Random phase perturbation method for phase shift matrix optimization of ris-assisted uav communication

Abstract

The application aims at the problem of high calculation complexity caused by the coupling of RIS phase shift matrix and unmanned aerial vehicle position height, and provides a random phase disturbance method for RIS assisted unmanned aerial vehicle communication phase shift matrix optimization, belonging to the field of wireless communication, which comprises the following steps: 1) establishing a MISO communication channel model according to the base station position, user position, RIS position, base station antenna quantity and RIS reflection element quantity, including the expression of the baseband signal received by the user, the channel model from the base station to the user, the channel model from the base station to the RIS and the channel model from the RIS to the user; 2) optimizing the RIS phase shift matrix in the expression of the baseband signal, randomly disturbing the solution of the RIS phase shift matrix, and continuously iterating and optimizing to obtain the solution of the optimal RIS phase shift matrix; 3) calculating the transmitting end active beamforming matrix according to the solution of the optimal RIS phase shift matrix, and realizing the communication between the unmanned aerial vehicle and the ground user. Compared with the commonly used convex optimization method, the application reduces the calculation amount and calculation cost.

Description

TECHNICAL FIELD

[0001] The present application relates to a step-reduced random phase disturbance method, in particular to a random phase disturbance method for RIS-assisted UAV communication phase shift matrix optimization. BACKGROUND

[0002] With the increasing demand for future wireless communication, unmanned aerial vehicles (UAVs) play an important role in improving the performance of communication systems and have been widely used in various scenarios such as emergency communication. Reconfigurable intelligent surfaces (RISs) integrate many reconfigurable passive reflecting elements, each of which can independently change the amplitude and phase of the incident electromagnetic wave. By jointly designing the phase shift of each element of the RIS, the phase alignment of different transmission path signals at the target receiver can be achieved, and the wireless transmission environment can be intelligently changed.

[0003] In complex urban environments, the line-of-sight (LoS) transmission between the UAV and the ground user is easily subject to random shielding, reducing the communication quality. Based on this, researchers consider using RIS-assisted UAV communication to utilize the passive beamforming characteristics of RIS when reflecting electromagnetic waves to construct a virtual LoS link, and to improve the communication rate of the ground user by jointly optimizing the RIS phase shift matrix and the UAV trajectory. In addition, the RIS-assisted UAV communication system can also provide secure communication for the ground user in the presence of eavesdroppers, and assist in data transmission and collection between devices in the Internet of Things.

[0004] In the optimization of such problems, the RIS phase shift matrix and the UAV position are closely coupled together, and involve many nonlinear and non-convex objective functions and constraints. To solve this problem, researchers often use the alternating optimization (AO) method to solve the target problem by splitting it into two sub-problems of RIS phase shift matrix design and UAV trajectory. When the reconstructed sub-problems are still non-convex problems, although methods such as successive convex approximation (SCA) and semi-definite relaxation (SDR) can obtain better optimization solutions, they will produce a large amount of calculation and beam training overhead. Therefore, it is particularly important to propose a simple and easy-to-implement RIS phase shift matrix optimization method with low computational overhead. SUMMARY

[0005] The present application aims at the deficiencies in the prior art, and provides a random phase perturbation method for RIS-assisted UAV communication phase shift matrix optimization, which reduces the calculation overhead of RIS phase shift matrix optimization through the method of random phase perturbation in view of the high calculation complexity caused by the coupling of RIS phase shift matrix and UAV position height.

[0006] To achieve the above object, the present application adopts the following technical solutions:

[0007] A random phase perturbation method for RIS-assisted UAV communication phase shift matrix optimization, comprising the following steps:

[0008] Step 1: establishing a MISO communication channel model for serving K users according to the base station position, user position, RIS position, base station antenna quantity and RIS reflecting element quantity, wherein the MISO communication channel model comprises a user-received baseband signal expression, a base station-to-user channel model, a base station-to-RIS channel model and a RIS-to-user channel model;

[0009] Step 2: optimizing the RIS phase shift matrix in the user-received baseband signal expression, randomly perturbing the solution of the RIS phase shift matrix, and continuously iterating and optimizing to obtain the optimal solution of the RIS phase shift matrix;

[0010] Step 3: calculating a transmitting active beamforming matrix according to the optimal solution of the RIS phase shift matrix, and realizing the communication between the UAV and the ground user.

[0011] To optimize the above technical solutions, the specific measures taken further include:

[0012] Further, in step 1, the MISO communication channel model is specifically:

[0013] If the RIS surface is carried by the UAV, the baseband signal y received by the kth user is represented as: k

[0014] y k = (h RIS-k ΦH BS-RIS +h BS-k )g k x k +n k (13)

[0015] In the formula, denotes a RIS-to-kth user Rayleigh fading channel model, denotes a base station-to-kth user Rayleigh fading channel model, denotes a base station-to-RIS Rayleigh fading channel model, denotes a RIS phase shift matrix, and θ N ​φn represents the phase shift of the Nth RIS reflecting element, x k denotes the transmitted baseband signal, satisfying denotes the transmit-side beamforming matrix; n k is a complex additive white Gaussian noise with mean 0 and variance σ 2 N is the total number of RIS reflecting elements, and M is the number of base station antennas, denotes an N x M-dimensional complex number set;

[0016] The Rayleigh fading channel model of the base station to the kth user is specifically:

[0017]

[0018] In the formula, ρ is the path loss under the reference distance D0=1m, a is the corresponding path loss index, denotes a random scattering component, which is subject to a complex Gaussian distribution with mean 0 and variance 1;

[0019] The Rician fading channel model of the base station to the RIS is:

[0020]

[0021] In the formula, K1 represents the K factor in the Rician fading channel, denotes the line-of-sight transmission component, denotes the non-line-of-sight transmission component, and the elements thereof are independent and identically distributed random variables, subject to a complex Gaussian distribution with mean 0 and variance 1;

[0022] The Rician fading channel model of the RIS to the kth user is specifically:

[0023]

[0024] In the formula, K2 represents the K factor in the Rician fading channel, denotes the line-of-sight transmission component, denotes the non-line-of-sight transmission component, and the elements thereof are independent and identically distributed random variables, subject to a complex Gaussian distribution with mean 0 and variance 1;

[0025] The N-element uniform linear array adopted at the base station has an array response represented by the following formula:

[0026]

[0027] In the formula, d represents the antenna element spacing, λ represents the wavelength of the transmitted electromagnetic wave, and θ represents the beam angle.

[0028] According to the array response of the N-element uniform linear array, the line-of-sight transmission component of the Rician fading channel of the base station to the RIS is

[0029]

[0030] wherein, denotes the beam transmit angle at the BS, denotes the angle of arrival of the BS transmit beam at the RIS;

[0031] The line-of-sight transmission component of the RIS to the kth user's Rayleigh fading channel is obtained according to the array response of an N-element uniform linear array

[0032]

[0033] wherein, denotes the beam transmit angle at the RIS for the kth user;

[0034] The maximum ratio transmission is adopted to improve the signal power; when transmitting with power p k , the transmit beamforming matrix g k is represented as:

[0035]

[0036] The kth user's received signal-to-noise ratio γ k is:

[0037]

[0038] In the formula, g i represents the transmit beamforming matrix corresponding to the ith user, σ 2 is the variance.

[0039] Further, step 2 specifically includes:

[0040] Step 2.1, generate N random numbers θ0= [θ 01 , θ 02 ,..., θ 0N ] on [0, 2π) obeying uniform distribution as the initial solution of the RIS phase shift matrix, and take as the initial RIS phase shift matrix, calculate the initial system communication efficiency R0;

[0041] Step 2.2, let the initial value of the iteration number t = 0, and the initial value of the disturbance failure number C T = 0;

[0042] Step 2.3, judge whether the disturbance failure number C T is greater than or equal to the threshold value τ; if C T ≤ τ, randomly perturb the elements in the solution θ t of the RIS phase shift matrix by a step size δ to obtain a new solution θt+1 ; if R T > τ, then perturb the elements in the solution θ D of the RIS phase shift matrix with step size R t δ, and obtain a new solution θ t+1 ; where R D ∈ (0, 1) represents the decay coefficient of the perturbation step size;

[0043] Step 2.4, calculate the new solution θ t+1 corresponding to the system communication efficiency R t+1 ; if R t+1 > R t , then the perturbation is successful, and the corresponding θ t+1 is retained; t = t + 1; if R t+1 ≤ R t , then the perturbation fails, θ t+1 = θ t , R t+1 = R t , C T = C T + 1, t = t + 1;

[0044] Step 2.5, determine whether t reaches the maximum number of iterations, if so, the iteration is terminated, if not, return to step 2.3.

[0045] Further, in step 2.3, the random phase perturbation of the elements in the solution θ t of the RIS phase shift matrix with step size R D δ is specifically:

[0046] An integer 1 is randomly generated with a probability of 0.5, an integer -1 is randomly generated with a probability of 0.5, and the generated integer is denoted as d; each element θ i in the solution θ t of the RIS phase shift matrix is randomly perturbed according to the following formula:

[0047] θ i,t+1 = θ i,t + dδ (22)

[0048] All elements are randomly perturbed to obtain a new solution θ t+1 .

[0049] Further, in step 2.3, the random phase perturbation of the elements in the solution θ t of the RIS phase shift matrix with step size R D δ is specifically:

[0050] An integer 1 is randomly generated with a probability of 0.5, an integer -1 is randomly generated with a probability of 0.5, and the generated integer is denoted as d; each element θ i in the solution θ t of the RIS phase shift matrix is randomly perturbed according to the following formula:t each element of i Random phase perturbation is performed:

[0051] θ i,t+1 = θ i,t +dR D delta (23)

[0052] where R D ∈(0,1), represents the attenuation coefficient of the perturbation step size; all elements complete random phase perturbation to obtain a new solution t+1 .

[0053] Further, the calculation formula of the system communication efficiency is as follows:

[0054]

[0055] In the formula, R t represents the system communication efficiency at the tth iteration, and gamma k is the received signal-to-noise ratio of the kth user.

[0056] The beneficial effects of the present application are:

[0057] (1) The RIS phase shift matrix optimization method proposed in the present application optimizes the RIS phase shift matrix in a random perturbation manner, and compared with the commonly used convex optimization method, the calculation amount and calculation overhead are reduced.

[0058] (2) The RIS phase shift matrix optimization method proposed in the present application considers the local optimal solution problem caused by using a constant step size in the random perturbation process. When the number of perturbation failures C T exceeds the specified threshold value tau, the perturbation step size delta is attenuated to R D delta, and by reducing the step size, the performance of the phase shift solution obtained after perturbation is better. BRIEF DESCRIPTION OF DRAWINGS

[0059] Figure 1 is a flowchart for obtaining the optimal RIS phase shift matrix in the present application scheme;

[0060] Figure 2 is an iteration curve diagram of the system communication efficiency optimization process. DETAILED DESCRIPTION

[0061] The present application will now be further described in detail in conjunction with the accompanying drawings.

[0062] The present application proposes a random phase perturbation method for RIS-aided unmanned aerial vehicle communication phase shift matrix optimization, Figure 1 and gives the specific implementation process of the present method.

[0063] This embodiment selects a MISO (Multiple Input Single Output) communication channel model serving 3 users, that is, K = 3. The number of transmitting antennas M = 64, the number of RIS reflecting elements N = 64, and the wireless signal transmission frequency f0= 300Mhz. The coordinates of the base station, the unmanned aerial vehicle and the 3 users are respectively: w BS =[0,0,0], w UAV =[12,12,25], w UE1 =[20,5,0], w UE2 =[20,20,0], w UE3 =[5,20,0]. The specific simulation parameters are shown in Table 1.

[0064] Table 1 Simulation parameters

[0065]

[0066]

[0067] The random phase disturbance method for RIS-aided unmanned aerial vehicle communication phase shift matrix optimization proposed in this embodiment includes the following steps:

[0068] Step 1, according to the base station position, user position, RIS position, base station antenna number and RIS reflecting element number, a MISO communication channel model serving 3 users is established, the MISO communication channel model includes the baseband signal expression received by the user, the channel model from the base station to the user, the channel model from the base station to the RIS and the channel model from the RIS to the user;

[0069] Assuming that the RIS surface is carried by the unmanned aerial vehicle, the baseband signal received by the kth (k = 1, 2, 3) user can be expressed as:

[0070] y k =(h RIS-k ΦH BS-RIS +h BS-k )g k x k +n k (1)

[0071] In the formula, represents the RIS to the kth user's Rice fading channel model, represents the base station to the kth user's Rayleigh fading channel model, represents the base station to the RIS's Rice fading channel model, represents the RIS phase shift matrix, θ N represents the phase shift of the Nth RIS reflecting element, x k represents the transmission baseband signal, satisfying represents the beamforming matrix of the transmitting end; n k is a complex additive white Gaussian noise with mean 0 and variance σ 2 ; N is the total number of RIS reflecting elements, and M is the number of base station antennas, represents a complex number set with dimension of N x M.

[0072] The channel model h BS-k from the base station to the kth user can be simulated by Rayleigh fading:

[0073]

[0074] where ρ is the path loss at the reference distance D0=1m, and a is the corresponding path loss exponent, represents the random scattering component, which is subject to a complex Gaussian distribution with mean 0 and variance 1.

[0075] Generate 3 sets of random numbers, each containing M random numbers, subject to a complex Gaussian distribution with mean 0 and variance 1, to simulate the corresponding random scattering component

[0076] According to formula (2) and the simulation parameters in Table 1, the channel matrix h BS-1 , h BS-2 , h BS-3 from the base station (BS) to the user are respectively:

[0077]

[0078]

[0079]

[0080] The Rician fading channel model from the base station to the RIS is:

[0081]

[0082] where K1 represents the K factor in the Rician fading channel, represents the line-of-sight (LoS) component, represents the non-line-of-sight (non-LoS, NLoS) component, and its elements are independent and identically distributed random variables, subject to a complex Gaussian distribution with mean 0 and variance 1.

[0083] The channel h RIS-1 , h RIS-2 , h RIS-3 from the RIS to the 3 users can be represented by Rician fading, given by formula (4):

[0084]

[0085] where K2 represents the K-factor in the RIS channel, represents the LoS component, represents the NLoS component, whose elements are independent and identically distributed random variables, following a complex Gaussian distribution with mean 0 and variance 1.

[0086] The N-element uniform linear array (ULA) employed at the base station, whose array response is characterized by:

[0087]

[0088] where d represents the antenna element spacing, λ represents the wavelength of the transmitted electromagnetic wave, and θ represents the beam angle.

[0089] The LoS component of the RIS-to-BS RIS channel is obtained according to the array response of the N-element ULA

[0090]

[0091] where, represents the beam transmit angle at the BS, represents the angle of arrival of the BS transmit beam at the RIS;

[0092] The LoS component of the RIS-to-BS RIS channel is obtained according to the array response of the N-element ULA

[0093]

[0094] where, represents the beam transmit angle at the BS,

[0095] The LoS component in the RIS channel is calculated according to formula (5), formula (6), and the simulation parameters in Table 1 where, and determined by the input BS and RIS position information, and its value is:

[0096]

[0097] A set of N random numbers is generated, following a complex Gaussian distribution with mean 0 and variance 1, and the NLoS component is:

[0098]

[0099] The channel matrix H of the BS-to-RIS channel is calculated according to formula (3) BS-RISis expressed as:

[0100]

[0101] The LoS component is calculated according to formula (7).

[0102] The corresponding is determined by the input position coordinate information. Then the corresponding are respectively:

[0103]

[0104]

[0105]

[0106] Generate 3 groups of N random numbers, subject to complex Gaussian distribution with mean 0 and variance 1, to simulate N LoS components, are respectively:

[0107]

[0108]

[0109]

[0110] According to formula (4), the channel h RIS-1 , h RIS-2 , h RIS-3 are respectively:

[0111] h RIS-1 = [0.0123-0.0016i,-0.0066+0.0061i,...,-0.0038-0.0093i] T

[0112] h RIS-2 = [0.0076+0.0004i,-0.011+0.0073i,...,0.0077-0.0021i] T

[0113] h RIS-3 = [0.096+0.0034i,-0.0043+0.0033i,...,-0.0020-0.0069i] T

[0114] Step 2, optimize the RIS phase shift matrix in the baseband signal expression received by the user, randomly perturb the solution of the RIS phase shift matrix, and continuously iterate and optimize to obtain the optimal solution of the RIS phase shift matrix. Specifically:

[0115] Step 2.1, generate N random numbers on [0, 2π) to form an initial solution vector θ0:

[0116] θ0=[1.8281,3.7461,0.5574,2.9744,1.2137,...,5.4636]

[0117] as the initial RIS phase shift matrix.

[0118] According to the maximum ratio transmission (MRT) principle, when transmitting with power p k , the transmit beamforming matrix g k is expressed as:

[0119]

[0120] According to the above formula, the transmit beamforming matrix corresponding to each user is calculated, g1, g2, g3 are respectively:

[0121] g1=[-0.1544+0.1264i,0.6266+0.0365i,...,-0.1800+0.1040i]

[0122] g2=[0.1636+0.0497i,0.0811+0.1126i,...,0.4228+0.3552i]

[0123] g3=[-0.1687+0.2540i,-0.1669-0.1749i,...,-0.1658+0.1780i]

[0124] According to the calculation formula of system communication efficiency:

[0125]

[0126] The initial system communication efficiency R0 is calculated to be 1.9895.

[0127] In the formula, γ k is the received signal-to-noise ratio of the kth user, and the calculation formula is as follows:

[0128]

[0129] In the formula, g i ​denotes the transmit beamforming matrix corresponding to the i-th user, σ 2 is the variance.

[0130] Step 2.2, let the initial value of the iteration number t = 0, and the initial value of the disturbance failure number C T = 0;

[0131] Step 2.3, judge whether the disturbance failure number C T is greater than or equal to the threshold value τ; if C T ≤ τ, disturb the elements in the solution θ t of the RIS phase shift matrix with a step size of δ to obtain a new solution θ t+1 ; Specifically, generate an integer 1 with a probability of 0.5, and generate an integer -1 with a probability of 0.5, and denote the generated integer as d; disturb each element θ i in the solution θ t of the RIS phase shift matrix according to the following formula:

[0132] θ i,t+1 = θ i,t + dδ (33)

[0133] If C T > τ, disturb the elements in the solution θ t of the RIS phase shift matrix with a step size of R D δ to obtain a new solution θ t+1 ; Specifically, generate an integer 1 with a probability of 0.5, and generate an integer -1 with a probability of 0.5, and denote the generated integer as d; disturb each element θ i in the solution θ t of the RIS phase shift matrix according to the following formula:

[0134] θ i,t+1 = θ i,t + dR D δ (34)

[0135] where R D ∈ (0, 1) represents the attenuation coefficient of the disturbance step size.

[0136] Step 2.4, calculate the system communication efficiency R t+1 corresponding to the new solution θ t+1 ; if R t+1 > R t , the disturbance is successful, and the corresponding θ t+1 is retained; t = t + 1; if R t+1 ≤ R t , the disturbance fails, and θ t+1 = θ t , R t+1 = Rt , C T =C T +1,t=t+1;

[0137] Step 2.5, judge whether t reaches the maximum iteration number, if yes, the iteration is terminated, if not, return to step 2.3.

[0138] For the convenience of illustration, this part only gives the process of the first iteration as an illustration. The perturbation method is as follows:

[0139] 1) judge whether the number of perturbation failures C T is greater than or equal to the threshold value τ; if C T ≤ τ, perturb the elements in θ0with δ as the step size. (If C T > τ, the corresponding perturbation step size is R D δ);

[0140] 2) randomly generate an integer 1 with a probability of 0.5, and randomly generate an integer -1 with a probability of 0.5, and denote the generated integer as d, d = 1;

[0141] 3) randomly perturb the elements θ 1,0 in θ0, and the expression is as follows:

[0142] θ 1,1 = θ 1,0 +dδ = 2.2051

[0143] Until each element in θ0is perturbed according to the step 2), a new solution θ1

[0144] θ1 = [2.2051, 4.1231, 0.9344, 2.5974,..., 5.8406] T

[0145] According to the perturbed θ1, the system communication rate R1 = 2.0352 at this time is calculated.

[0146] Judge the size relationship between R1 and R0, if R1 > R0, the perturbation is successful, t = t + 1, and θ1 is retained; if R1 ≤ R0, the perturbation fails, θ1 = θ0, R1 = R0, C T = C T +1; t = t + 1, until the iteration number is reached, and the iteration termination condition is met.

[0147] Step 3, calculate the transmit end active beamforming matrix according to the optimal RIS phase shift matrix solution, and realize the communication between the unmanned aerial vehicle and the ground user.

[0148] The effect obtained in this example can be achieved by Figure 2The specific data obtained in the simulation experiment are further described. In order to more accurately describe the effectiveness of the method in optimizing the RIS phase shift matrix, considering the randomness generated by the method, the system communication rate value of each iteration will be given by the average value of 1500 random trials, and the channel conditions of each experiment are the same. In addition, the present example gives the comparison between the present method and the random phase perturbation method with constant step size, and the effectiveness of the "step size reduction" method is illustrated by comparing the two different iteration curves. Figure 2 The comparison between the two curves shows that, compared with the phase perturbation method with constant step size, the "step size reduction" method can make the iteration curve jump up, which shows that after the perturbation failure reaches the threshold, reducing the perturbation step size can make the random perturbation more accurately approach the optimal solution. It should be particularly noted that the number of iterations in the present method can be increased according to the optimization requirements in engineering practice, but it should not be set too low, otherwise the output result will be poor.

[0149] The above is only the preferred embodiment of the present application, and the protection scope of the present application is not limited to the above-mentioned embodiments. Any technical solution falling within the scope of the present application is within the protection scope of the present application. It should be noted that, for ordinary skilled persons in the art, some improvements and refinements without departing from the principles of the present application should be considered as the protection scope of the present application.

Claims

1. A random phase perturbation method for optimizing the phase shift matrix of RIS-assisted UAV communication, characterized in that, Includes the following steps: Step 1: Establish a MISO communication channel model serving K users based on the base station location, user location, RIS location, number of base station antennas, and number of RIS reflective elements. The MISO communication channel model includes the expression of the baseband signal received by the user, the channel model from the base station to the user, the channel model from the base station to the RIS, and the channel model from the RIS to the user. The MISO communication channel model is specifically as follows: If the RIS surface is carried by a drone, then the baseband signal y received by the k-th user is... k Represented as: y k =(h RIS-k ΦH BS-RIS +h BS-k )g k x k +n k (1) In the formula, This represents the Ricean fading channel model from RIS to the k-th user. This represents the Rayleigh fading channel model from the base station to the k-th user. This represents the Ricean fading channel model from the base station to the RIS. Represents the RIS phase shift matrix, θ N Let x represent the phase shift of the Nth RIS reflector. k Represents the baseband signal transmission, satisfying Ε{x k } = 1; Represents the beamforming matrix at the transmitter; n k The mean is 0 and the variance is σ. 2 The additive white Gaussian noise; N is the total number of RIS reflective elements, and M is the number of base station antennas. Represents the set of complex numbers with N×M dimensions; The Rayleigh fading channel model from the base station to the k-th user is specifically as follows: In the formula, ρ is the path loss at a reference distance D0 = 1m, and α is the corresponding path loss exponent. It represents the random scattering component, which follows a complex Gaussian distribution with a mean of 0 and a variance of 1; The Ricean fading channel model from the base station to the RIS is as follows: Where K1 represents the K-factor in the Ricean fading channel, Represents the component of eye-tracking transmission. This represents the non-line-of-sight transmission component, whose elements are independent and identically distributed random variables that follow a complex Gaussian distribution with a mean of 0 and a variance of 1. The Ricean fading channel model from the RIS to the k-th user is specifically as follows: Where K2 represents the K-factor in Ricean fading channels, Represents the component of eye-tracking transmission. This represents the non-line-of-sight transmission component, whose elements are independent and identically distributed random variables that follow a complex Gaussian distribution with a mean of 0 and a variance of 1. The N-element uniform linear array used at the base station has an array response characterized by the following equation: Where d represents the spacing between antenna elements, λ represents the wavelength of the transmitted electromagnetic wave, and θ represents the beam angle; The line-of-sight transmission component of the Ricean fading channel from the base station to the RIS is obtained based on the array response of the N-ary uniform linear array. in, Represents the beam emission angle at BS. This represents the angle of arrival of the BS transmitted beam at the RIS. The line-of-sight transmission component of the Ricean fading channel from the RIS to the k-th user is obtained based on the array response of the N-ary uniform linear array. in, This represents the beam emission angle of the RIS for the k-th user; Maximum ratio transmission is used to increase signal power; when power p k During transmission, the beamforming matrix g at the transmitting end k Represented as: The signal-to-noise ratio γ received by the kth user k for: In the formula, g i Let σ represent the beamforming matrix of the transmitter corresponding to the i-th user. 2 For variance; Step 2: Optimize the RIS phase shift matrix in the expression of the baseband signal received by the user. Randomly perturb the solution of the RIS phase shift matrix and iteratively optimize to obtain the optimal solution of the RIS phase shift matrix. Step 2 specifically includes: Step 2.1: Generate N random numbers θ0 = [θ0, 2π) that follow a uniform distribution on the interval [0, 2π). 01 ,θ 02 ,...,θ 0N As the initial solution for the RIS phase shift matrix, As the initial RIS phase shift matrix, the initial system communication efficiency R0 is calculated; Step 2.2: Set the initial value of the iteration count to t = 0, and the initial value of the number of perturbation failures to C. T =0; Step 2.3: Determine the number of perturbation failures (C). T Is it greater than or equal to the threshold value τ? If C T ≤τ, with a step size of δ, the solution θ of the RIS phase shift matrix t By randomly perturbing the elements in the solution, a new solution θ is obtained. t+1 If C T >τ, with R D δ is the step size, and θ is the solution to the RIS phase shift matrix. t By randomly perturbing the elements in the solution, a new solution θ is obtained. t+1 ;where R D ∈(0,1), representing the decay coefficient of the perturbation step size; Step 2.4: Calculate the new solution θ t+1 The corresponding system communication efficiency R t+1 If R t+1 >R t If the perturbation is successful, the corresponding θ is retained. t+1 ; t = t + 1; If R t+1 ≤R t If the perturbation fails, let θ t+1 =θ t R t+1 =R t C T =C T +1, t=t+1; Step 2.5: Determine if t has reached the maximum number of iterations. If yes, the iteration terminates; otherwise, return to step 2.

3. Step 3: Calculate the active beamforming matrix of the transmitter based on the solution of the optimal RIS phase shift matrix to realize communication between the UAV and the ground user.

2. The random phase perturbation method for optimizing the phase shift matrix of RIS-assisted UAV communication as described in claim 1, characterized in that, In step 2.3, the solution θ of the RIS phase shift matrix is ​​obtained with a step size of δ. t The elements in the data undergo random phase perturbation as follows: An integer 1 is randomly generated with a probability of 0.5, and an integer -1 is randomly generated with a probability of 0.

5. Let the generated integers be denoted as d. The solution θ of the RIS phase shift matrix is ​​obtained according to the following formula. t Each element θ in i Perform random phase perturbation: i i,t+1 =θ i,t +dδ (10) All elements undergo random phase perturbation to obtain a new solution θ. t+1 .

3. The random phase perturbation method for optimizing the phase shift matrix of RIS-assisted UAV communication as described in claim 1, characterized in that, In step 2.3, the step of using R D δ is the step size, and θ is the solution to the RIS phase shift matrix. t The elements in the data undergo random phase perturbation as follows: An integer 1 is randomly generated with a probability of 0.5, and an integer -1 is randomly generated with a probability of 0.

5. Let the generated integers be denoted as d. The solution θ of the RIS phase shift matrix is ​​obtained according to the following formula. t Each element θ in i Perform random phase perturbation: i i,t+1 =θ i,t +dR D d (11) Where R D ∈(0,1), representing the decay coefficient of the perturbation step size; all elements undergo random phase perturbation to obtain a new solution θ. t+1 .

4. The random phase perturbation method for optimizing the phase shift matrix of RIS-assisted UAV communication as described in claim 1, characterized in that, The formula for calculating the system communication efficiency is as follows: In the formula, R t γ represents the system communication efficiency at the t-th iteration. k The signal-to-noise ratio received by the k-th user.

Images