Intelligent reflector assisted low latency wireless communication and secure offload method and system

Abstract

The invention belongs to the field of network security, and particularly relates to an intelligent reflecting surface (RIS)-assisted low-delay wireless communication and secure unloading method and system, and the method comprises the following steps: S1, constructing an intelligent reflecting surface (RIS)-assisted mobile edge computing (MEC) system, modulating a signal to be transmitted, and transmitting the modulated signal to an RIS and an edge node through a free transmission channel space; s2, establishing an MEC safety unloading optimization method according to an optimization target of maximizing the signal quality of a receiving end and minimizing the bit error rate; s3, executing a beam forming vector optimization method, wherein an output element is a complex vector w; s4, taking the beam forming vector w obtained in the step S3 as an input parameter, executing an RIS phase shift matrix optimization method, and taking an output element as a complex matrix q; and S5, taking w and q obtained in the steps S3 and S4 as input parameters, and calculating the maximum secrecy rate Rsec of the current system. According to the invention, the secrecy performance of the MEC system assisted by the intelligent reflecting surface can be enhanced, and the technical effect of optimal compromise between system delay and secure communication is realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of network security, and particularly relates to an intelligent reflecting surface (RIS)-assisted low-latency wireless communication and secure offloading method and system. Background Art

[0002] The increasingly rich application requirements and extreme performance requirements of future mobile communications will continuously promote the development and upgrade of mobile edge computing (MEC) technology. Among them, key technologies such as computing offloading, resource allocation, cache management, and security protection play important roles in efficiently allocating network resources and meeting the digital requirements of 6G full connectivity. However, due to the openness of the wireless transmission medium, information is easily eavesdropped. Usually, the secrecy rate (SR) is used to measure the secrecy performance of a wireless communication system, that is, the difference between the legitimate user information rate and the eavesdropper eavesdropping rate. However, in the prior art, the focus is mainly on resource allocation algorithms rather than ensuring the security of the MEC server. In the edge computing architecture, performing calculations near the data source is an effective method to protect privacy and data security. However, due to the limited resources of network edge devices, existing data security protection methods are not fully applicable to the MEC architecture. Therefore, the prior art has problems such as the latency rate of the communication system while ignoring security.

[0003] Chinese Patent Publication No. CN118632254A discloses a low-latency secure computing offloading method for an RIS-assisted MEC system. After simplifying the problem through variable substitution, this technical solution uses convex optimization methods such as alternating optimization, quadratic transformation, and Lagrangian dual transformation. However, the iterative process using methods such as alternating optimization and Lagrangian dual transformation may lead to high computational complexity, especially in large-scale networks, which will affect real-time performance. Summary of the Invention

[0004] Aiming at the defects existing in the prior art, the present invention provides an intelligent reflecting surface-assisted low-latency wireless communication and secure offloading method and system.

[0005] The present invention adopts the following technical solutions:

[0006] First, define the following sub-methods, and the detailed steps will be given respectively later: the optimization method of the beamforming vector w; the optimization method of the RIS reflection phase shift q.

[0007] An intelligent reflecting surface-assisted low-latency wireless communication and secure offloading method of the present invention specifically includes the following steps:

[0008] S1: Construct a mobile edge computing (MEC) system assisted by a reflecting intelligent surface (RIS). Modulate the signal to be transmitted and then send it to the RIS and the edge node;

[0009] S2: Based on the optimization goal of maximizing the signal quality at the receiving end and minimizing the bit error rate, establish an MEC secure offloading optimization method;

[0010] S3: Execute the beamforming vector optimization method, and output a complex vector w as an element;

[0011] S4: Take the beamforming vector w obtained in step S3 as an input parameter, execute the RIS phase shift matrix optimization method, and output a complex matrix q as an element;

[0012] S5: Take w and q obtained in steps S3 and S4 respectively as input parameters, and calculate the maximum secrecy rate R of the current system sec 。

[0013] Preferably, in step S1: Construct a mobile edge computing (MEC) system assisted by a reflecting intelligent surface (RIS). The user performs modulation on the signal to be transmitted through technologies such as quadrature amplitude modulation (QAM), and then sends it to the RIS and the edge node through the free transmission channel space.

[0014] Preferably, the RIS-assisted MEC system in step S1 includes a base station with a positive integer M number of antennas, a RIS with a positive integer N number of reflection units, an edge node, multiple single-antenna users, and eavesdroppers. Assume that the user / RIS jointly designs the transmit / reflection beamforming.

[0015] Preferably, in step S2, to establish the MEC secure offloading optimization method, the following steps are adopted:

[0016] S2.1: The user transmits the confidential information with zero mean and cell variance to the edge node through beamforming. The beamforming vector is represented as w and satisfies the following constraint conditions:

[0017] ||w|| 2 ≤P U (1)

[0018] Among them, while represents the set of M×1 dimensional composite matrices, ||.|| represents the two-norm, and P U is the maximum transmit power of the user. Use to simulate the RIS reflection. Moreover, in order to obtain the maximum reflection power gain, q should satisfy:

[0019]

[0020] wherein, θ n ∈[0, 2π), β n ∈[0, 1], n = 1, ... N, respectively represent the phase shift and amplitude reflection coefficient of the nth unit. The transpose operation is denoted by T. For ease of operation, it is stipulated that β n = 1, Each transmitting unit of the RIS has an adjustable phase shift to eliminate the received signal offset.

[0021] S2.2: At the edge node and the eavesdropper, the received signal is:

[0022] s E = (h RE Qh UR + h UE )ws + n E (3)

[0023] s ED = (h RED Qh UR + h UED )ws + n ED (4)

[0024] wherein, respectively represent the channel coefficients from the user to the RIS, from the user to the eavesdropper, from the RIS to the eavesdropper, from the RIS to the edge node, and from the user to the edge node, represents the set of M×N dimensional complex-valued matrices. When the eavesdropper is an active user of the system but not trusted by the legitimate receiver, let h UED and h RED be known and reasonable. represents a diagonal matrix, and the diagonal elements are the corresponding elements of the vector q. n U , n E respectively represent the Gaussian noise at the edge node and the eavesdropper, with a mean of 0 and a variance of and Therefore, the secrecy rate from the user to the edge node is expressed as:

[0025] R sec = [R E - R ED + (5)

[0026] wherein, R sec is in units of bits per second per hertz (bps / Hz), ​"+" ensures that the final result is non - negative. According to the formula for calculating the Signal to Noise Ratio (SNR), SNR = signal power / noise power. Then the signal - to - noise ratios at the edge node and the eavesdropper are as follows:

[0027]

[0028]

[0029] According to Shannon's formula: R = log 2 (1 + SNR), then we have:

[0030]

[0031] where R E and R ED represent the achievable rates of the legitimate link and the eavesdropping link respectively. To maximize the transmission beamforming vector for a given user, for the given user transmission beamforming vector w, the reflection beamforming vector q needs to satisfy the following two conditions: the reflected channel h RE Qh UR is aligned with the direct channel h UE to maximize the received signal power of the user and R E ; the reflected channel h RED Qh UR is opposite to the direct channel h UED of the eavesdropper to cancel the signal, so as to minimize R ED .

[0032] S2.3: Model the system secrecy optimization problem as:

[0033]

[0034] s.t.

[0035] C1: ||w|| 2 ≤P U

[0036]

[0037] where the constraint C1 ensures that the transmit power of the user is within the range specified by the system, and the constraint C2 ensures that the amplitude of the RIS reflection unit is fixed at 1 and only the phase can be adjusted.

[0038] Preferably, in step S3, the beamforming vector w optimization method specifically adopts the following steps:

[0039] First, fix the adjustable phase shift q of the RIS, and introduce two intermediate variables X and Y:

[0040]

[0041] Among them, the superscript H represents the conjugate transpose operation. The optimization problem is re-described as follows:

[0042]

[0043] s.t.

[0044] C3: w H w ≤ P U (13)

[0045] Solving equation (13) gives the optimal solution as:

[0046]

[0047] Among them, j max is the normalized eigenvector corresponding to the largest eigenvalue of the matrix . I M represents an m×m identity matrix.

[0048] Preferably, in step S4, the RIS phase shift matrix optimization method specifically adopts the following steps:

[0049] Fix the beamforming vector w and optimize the RIS reflection unit. Since:

[0050] h RE Qh UR = q T diag(h UE )h UR (15)

[0051] h RED Qh UR = q T diag(h UED )h UR (16)

[0052] The problem is re-described as:

[0053]

[0054] s.t. (10)(17) Therefore, the following equation holds:

[0055]

[0056] Among them, s = [q T , 1] T represents the beamforming vector at the transmitter or the weight vector at the receiver, and there is:

[0057]

[0058] Among them, the superscript "*" represents the conjugate operation.

[0059]

[0060] Among them, G E and G ED respectively represent the effective signal gain matrix received at the legitimate user and the interference signal gain matrix between the eavesdropper and the legitimate user.

[0061] By substituting equations (18) and (19) into equation (17), it is rewritten in a more easily processed form:

[0062]

[0063] Among them, E n represents the noise covariance matrix, and the (i, j)-th element in E n is denoted as [E n i,j , and satisfies:

[0064]

[0065] Preferably, step S5: Solve the problem by processing the non-convexity and coupled variables in the problem through algorithms such as alternating optimization. Obtain the maximum secure rate R sec .

[0066] Preferably, in step S5, calculate the maximum secrecy rate R sec of the current system, and solve it using the following steps:

[0067] S5.1: Since each constraint in the problem is a non-convex quadratic equality constraint for n, and the objective function is a fractional function and non-concave with respect to s. Therefore, first use tr(Z) and rank(Z) to represent the trace and rank of matrix Z respectively.

[0068] Among them, Z represents an arbitrary matrix.

[0069] S5.2: Use the semidefinite relaxation technique to overcome the non-convexity of the problem, introduce Remove the rank(S)=1 constraint condition, and equation (24) is reformulated into the following relaxed form:

[0070]

[0071] S5.3: Use the Charnes-Cooper transformation, that is, set the variables Π and μ to Π = μS and μ = 1 / [tr(G ED s)+h ED ​If [ [ ID = 0 ] ] + 1, then equation (26) is converted into an equivalent non - fractional form:

[0072]

[0073] s.t.

[0074] C6: tr(G ED Π)+μ(h ED + 1)=1

[0075]

[0076] S5.4: Use the interior - point method to perform optimal solution for equation (27).

[0077] S5.5: In order to solve the omitted constraint rank(S)=1, apply the standard Gaussian randomization method to obtain an approximate solution of equation (17).

[0078] The present invention also discloses an intelligent reflecting surface - assisted low - latency wireless communication and secure offloading system for performing the above - mentioned method, which includes the following modules:

[0079] Signal transmission module: Construct an intelligent reflecting surface RIS - assisted mobile edge computing MEC system, modulate the signal to be transmitted, and then send it to the RIS and the edge node;

[0080] MEC secure offloading optimization method establishment module: Establish an MEC secure offloading optimization method according to the optimization objectives of maximizing the signal quality at the receiving end and minimizing the bit error rate;

[0081] RIS phase - shift optimization module: Execute the beamforming vector optimization method, and output a complex vector w as an element; Beamforming vector optimization module: Take the obtained beamforming vector w as an input parameter, execute the RIS phase - shift matrix optimization method, and output a complex matrix q as an element;

[0082] Output module: Take the obtained w and q as input parameters, and calculate the maximum secrecy rate R of the current system sec . The introduction of the prior art related to the present invention is as follows:

[0083] 1. Alternating optimization algorithm

[0084] The Alternating Optimization (AO) algorithm is a method widely used in multi-variable optimization problems. It gradually improves the solution by alternately fixing some variables and optimizing others. This method is particularly suitable for dealing with cases where there are complex interdependencies between variables. The core of the AO algorithm lies in decomposing complex optimization problems into more tractable sub-problems. In each iteration, the algorithm fixes a part of the variables and optimizes the others. In this way, the coupling between variables can be gradually reduced, thus approaching the global optimal solution or an approximate optimal solution. The advantage of this method is its high computational efficiency, because only a part of the variables are processed each time, thereby reducing the complexity of the problem. For details, see "R. Blanquero, E. Carrizosa, A. Jiménez-Cordero, et al. Functional-bandwidth Kernel for Support Vector Machine with Functional Data: an Alternating Optimization Algorithm [J]. European Journal of Operational Research, 2018, 275(1): 195-207".

[0085] 2. Semidefinite Relaxation Algorithm

[0086] The Semidefinite Relaxation (SDR) algorithm is a method for dealing with non-convex quadratic constraint quadratic programming (QCQP) problems. It solves the problem by converting it into a convex problem. The core of the SDR algorithm lies in transforming the original QCQP problem into a convex optimization problem, and then using the methods of convex optimization to solve it. This transformation is achieved by introducing a new variable X = xx TIt is achieved by letting \(x\) be the optimization variable of the original problem. Through this transformation, the non-convex constraints in the original problem are converted into convex constraints with respect to \(X\), and the objective function of the original problem becomes a linear function with respect to \(X\). During the solution process, first, the transformed convex problem is solved by a convex optimization method to obtain a solution \(X\). If the rank of this solution is 1, then the optimal solution \(x\) of the original problem can be directly obtained through eigenvalue decomposition. If the rank of the solution is not 1, then other methods need to be used to find an approximate solution. For details, see "LUO Z Q, MA W K, A.M.C. So, YE Y, Zhang S. Semidefinite Relaxation of Quadratic Optimization Problems[J]. in IEEE Signal Processing Magazine, vol. 27, no. 3, pp. 20 - 34, May 2010".

[0087] 3. Gaussian Randomization Method

[0088] The Gaussian Random Process (GRP) is a common random process in the fields of statistics and signal processing, where the value at each time point or spatial point follows a Gaussian distribution (i.e., normal distribution). The Gaussian random process can be stationary (or weakly stationary), which means that its mean and covariance functions do not depend on the absolute position of time but only on the difference between time points. Any linear combination of it is also Gaussian distributed.

[0089] 4. Interior Point Method

[0090] The Interior Point Method (IPM) is a class of numerical optimization algorithms for solving linear programming, non-linear programming, and other optimization problems. Usually, a barrier function is constructed, and this function will increase rapidly when approaching the boundary, thus avoiding hitting the constraint boundary during the iteration process. The use of the barrier function makes the optimization problem easier to handle because in each iteration, the solution will be kept inside the feasible region. For details, see "Lin T, Ma S, Ye Y, Zhang S. An ADMM-based interior-point method for large-scale linear programming. Optimization Methods and Software, 36(2–3): 389–424.".

[0091] 5. Charnes-Cooper Transformation

[0092] The Charnes-Cooper Transformation is a technique widely used in operations research and optimization theory. It is mainly used to transform the constraint conditions of a linear programming problem into an equivalent standard form problem, usually to simplify the problem-solving process. This transformation makes the originally complex constraint problem into an easily solvable form by making appropriate substitutions for the variables in the original problem. For details, see "HUANG B D, SHEN P P. An efficient branch and bound reduction algorithm for globally solving linear fractional programming problems[J]. Chaos, Solitons &

[0093] Fractals, Volume 182, 2024, 114757, ISSN 0960-0779.”

[0094] 6. QAM Transformation

[0095] Quadrature Amplitude Modulation (QAM) is a modulation technique widely used in communication systems, especially in digital signal transmission. QAM combines amplitude modulation (AM) and phase modulation (PM) and transmits information by modulating both the amplitude and phase on the same signal simultaneously. The signal is modulated by two orthogonal carriers, namely the sine wave and the cosine wave. Each carrier represents a different signal component (usually the I component and the Q component). Each symbol is represented as a specific amplitude and phase combination, usually shown on a coordinate graph called a "constellation diagram". The points in the diagram represent different symbols, and each point corresponds to a specific I and Q value. In QAM, each point in the constellation diagram represents a possible symbol value. The higher the density of the points in the constellation diagram, the more data can be transmitted, but the sensitivity to noise also increases.

[0096] In the present invention, the RIS reflection phase shift matrix is first fixed, and the successive convex approximation (SCA) method is used to solve the optimal transmit beamforming vector. Secondly, the beamforming vector is fixed, and the semi-definite relaxation (SDR) method is used to solve the RIS reflection phase shift matrix. Finally, the optimal secrecy energy efficiency value is obtained, realizing a low-latency wireless communication and secure offloading scheme.

[0097] In summary, the present invention can enhance the secrecy performance of the intelligent reflecting surface-assisted MEC system, achieving the optimal trade-off technical effect between system latency and secure communication. Description of the Drawings

[0098] Figure 1 It is a model diagram of the RIS-assisted MEC communication system according to the preferred embodiment of the present invention;

[0099] Figure 2 It is an optimization flowchart for beamforming vector optimization, RIS phase shift matrix optimization, and calculation latency in the RIS-assisted MEC low-latency secure offloading method according to the preferred embodiment of the present invention;

[0100] Figure 3 It is a flowchart of the steps for optimizing the transmit beamforming vector according to the preferred embodiment of the present invention;

[0101] Figure 4 It is a flowchart of the steps for optimizing the RIS reflection phase shift matrix according to the preferred embodiment of the present invention;

[0102] Figure 5 It is a comparison chart of the algorithm convergence performance under different schemes;

[0103] Figure 6 It is a comparison chart of the secrecy performance with the increase of the number of RIS reflection units under different schemes;

[0104] Figure 7 It is a system block diagram of the RIS-assisted MEC low-latency secure offloading method according to the preferred embodiment of the present invention. Detailed Embodiments

[0105] The present invention will be further described below in conjunction with specific embodiments. The specific embodiments of the present invention can be detailedly illustrated by the following embodiment diagrams.

[0106] Figure 1 It is a model diagram involved in the RIS-assisted low-latency wireless communication and secure offloading method according to the embodiment of the present invention. The system includes a base station with M antennas, a RIS with N reflection units, an edge node, multiple single-antenna users, and eavesdroppers. The users and the RIS jointly design the transmit and reflection beamforming.

[0107] A method for intelligent reflecting surface-assisted low-latency wireless communication and secure offloading in this embodiment includes the following steps:

[0108] S1: Construct a mobile edge computing MEC system assisted by an intelligent reflecting surface RIS (see Figure 1 ), modulate the signal to be transmitted, and then send it to the RIS and the edge node;

[0109] S2: Based on the optimization objective of maximizing the signal quality at the receiver and minimizing the bit error rate, establish an MEC secure offloading optimization method; specifically, it is implemented through the following steps:

[0110] S2.1: The user transmits the confidential information with zero mean and cell variance to the edge node through beamforming. The beamforming vector is denoted as w, which satisfies the following constraint conditions:

[0111] ||w|| 2 ≤P U (1)

[0112] where, represents the set of M×1 dimensional composite matrices, ||.|| represents the second norm, and P U is the maximum transmit power of the user; adopt to simulate RIS reflection; q satisfies:

[0113]

[0114] where, θ n ∈[0, 2π), β n ∈[0, 1], n = 1,... N, respectively represent the phase shift and amplitude reflection coefficient of the nth unit; T represents the transpose operation; assume β n = 1, each transmitting unit of the RIS has an adjustable phase shift;

[0115] S2.2: At the edge node and the eavesdropper, the received signal is:

[0116] s E =(h RE Qh UR +h UE )ws + n E (3)

[0117] s ED =(h RED Qh UR +h UED )ws + n ED (4)

[0118] where, respectively represent the channel coefficients from the user to the RIS, from the user to the eavesdropper, from the RIS to the eavesdropper, from the RIS to the edge node, and from the user to the edge node, represents the set of M×N dimensional complex-valued matrices; when the eavesdropper is an active user but not trusted by the legitimate receiver, assume h UED and h RED are known; denotes a diagonal matrix, and the diagonal elements are the corresponding elements of the vector q; n U , n E respectively represent the Gaussian noise of the edge node and the eavesdropper, with a mean of 0 and a variance of and Therefore, the secrecy rate from the user to the edge node is expressed as:

[0119] R sec = E -R ED + (5)

[0120] where R sec is in bits per second per hertz (bps / Hz), + ensures that the final result is non - negative; according to the calculation formula of the signal - to - noise ratio SNR, SNR = signal power / noise power; then the signal - to - noise ratios at the edge node and the eavesdropper are:

[0121]

[0122] According to the Shannon formula: R = log 2 (1 + SNR), then there is:

[0123]

[0124] where R E and R ED respectively represent the achievable rates of the legitimate link and the eavesdropping link; the reflection beamforming vector q satisfies the following two conditions: the reflection channel h RE Qh UR is aligned with the direct channel h UE to maximize the received signal power of the user and R E ; the reflection channel h RED Qh UR is opposite to the direct channel h UED of the eavesdropper, to minimize R ED ;

[0125] S2.3: Model the secrecy optimization problem as:

[0126]

[0127] s.t.

[0128] C1: ||w|| 2 ≤ P U

[0129]

[0130] ​Among them, constraint C1 ensures that the transmission power of the user is within the specified range, and constraint C2 ensures that the amplitude of the RIS reflection unit is fixed at 1 and only the phase can be adjusted.

[0131] S3: Execute the beamforming vector optimization method, and output the complex vector w as an element;

[0132] S4: Use the beamforming vector w obtained in step S3 as the input parameter, execute the RIS phase shift matrix optimization method, and output the complex matrix q as an element;

[0133] S5: Use w and q obtained in step S3 and step S4 respectively as input parameters to calculate the maximum secrecy rate R of the current system sec . In this step, the non-convexity and coupled variables in the problem are processed by the alternating optimization method to solve the problem and obtain the maximum secrecy rate R that satisfies all constraint conditions sec . Specifically as follows:

[0134] S5.1: Use tr(Z) and rank(Z) to represent the trace and rank of matrix Z respectively;

[0135] S5.2: Introduce Removing the rank(S)=1 constraint condition, equation (24) is reformulated into the following relaxed form:

[0136]

[0137] S5.3: Set the variables μ and Π to μ = 1 / [tr(G ED s)+h ED +1] and Π = μS respectively, then equation (26) is converted into an equivalent non-fractional form:

[0138]

[0139] s.t.

[0140] C6: tr(G ED Π)+μ(h ED +1)=1

[0141]

[0142] S5.4: Use the interior point method to perform optimal solution for equation (27);

[0143] S5.5: Use the Gaussian randomization method to obtain an approximate solution of equation (17).

[0144] Figure 2Shows the optimization process of beamforming vector optimization, RIS phase shift matrix optimization, and calculation delay in the intelligent surface reflection-assisted low-latency wireless communication and secure offloading method in the embodiments of the present invention. The specific steps are as follows:

[0145] Step 1: The user sends a signal, which is digitally modulated and transmitted through the free channel space to the intelligent reflecting surface RIS and the edge node.

[0146] Step 2: Set the initial iteration value k = 0 and the convergence constant ε = 0.01, and input the beamforming vector w (0) , the RIS phase shift matrix q (0) ; R (0) = f(w (0) , q (0) );

[0147] Step 3: Judge whether it holds, where R (k) refers to the secrecy rate calculated in the k-th iteration. If it holds, directly end; otherwise, execute Step 4;

[0148] Step 4: Obtain the beamforming vector w according to the beamforming vector optimization algorithm k ;

[0149] Step 5: Obtain the phase shift matrix q according to the phase shift matrix optimization algorithm k ;

[0150] Step 6: Output the obtained calculation delay.

[0151] Figure 3 This is the flow chart for optimizing the transmit beamforming vector in the embodiments of the present invention, which is mainly completed through the following steps:

[0152] Step 1: Randomly initialize the phase vector q;

[0153] Step 2: Construct the phase matrix Q;

[0154] Step 3: Calculate the matrices X and Y;

[0155] Step 4: Calculate the matrix C;

[0156] Step 5: Perform eigenvalue decomposition;

[0157] Step 6: Calculate the optimal weight w;

[0158] Step 7: Return the optimized weight w.

[0159] Figure 4 This is the flow chart for optimizing the RIS reflection phase shift matrix in the embodiments of the present invention, which is mainly completed through the following steps:

[0160] Step 1: Input the optimized w;

[0161] Step 2: Calculate the interference matrix G E and G ED ;

[0162] Step 3: Optimize the phase q using SDR;

[0163] Step 4: Update the phase matrix Q;

[0164] Step 5: Return the optimized phase q.

[0165] Figure 5 It is a comparison graph of the algorithm convergence performance under different scenarios. It can be seen from the graph that the method proposed by the present invention (Alternating Optimization with RIS) has a fast convergence speed and can basically achieve convergence after 20 iterations. At the same time, the transmission delay is reduced by 0.003 s compared with the control group. This verifies the effectiveness and convergence of the method of the present invention.

[0166] Figure 6 It is a comparison graph of the secrecy performance when the number of RIS reflection units increases under different scenarios. The maximum power of the user is set to 15 dBm. It can be seen that the delay of the system can be reduced as the number of RIS reflection units increases, and the more users there are, the higher the secrecy energy efficiency of the system. The secrecy energy efficiency of the system with RIS deployment is much higher than that without RIS deployment, which indicates that RIS can greatly improve the energy efficiency of the system and enhance the security of the system.

[0167] As Figure 7 shown, this embodiment discloses a RIS-assisted MEC low-latency wireless communication and secure offloading system for implementing the above method embodiment, including the following modules:

[0168] Signal transmission module: Construct a mobile edge computing (MEC) system assisted by a reconfigurable intelligent surface (RIS), modulate the signal to be transmitted, and then send it to the RIS and the edge node;

[0169] MEC secure offloading optimization method establishment module: Establish an MEC secure offloading optimization method according to the optimization objectives of maximizing the signal quality at the receiving end and minimizing the bit error rate;

[0170] RIS phase shift optimization module: Execute the beamforming vector optimization method and output a complex vector w as the element;

[0171] Beamforming vector optimization module: Take the obtained beamforming vector w as the input parameter, execute the RIS phase shift matrix optimization method, and output a complex matrix q as the element;

[0172] Output module: Using the obtained \(w\) and \(q\) as input parameters, calculate the maximum secrecy rate \(R\) of the current system sec 。

[0173] For other content of this embodiment, reference can be made to the above method embodiment.

[0174] Although the embodiments of the present invention have been clearly described above. However, for those skilled in the art, without departing from the principle and spirit of the method of the present invention, various changes, modifications, substitutions, and variations can be made to these embodiments. The scope of the present invention is defined by the appended claims and their equivalents, and still belongs to the scope of the method of the present invention and is still regarded as the protection scope of the present invention.

Claims

1. Low-latency wireless communication and secure offloading method assisted by intelligent reflective surface, characterized in that: The following steps are involved: S1: Build a mobile edge computing MEC system assisted by an intelligent reflective surface RIS, modulate the signal to be transmitted, and then send it to the RIS and edge nodes; S2: Establish a MEC security offloading optimization method based on the optimization goals of maximizing the signal quality at the receiving end and minimizing the bit error rate; S3: Execute the beamforming vector optimization method, and the output element is a complex vector w; S4: using the beamforming vector w obtained in step S3 as an input parameter, executing the RIS phase shift matrix optimization method, and the output element is a complex matrix q; S5: Take w and q obtained in step S3 and step S4 respectively as input parameters to calculate the maximum confidentiality rate R of the current system sec .

2. The low-latency wireless communication and secure unloading method assisted by the intelligent reflective surface as described in claim 1 is characterized in that: In step S1, the constructed mobile edge computing MEC system assisted by the intelligent reflecting surface RIS includes a base station with a positive integer M of antennas, a RIS with a positive integer N of reflection units, an edge node, multiple single-antenna users and an eavesdropper; it is assumed that the user and the RIS jointly design the transmission and reflection beamforming.

3. The low-latency wireless communication and secure unloading method assisted by the intelligent reflective surface as described in claim 2 is characterized in that: In step S2, a MEC secure offloading optimization method is established, which is implemented by the following steps: S2.1: The user transmits confidential information with zero mean and cell variance to the edge node through beamforming. The beamforming vector is denoted as w, which satisfies the following constraints: ||in|| 2 ≤P U (1) in, represents the set of M×1 dimensional composite matrices, ||.|| represents the bi-norm, P U is the maximum transmit power of the user; Simulate RIS reflection; q satisfies: in, θ n ∈[0,2π),β n ∈[0,1], n=1,...N, respectively representing the phase shift and amplitude reflection coefficient of the nth unit; T represents the transposition operation; let β n =1, Each transmitting unit of RIS has an adjustable phase shift; S2.2: At the edge node and the eavesdropper, the received signal is: s E =(h RE Qh UR +h UE )ws+n E (3) s ED =(h RED Qh UR +h UED )ws+n ED (4) in, They represent the channel coefficients from user to RIS, from user to eavesdropper, from RIS to eavesdropper, from RIS to edge node, and from user to edge node, respectively. represents a set of M×N dimensional complex-valued matrices; when the eavesdropper is an active user but is not trusted by the legitimate receiver, let h UED and h RED is known; represents a diagonal matrix, the diagonal elements are the corresponding elements of the vector q; n U ,n E They represent the Gaussian noise of edge nodes and eavesdroppers, with a mean of 0 and a variance of and Therefore, the confidentiality rate from the user to the edge node is expressed as: R sec =[R E -R ED ] + (5) Among them, R sec The unit is bits per second per Hertz (bps / Hz). + Ensure that the final result is non-negative; According to the calculation formula of signal-to-noise ratio SNR, SNR = signal power / noise power; then the signal-to-noise ratio at the edge node and the eavesdropper is: According to Shannon's formula: R = log2 (1 + SNR), we have: Among them, R E and R ED represent the achievable rates of the legitimate link and the eavesdropped link respectively; the reflection beamforming vector q satisfies the following two conditions: the reflection channel h RE Q UR With direct channel h UE Aligned to maximize user and R E The received signal power of the reflection channel h RED Q UR Direct channel with the eavesdropper UED On the contrary, R ED Minimum; S2.3: Model the confidentiality optimization problem as: Among them, constraint C1 ensures that the user's transmission power is within the specified range, and constraint C2 ensures that the amplitude of the RIS reflection unit is fixed to 1, and only the phase can be adjusted.

4. The low-latency wireless communication and secure unloading method assisted by an intelligent reflective surface as described in claim 3 is characterized in that: In step S3, the beamforming vector optimization method is specifically implemented by the following steps: The adjustable phase shift q of RIS is fixed, and two intermediate variables X and Y are introduced; Wherein, the superscript H represents the conjugate transpose operation; the optimization problem is restated as follows: Solving equation (13) yields the optimal solution: Among them, j max is the same as the matrix The normalized eigenvector corresponding to the maximum eigenvalue; I M Represents the m×m identity matrix.

5. The low-latency wireless communication and secure unloading method assisted by an intelligent reflective surface as described in claim 4 is characterized in that: In step S4, the RIS phase shift matrix optimization method is specifically implemented by the following steps: The beamforming vector w is fixed and the RIS reflector unit is optimized because: h RE Qh UR =q T diag(h UE )h UR (15) h RED Qh UR =q T diag(h UED )h UR (16) Let's describe the problem as follows: Therefore, the following equation holds: Where s = [q T ,1] T Represents the beamforming vector at the transmitter or the weight vector at the receiver, which is: Among them, the superscript * represents the conjugate operation; Among them, G E and G ED They represent the effective signal gain matrix received by the legitimate user and the interference signal gain matrix between the eavesdropper and the legitimate user respectively; By substituting formula (18) and formula (19) into formula (17), it can be rewritten as: Among them, E n represents the noise covariance matrix, E n The (i,j)th element in is represented by [E n ] i,j ,satisfy:

6. The low-latency wireless communication and secure unloading method assisted by an intelligent reflective surface as described in claim 5 is characterized in that: In step S5, the non-convexity and coupling variables in the problem are processed by the alternating optimization method, and the problem is solved to obtain the maximum confidentiality rate R that satisfies all constraints. sec .

7. The low-latency wireless communication and secure unloading method assisted by an intelligent reflective surface as described in claim 6 is characterized in that: In step S5, an alternating optimization method is used to deal with non-convexity and coupled variables in the problem, which is specifically implemented by the following steps: S5.1: Let tr(Z) and rank(Z) denote the trace and rank of the matrix Z, respectively. S5.2: Introduction Removing the rank (S) = 1 constraint, equation (24) can be restated in a relaxed form as follows: S5.3: Variables μ and Π are set to μ = 1 / [tr(G ED s)+h ED +1] and Π=μS, then equation (26) is converted into an equivalent non-fractional form: S5.4: Use the interior point method to find the optimal solution for equation (27); S5.5: Use the Gaussian randomization method to obtain an approximate solution to equation (17).

8. A low-latency wireless communication and secure offloading system assisted by an intelligent reflective surface, used to execute the method according to any one of claims 1 to 7, characterized in that: Includes the following modules: Signal transmission module: Build a mobile edge computing MEC system assisted by an intelligent reflective surface RIS, modulate the signal to be transmitted, and then send it to the RIS and edge nodes; MEC security offloading optimization method establishment module: Establish a MEC security offloading optimization method based on the optimization goals of maximizing the signal quality of the receiving end and minimizing the bit error rate; RIS phase shift optimization module: executes the beamforming vector optimization method, and the output element is a complex vector w; Beamforming vector optimization module: takes the obtained beamforming vector w as input parameter, executes RIS phase shift matrix optimization method, and the output element is complex matrix q; Output module: Take the obtained w and q as input parameters to calculate the maximum confidentiality rate R of the current system sec .

Images