a lower achievable rate to reduce the probability of system security rate interruption, thereby consuming less power. As shown in the figure, compared with the system without RIS and without AN, the system of this paper has better performance. Under this channel uncertainty condition, the number of RIS elements is set to 4 and 9 for comparison. The experimental results show that the minimum BS transmission power of the system will decrease with the increase of the number of RIS elements. The experimental results show that the robust design we proposed compensates for the impact of imperfect CSI at the cost of more transmission power, thereby meeting the system security rate interruption probability. requirements.

[0035] Figure 5 The minimum BS transmission power and channel uncertainty are shown in First, it can be observed that as the channel uncertainty level increases, the minimum BS transmit power also increases. Second, when there is no RIS assistance in this system, the robust beamforming design cannot effectively compensate for the loss caused by CSI errors. When the system is not AN-assisted, the performance will be degraded to a certain extent. And we can find that as the number of RIS elements increases, the system requires a larger rate of increase in the minimum BS transmit power, that is, the slope becomes larger. When the minimum BS transmission power required by the system decreases with the increase of the number of RIS elements, but As they get bigger, their size relationship may change, which also proves Figure 1 The conclusion is that too large an imperfect CSI may cause the RIS element to be unable to accurately adjust its phase shift, thus failing to obtain RIS gain. In addition, compared with the non-robust optimization scheme and the bounded CSI error model, it can be found that the algorithm proposed in this paper performs better.

[0036] Figure 6 It represents the change of the feasible rate as the threshold of the safety rate required by the system changes. The feasible rate is defined as the ratio of the number of feasible channel realizations to the total number of channels, where a feasible channel refers to a channel with a feasible solution under the requirements of power limit and safety rate interruption constraint. When the number of RIS elements N=4 is fixed, as the degree of channel uncertainty increases, the system feasible rate will gradually decrease, and when the safety rate reaches When the channel error uncertainty is fixed at 0.01, the system feasibility increases with the number of RIS elements. Compared with the non-robust optimization scheme and the bounded CSI error model, this system has a great advantage.

[0037] Figure 7 The effects of different channel uncertainty levels and changes in the number of RIS elements on system performance are shown. The case of can be regarded as an ideal CSI case, and its minimum BS transmit power decreases with the increase of the number of RIS elements N. When , the change trend of the minimum BS transmit power required by the system with the number of RIS elements is the same as that of the perfect CSI case, and is higher than the minimum BS transmit power in the perfect CSI case. The reason for this is that the BS needs to consume more power to compensate for the safety rate loss caused by the channel estimation error, and under this degree of channel error, the benefits brought by the number of RIS elements outweigh its impact. When the system's minimum BS transmit power increases gradually with the number of RIS elements, this is because the degree of imperfect CSI is large, resulting in the RIS elements being unable to accurately adjust their phase shifts, so under a higher degree of channel error, RIS may lose its performance gain advantage. Therefore, the experimental results show that for different degrees of imperfect CSI, the number of reflection elements should be carefully selected, and the accuracy of CSI estimation is crucial to obtaining RIS gain.

[0038] In summary, the present invention studies a robust beamforming method for enhancing the communication security of MIMO system with imperfect CSI RIS assisted by AN. In the case of inaccurate CSI of the eavesdropping channel, a robust security scheme is designed. The error of CSI is described by a statistical error model and is a random variable obeying a Gaussian distribution. By optimizing transmit beamforming, AN beamforming and RIS phase shift, a robust optimization problem of minimizing BS transmit power under a given safety rate interruption probability constraint is constructed, but this optimization problem is unsolvable. In order to solve this problem, slack variables are first introduced, and the Bernstein-Type inequality is used to approximate the robustness constraint, the problem is reformulated, and the sub-problem is converted into a convex problem using SDR, and under the AO framework, for a given beamforming matrix, the sub-problem of optimizing the RIS phase shift matrix is ​​introduced, and the original feasibility verification problem is reformulated by introducing slack variables. By alternately iterating and optimizing signal beamforming, AN beamforming and IRS phase shift, the problem is effectively solved. The simulation results show that the scheme has good robustness and has better performance than the non-robust optimization scheme and the scheme based on the bounded CSI error model.

[0039] The above description is only a preferred embodiment of the present invention and does not constitute any other form of limitation to the present invention. Any modification or equivalent change made based on the technical essence of the present invention still falls within the scope of protection required by the present invention.

Claims

1. AN enhanced RIS-assisted physical layer secure communication robust beamforming method, characterized by: The following steps are involved: Step 1: construct a system model for secure communication of AN-enhanced RIS-assisted MIMO system, and construct an optimization problem with the goal of minimizing the system transmit power; Step 2, introduce the imperfect CSI situation, construct the security rate interruption constraint based on the imperfect CSI channel model of the BS-RIS-eavesdropper cascade channel transmitter, that is, the robustness constraint of the system, and approximate the security rate interruption probability constraint based on the Bernstein type inequality; Step 3: Under the AO framework, construct the given optimization Subproblems and constructs given optimization Sub-problem, iteratively optimize the BS transmit beamforming vector, AN beamforming vector and RIS reflection beamforming matrix.

2. The AN enhanced RIS assisted physical layer secure communication robust beamforming method according to claim 1, characterized in that: In step 1, a system model for secure communication of AN-enhanced RIS-assisted MIMO system is constructed. The constructed system model is: ； in , , , , Respectively represent BS to RIS, BS to k users, BS to eavesdropper, RIS to k The channel from user RIS to eavesdropper, the channel model includes large-scale fading and small-scale fading. The large-scale fading model is expressed as ,in is the path loss exponent, is the link distance in meters. The small-scale fading is assumed to be Rayleigh fading distribution. The transmission signal is expressed as ,in represents the BS beamforming vector, represents the AN beamforming vector, represent information signal and AN respectively, represents the phase shift of the reflective element of the RIS, and are the noises at the user and the eavesdropper, respectively, with a mean of zero and a variance of , the achievable rates of the user and the eavesdropper are expressed as: 。 3. The AN enhanced RIS assisted physical layer secure communication robust beamforming method according to claim 1, characterized in that: In step 1, the optimization problem is constructed with the goal of minimizing the system transmit power, as follows: ； in, represents the lower bound of the user achievable rate required by the system, represents the upper bound of the eavesdropper's achievable rate required by the system, Indicates the interruption probability requirement coefficient that the system needs to meet.

4. The AN enhanced RIS assisted physical layer secure communication robust beamforming method according to claim 1, characterized in that: In step 2, the imperfect CSI situation is introduced to construct the security rate interruption constraint based on the imperfect CSI channel model of the transmitter of the BS-RIS-eavesdropper cascade channel, where the error of the imperfect CSI channel model is described by the statistical CSI error model: ； in Indicates k The estimated channel of the reflected cascade channel of the eavesdropper is expressed as For the k The reflected cascade channel estimation error of each eavesdropper, each CSI error vector follows a circularly symmetric complex Gaussian distribution, that is: ; In the formula, represents the error covariance matrix.

5. The AN enhanced RIS assisted physical layer secure communication robust beamforming method according to claim 1, characterized in that: In step 2, the safe rate interruption probability constraint is approximated based on the Bernstein type inequality, as follows: ； in , For convenience of expression, is the introduced slack variable, ,in , represents the beamforming matrix of other users except the kth user, and I represents the identity matrix.

6. The AN enhanced RIS assisted physical layer secure communication robust beamforming method according to claim 1, characterized in that: In step 3, under the AO framework, construct the given optimization The sub-questions are as follows: ； in ,and , and use SDR to transform the subproblem into a convex problem.

7. The AN enhanced RIS assisted physical layer secure communication robust beamforming method according to claim 1, characterized in that: In step 3, construct the given optimization Slack variables are introduced into the sub-problem, as follows, to reformulate the original feasibility verification problem: ； in is the slack variable introduced to transform the original feasible line verification problem, and .

Images