Method for cooperative concealment transmission of unmanned aerial vehicle and smart reflector under variable detection threshold

Abstract

The application provides a UAV and intelligent reflector collaborative concealment transmission method with variable detection threshold, and the method comprises the following steps: establishing a system model; initializing system parameters, entering an iteration cycle, and updating data in each time slot; calculating the channel state of each time slot; calculating the achievable rate of the ground user; calculating the received signal of the eavesdropper, calculating the optimal detection threshold, and counting the number of false alarms and missed detections under the optimal detection threshold; optimizing the phase matrix of the intelligent reflector to satisfy the maximum achievable rate of the ground user under the false detection threshold; according to the model predictive control framework, the UAV motion and communication performance in the future time slots are predicted, the acceleration, speed and UAV position coordinates are updated; the transmission power of the UAV in each time slot is updated to satisfy the maximum average achievable rate of the ground user under the false detection threshold. The application improves the achievable rate of the legitimate user under the false detection probability constraint.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology, and in particular relates to a method for covert transmission of drones and intelligent reflective surfaces in coordination under variable detection thresholds. Background Technology

[0002] Drone communication and smart reflector-assisted communication expand the air-to-ground communication scenarios of 6G mobile communication. Communication using drones and smart reflectors increases the probability of air-to-ground line-of-sight communication, thereby improving the effectiveness of air-to-ground communication. However, the increased probability of line-of-sight communication also increases the risk of eavesdropping by malicious eavesdroppers. Therefore, it is necessary to improve the stealth of air-to-ground communication in scenarios where drones and smart reflectors cooperate in transmission. Assuming the detection threshold of eavesdroppers is fixed, the stealth of air-to-ground communication can be effectively improved through joint optimization of the drone's flight path and the passive beam of the smart reflector. However, considering that eavesdroppers are becoming more intelligent and their detection threshold can change dynamically according to the air-to-ground link, using a fixed drone flight path and the phase matrix of the smart reflector cannot counteract this in real time, thus failing to guarantee stealth communication performance at every moment.

[0003] Therefore, when the detection threshold is variable, it is necessary to dynamically adjust the collaborative transmission strategy of the drone and the smart reflective surface in order to achieve robust covert communication performance. Summary of the Invention

[0004] The purpose of this invention is to propose a method for covert transmission using a drone and an intelligent reflective surface with a variable detection threshold, which addresses situations where the eavesdropper can dynamically adjust its detection threshold.

[0005] To achieve the objective of this invention, this invention provides a method for covert transmission using a drone and a smart reflective surface under a variable detection threshold. The method includes the following steps:

[0006] Step 1: Describe and establish a system model of the intelligent reflective surface-assisted UAV covert transmission system;

[0007] Step 2: Initialize system parameters, enter the iterative loop, and update data in each time slot;

[0008] Step 3: Calculate the channel state for each time slot, including large-scale path loss and small-scale fading that follows a Rayleigh distribution;

[0009] Step 4: Calculate the reachability rate for ground users ;

[0010] Step 5: Calculate the eavesdropper's received signal Calculate the optimal detection threshold and count the number of false alarms and missed detections under the optimal detection threshold;

[0011] Step 6: Optimize the phase matrix of the smart reflector To maximize the reachability of ground users under the error detection threshold;

[0012] Step 7: Based on the Model Predictive Control (MPC) framework, predict the future... The drone's motion and communication performance in each time slot is analyzed, and the action that maximizes the cumulative discounted concealment rate is selected from a set of candidate accelerations to update the acceleration. ,speed and drone location coordinates ;

[0013] Step 8: Update the drone's transmit power for each time slot. To maximize the average reachability of ground users under the error detection threshold;

[0014] Step 9: Once the maximum number of iterations is reached, the loop iteration stops.

[0015] The significant advancement of this invention compared to existing technologies lies in:

[0016] This invention considers the impact of changes in air-to-ground communication links on the probability of correct detection of eavesdroppers. Under the condition of dynamic changes in the eavesdropper detection threshold, it jointly optimizes the phase matrix of UAV transmission power, trajectory and intelligent reflector, thereby improving the reachability rate of legitimate users under the constraint of false detection probability.

[0017] This invention considers the continuous-time movement model of UAVs, uses differential game theory for modeling, and optimizes the UAV's trajectory through model predictive control methods, thereby improving its ability to dynamically counter intelligent eavesdroppers.

[0018] To more clearly illustrate the functional characteristics and structural parameters of the present invention, further explanation is provided below in conjunction with the accompanying drawings and specific embodiments. Attached Figure Description

[0019] Figure 1 This is a model diagram of the UAV and intelligent reflective surface collaborative covert transmission system provided in this application;

[0020] Figure 2 This is a diagram of the optimized flight path and uniform straight-line flight trajectory of the UAV provided in this application;

[0021] Figure 3 This is a graph showing the change in air-to-ground line-of-sight transmission probability over time, provided in this application.

[0022] Figure 4 This is a comparison chart of the average cumulative transmit power under optimized UAV trajectory and uniform straight flight provided in this application;

[0023] Figure 5 This is a comparison chart of the average stealth rate under optimized UAV trajectory and uniform straight flight provided in this application. Detailed Implementation

[0024] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] This application employs differential game theory and cross-iteration theory to jointly optimize the UAV's trajectory, transmission power, and intelligent reflector phase matrix, thereby combating passive eavesdropping by eavesdroppers and improving the covert communication performance of air-to-ground communication networks.

[0026] The system block diagram of the present invention is as follows: Figure 1 As shown, a method for covert transmission using a drone and a smart reflective surface in collaboration under a variable detection threshold is proposed, including the following steps:

[0027] Step 1: Describe and establish a system model of the intelligent reflective surface-assisted UAV covert transmission system;

[0028] Step 2: Initialize system parameters, enter the iterative loop, and update data in each time slot;

[0029] Step 3: Calculate the channel state for each time slot, including large-scale path loss and small-scale fading that follows a Rayleigh distribution;

[0030] Step 4: Calculate the reachability rate for ground users ;

[0031] Step 5: Calculate the eavesdropper's received signal Calculate the optimal detection threshold and count the number of false alarms and missed detections under the optimal detection threshold;

[0032] Step 6: Optimize the phase matrix of the smart reflector To maximize the reachability of ground users under the error detection threshold;

[0033] Step 7: Based on the Model Predictive Control (MPC) framework, predict the future... The drone's motion and communication performance in each time slot is analyzed, and the action that maximizes the cumulative discounted concealment rate is selected from a set of candidate accelerations to update the acceleration. ,speed and drone location coordinates ;

[0034] Step 8: Update the drone's transmit power for each time slot. To maximize the average reachability of ground users under the error detection threshold;

[0035] Step 9: Once the maximum number of iterations is reached, the loop iteration stops.

[0036] Furthermore, the description and system model of the intelligent reflective surface-assisted UAV covert transmission system described in step 1 include:

[0037] In the IRS-assisted covert communication model, the system includes a UAV, a receiver Bob, a listener Willie, and a smart reflector IRS; assuming the transmit power is... The drone transmitter, receiver, and eavesdropper are all equipped with single antennas. The channels from the drone to the receiver and eavesdropper include both direct links and links reflected by the IRS. The drone communicates covertly with the receiver Bob, while the eavesdropper Willie attempts to detect the data transmission between them. The intelligent eavesdropper can dynamically adjust the detection threshold to maximize the detection probability. Assume the diagonal reflection matrix of the IRS is... Assuming the amplitude of the reflection coefficient is 1, and the noise at the receiver and listener has a mean of 0 and variances of [missing information], respectively. , Gaussian distribution;

[0038] Assuming the drone's altitude is fixed at At any moment The two-dimensional position coordinates, velocity, and acceleration are expressed as follows: , and Unlike traditional discrete-time models, this application considers a continuous-time movement model of the UAV, and the movement of the UAV is represented by the following differential equation:

[0039] (1);

[0040] (2);

[0041] And assume the drone's initial position and velocity are:

[0042] (3);

[0043] (4);

[0044] Assuming both the legitimate user and the eavesdropper are located on the ground, their two-dimensional coordinates are represented as follows: and The height of the intelligent reflective surface is fixed at Assuming the distance between the drone and the intelligent reflector is much greater than the distance between the reflector units of the intelligent reflector, therefore the distance from the drone to each reflector unit is approximately equal, then the two-dimensional coordinates of the intelligent reflector are assumed to be... .

[0045] Furthermore, the specific steps for initializing system parameters and entering the iterative loop in step 2, and updating data in each time slot, are as follows:

[0046] Initialize system parameters and set the number of loop iterations to [value]. initial value During iteration The maximum number of loop iterations is Set the drone's initialization status, including its initial position. Initial velocity and initial acceleration and the drone's launch power Minimum transmit power and maximum transmit power and the altitude of the drone Set the two-dimensional coordinates of the ground user and the eavesdropper as follows: and Set the parameters of the smart reflective surface, including its two-dimensional coordinates. and height Number of reflective units and phase matrix ,in Initialize to random phase; set the false detection probability threshold for the eavesdropper to [value missing]. .

[0047] Furthermore, step 3, which involves calculating the channel state for each time slot, including large-scale path loss and small-scale fading that follows a Rayleigh distribution, is described in the following steps:

[0048] Step 3a, assuming there is a direct link between the drone and the ground user and the eavesdropper, the distances between the drone and the ground user and the eavesdropper are respectively:

[0049] (5);

[0050] (6);

[0051] There exists a reflection link from the drone to the ground user and the eavesdropper, reflected by the smart reflector. The distances from the drone to the smart reflector, from the smart reflector to the ground user, and from the smart reflector to the eavesdropper are respectively:

[0052] (7);

[0053] (8);

[0054] (9);

[0055] Step 3b, simultaneously assuming that both air-to-ground links have a certain probability of having line-of-sight links and non-line-of-sight links, and that the probability of line-of-sight transmission... The following relationship must be satisfied:

[0056] (10);

[0057] Among them Both are constants, especially in dense urban environments. , These represent ground users, eavesdroppers, and intelligent reflectors, respectively. The pitch angle of the drone relative to the legitimate user, the eavesdropper, and the intelligent reflector. ;

[0058] Therefore, the path loss and channel gain between the nodes of the air-to-ground link are:

[0059] (11);

[0060] (12);

[0061] (13);

[0062] in and These represent the path loss for line-of-sight and non-line-of-sight transmission, respectively. Let these represent the path attenuation factors for line-of-sight and non-line-of-sight links, respectively. , , Let these represent the line-of-sight and non-line-of-sight channel gains at a reference distance of 1 meter, respectively. , . This represents the small-scale fading channel gain between the drone and ground users and eavesdroppers. This represents the small-scale fading channel gain vector from the UAV to each smart reflection unit; This represents the small-scale fading channel gain vector from each smart reflection unit to the ground user and the eavesdropper; This represents the path loss and channel gain between the drone and ground users and eavesdroppers. This represents the path loss and channel gain from the UAV to each intelligent reflection unit. This represents the path loss and channel gain from each smart reflector unit to the ground user and the eavesdropper.

[0063] Furthermore, step 4 involves calculating the reachability rate of ground users. It is determined by the following method:

[0064] Step 4a, the signal received by the ground user from the drone is:

[0065] (14);

[0066] in Indicates that drones are in Transmission power at any given time Indicates that drones are in The normalized signal emitted at all times, This represents additive white Gaussian noise at the ground user location;

[0067] Step 4b, calculate the reachable rate for ground users:

[0068] (15);

[0069] in This indicates the noise power of ground users.

[0070] Furthermore, step 5 involves calculating the eavesdropper's received signal. The optimal detection threshold is calculated, and the number of false alarms and missed detections occurring under the optimal detection threshold is counted. The specific implementation is as follows:

[0071] Step 5a, the signal received by the eavesdropper from the drone is:

[0072] (16);

[0073] in This represents additive white Gaussian noise at the location of the eavesdropper.

[0074] Step 5b, assuming the drone is transmitting a signal, calculate the eavesdropper's received power as follows:

[0075] (17);

[0076] When the drone is not transmitting a signal, the signal received by the eavesdropper is its noise power, i.e. ;

[0077] Step 5c, assuming the eavesdropper receives... The signal power of each sample is then used to calculate the average power received:

[0078] (18);

[0079] in Indicates the first time the eavesdropper received The signal of each sample;

[0080] In step 5d, according to the Neyman-Pearson criterion, likelihood detection is considered the optimal rule for minimizing the probability of false detection. At this point, the optimal detection threshold for the eavesdropper is calculated as follows:

[0081] (19);

[0082] Step 5e: Based on obtaining the optimal detection threshold, calculate the probability of missed detection of the eavesdropper as follows:

[0083] (20);

[0084] Among them, the probability of missed detection It is the probability that a legitimate user actually transmitted information, but the eavesdropper determined that no information was transmitted. Describing the degrees of freedom as The chi-square distribution function, This indicates the detection threshold for eavesdroppers. ;

[0085] Step 5f: Based on obtaining the optimal detection threshold, calculate the false alarm probability of the eavesdropper as follows:

[0086] (twenty one);

[0087] Among them, false alarm probability The probability that a legitimate user did not transmit any information, but an eavesdropper mistakenly believed that information was being transmitted. ;

[0088] Step 5g, calculate the false detection probability of the eavesdropper:

[0089] (twenty two);

[0090] To characterize the stealth performance of ground users, the false detection probability of eavesdroppers is required. ,in .

[0091] Furthermore, the phase matrix of the optimized smart reflective surface described in step 6... To maximize the reachability of ground users under the error detection threshold, the specific implementation is as follows:

[0092] Step 6a, calculate the first intelligent reflective surface Phase offset of each reflecting unit:

[0093] (twenty three);

[0094] Step 6b: Calculate the optimal phase for each reflective element of the smart reflective surface:

[0095] (twenty four);

[0096] in The function guarantees that the phase of each reflection unit is within the range of the reflection unit. Within the range.

[0097] Furthermore, step 7 describes predicting the future. The motion and communication performance of the UAV in each time slot is analyzed. The action that maximizes the cumulative discounted concealment rate is selected from a set of candidate accelerations, thereby updating the acceleration, velocity, and UAV position coordinates. The specific implementation is as follows:

[0098] Step 7a, set the prediction time slot length Sixteen candidate directions for acceleration are set and distributed at equal intervals. The magnitude of the acceleration is Therefore, the candidate acceleration is Initialize optimal acceleration And the best reward is ;

[0099] Step 7b, set the number of loops. Enter the loop iteration:

[0100] Step 7c, in each candidate Set initial rewards Then, the concealment rate for each time slot is calculated separately. and cumulative rewards :

[0101] Step 7d: Using Euler's formula, update the speed and position of the UAV in each time slot:

[0102] (25);

[0103] (26);

[0104] Step 7e: Based on the predicted drone coordinates, calculate the distances from the drone to the ground user and the eavesdropper as follows:

[0105] (27);

[0106] (28);

[0107] Simultaneously, there exists a reflection link from the drone, reflected by the smart reflector, to the ground user and the eavesdropper, and the distance from the drone to the smart reflector, and from the smart reflector to the ground user and the eavesdropper, is...

[0108] (29)

[0109] Step 7f: Predict the line-of-sight transmission probability and path loss / channel gain between air-to-ground links, where the air-to-ground line-of-sight transmission probability... Satisfying the following formula

[0110] (30)

[0111] Among them Both are constants, especially in dense urban environments. , These represent the ground user, the eavesdropper, and the intelligent reflector, respectively. .

[0112] Therefore, the path loss and channel gain between air-to-ground links are

[0113] (31)

[0114] (32)

[0115] in and These represent the path loss for line-of-sight and non-line-of-sight transmission, respectively. Let these represent the path attenuation factors for line-of-sight and non-line-of-sight links, respectively. , , Let these represent the line-of-sight and non-line-of-sight channel gains at a reference distance of 1 meter, respectively. , , This represents the small-scale fading channel gain between the drone and ground users and eavesdroppers. These represent the small-scale fading channel gain vector from the UAV to each intelligent reflection unit;

[0116] Step 7g, predict the user's reachability rate:

[0117] (33);

[0118] in Indicates the noise power of ground users;

[0119] Step 7h: Use the Sigmoid function to predict the probability of false detection of the eavesdropper.

[0120] (34);

[0121] in ;

[0122] Step 7i, predict the user's concealment rate, when hour, ,otherwise ;

[0123] Step 7j, calculate the cumulative reward with discount factor:

[0124] (35);

[0125] in Discount factor;

[0126] Step 7k, update the drone's acceleration, when ,

[0127] (36);

[0128] (37);

[0129] in It is the cumulative reward value. That is the optimal reward value. It is the candidate acceleration. It is the optimal acceleration;

[0130] Step 71: Update the drone's speed and position based on the optimal acceleration.

[0131] (38);

[0132] (39).

[0133] Furthermore, step 8 involves updating the UAV's transmit power for each time slot. To maximize the average reachability of ground users under the error detection threshold, the specific implementation is as follows:

[0134] Step 8a, set accuracy requirements and number of iterations ;

[0135] Step 8b, enter the loop iteration, set ,in and These are the initial minimum and maximum transmit power values. and It is the first The minimum and maximum transmit power in each iteration;

[0136] Step 8c, set the temporary drone transmit power, and Calculate the achievable rate and the probability of false detection of the eavesdropper:

[0137] (40);

[0138] (41);

[0139] in , ;

[0140] Step 8d, update the drone's transmit power, when and ,

[0141] (42);

[0142] (43);

[0143] otherwise

[0144] (44);

[0145] in It is the first The optimal transmit power for the next iteration;

[0146] Execute the loop iteratively until the maximum number of iterations is reached or The loop ends.

[0147] Figures 2-5 This is the simulation curve of the present invention. Figure 2 This is a diagram showing the optimized flight path and uniform linear flight trajectory of a drone. As can be seen from the diagram, the flight trajectory after the drone's acceleration optimization is different from the flight trajectory of uniform linear motion. Due to the updated acceleration, the flight distance is greater, and the distance from the eavesdropper is also greater, making it less likely for the eavesdropper to detect its transmission behavior. Figure 3 It is a curve showing the change in the probability of air-to-ground line-of-sight transmission over time slots. As time slots change, the probability of UAV line-of-sight transmission increases, which can improve the user's reachability while ensuring concealment constraints. Figure 4 and Figure 5 The simulation results are The average of the independent experiments, of which Figure 4This is a comparison chart of the average cumulative transmit power under optimized drone flight path and uniform straight-line flight. Figure 5 This is a comparison chart of the average stealth rate under UAV trajectory optimization and uniform straight flight. After UAV trajectory optimization, the cumulative transmission power is reduced, while the stealth rate is improved. The algorithm proposed in this invention is superior to the uniform straight flight strategy, thus verifying that the stealth transmission performance of UAVs can be effectively improved by jointly optimizing the UAV's transmission power and trajectory.

[0148] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for covert transmission using a drone and a smart reflective surface in coordination under a variable detection threshold, characterized in that, The method includes the following steps: Step 1: Describe and establish a system model of the intelligent reflective surface-assisted UAV covert transmission system; Step 2: Initialize system parameters, enter the iterative loop, and update data in each time slot; Step 3: Calculate the channel state for each time slot, including large-scale path loss and small-scale fading that follows a Rayleigh distribution; Step 4: Calculate the reachability rate for ground users ; Step 5: Calculate the eavesdropper's received signal Calculate the optimal detection threshold and count the number of false alarms and missed detections under the optimal detection threshold; Step 6: Optimize the phase matrix of the smart reflector To maximize the reachability of ground users under the error detection threshold; Step 7: Based on the Model Predictive Control (MPC) framework, predict the future... The drone's motion and communication performance in each time slot is analyzed, and the action that maximizes the cumulative discounted concealment rate is selected from a set of candidate accelerations to update the acceleration. ,speed and drone location coordinates ; Step 8: Update the drone's transmit power for each time slot. To maximize the average reachability of ground users under the error detection threshold; Step 9: Once the maximum number of iterations is reached, the loop iteration stops.

2. The method according to claim 1, characterized in that, Step 1: Describe and establish a system model for the intelligent reflective surface-assisted UAV covert transmission system, including: In the IRS-assisted covert communication model, the system includes a UAV, a receiver Bob, a listener Willie, and a smart reflector IRS; assuming the transmit power is... The drone transmitter, receiver, and eavesdropper are all equipped with single antennas. The channels from the drone to the receiver and eavesdropper include both direct links and links reflected by the IRS. The drone communicates covertly with the receiver Bob, while the eavesdropper Willie attempts to detect the data transmission between them. The intelligent eavesdropper can dynamically adjust the detection threshold to maximize the detection probability. Assume the diagonal reflection matrix of the IRS is... Assuming the amplitude of the reflection coefficient is 1, and the noise at the receiver and listener has a mean of 0 and variances of 1 and 2, respectively, the following conditions apply. , Gaussian distribution; Assuming the drone's altitude is fixed at At any moment The two-dimensional position coordinates, velocity, and acceleration are expressed as follows: , and Unlike traditional discrete-time models, this application considers a continuous-time movement model of the UAV, and the movement of the UAV is represented by the following differential equation: （1）； （2）； And assume the drone's initial position and velocity are: （3）； （4）； Assuming both the legitimate user and the eavesdropper are located on the ground, their two-dimensional coordinates are represented as follows: and The height of the intelligent reflective surface is fixed at Assuming the distance between the drone and the intelligent reflector is much greater than the distance between the reflector units of the intelligent reflector, therefore the distance from the drone to each reflector unit is approximately equal, then the two-dimensional coordinates of the intelligent reflector are assumed to be... .

3. The method according to claim 2, characterized in that, Step 2: Initialize system parameters, enter the iterative loop, and update data in each time slot, including: Initialize system parameters and set the number of loop iterations to [value]. initial value During iteration The maximum number of loop iterations is Set the drone's initialization status, including its initial position. Initial velocity and initial acceleration and the drone's launch power Minimum transmit power and maximum transmit power and the altitude of the drone Set the two-dimensional coordinates of the ground user and the eavesdropper as follows: and Set the parameters of the smart reflective surface, including its two-dimensional coordinates. and height Number of reflective units and phase matrix ,in Initialize to random phase; set the false detection probability threshold for the eavesdropper to [value missing]. .

4. The method according to claim 3, characterized in that, Step 3: Calculate the channel state for each time slot, including large-scale path loss and small-scale fading following a Rayleigh distribution, including: Step 3a, assuming there is a direct link between the drone and the ground user and the eavesdropper, the distances between the drone and the ground user and the eavesdropper are respectively: （5）； （6）； There exists a reflection link from the drone to the ground user and the eavesdropper, reflected by the smart reflector. The distances from the drone to the smart reflector, from the smart reflector to the ground user, and from the smart reflector to the eavesdropper are respectively: （7）； （8）； （9）； Step 3b, simultaneously assuming that both air-to-ground links have a certain probability of having line-of-sight links and non-line-of-sight links, and that the probability of line-of-sight transmission... The following relationship must be satisfied: （10）； Among them Both are constants, especially in dense urban environments. , These represent ground users, eavesdroppers, and intelligent reflectors, respectively. The pitch angle of the drone relative to the legitimate user, the eavesdropper, and the intelligent reflector. ; Therefore, the path loss and channel gain between the nodes of the air-to-ground link are: （11）； （12）； （13）； in and These represent the path loss for line-of-sight and non-line-of-sight transmission, respectively. Let these represent the path attenuation factors for line-of-sight and non-line-of-sight links, respectively. , , Let these represent the line-of-sight and non-line-of-sight channel gains at a reference distance of 1 meter, respectively. , . This represents the small-scale fading channel gain between the drone and ground users and eavesdroppers. This represents the small-scale fading channel gain vector from the UAV to each smart reflection unit; This represents the small-scale fading channel gain vector from each smart reflection unit to the ground user and the eavesdropper; This represents the path loss and channel gain between the drone and ground users and eavesdroppers. This represents the path loss and channel gain from the UAV to each intelligent reflection unit. This represents the path loss and channel gain from each smart reflector unit to the ground user and the eavesdropper.

5. The method according to claim 4, characterized in that, Step 4: Calculate the reachability rate for ground users. It is determined by the following methods, including: Step 4a, the signal received by the ground user from the drone is: （14）； in Indicates that drones are in Transmission power at any given time Indicates that drones are in The normalized signal emitted at all times, This represents additive white Gaussian noise at the ground user location; Step 4b, calculate the reachable rate for ground users: （15）； in This indicates the noise power of ground users.

6. The method according to claim 5, characterized in that, Step 5: Calculate the eavesdropper's received signal Calculate the optimal detection threshold and count the number of false alarms and missed detections occurring at the optimal detection threshold, including: Step 5a, the signal received by the eavesdropper from the drone is: （16）； in This represents additive white Gaussian noise at the location of the eavesdropper. Step 5b, assuming the drone is transmitting a signal, calculate the eavesdropper's received power as follows: （17）； When the drone is not transmitting a signal, the signal received by the eavesdropper is its noise power, i.e. ; Step 5c, assuming the eavesdropper receives... The signal power of each sample is then used to calculate the average power received: （18）； in Indicates the first time the eavesdropper received The signal of each sample; In step 5d, according to the Neyman-Pearson criterion, likelihood detection is considered the optimal rule for minimizing the probability of false detection. At this point, the optimal detection threshold for the eavesdropper is calculated as follows: （19）； Step 5e: Based on obtaining the optimal detection threshold, calculate the probability of missed detection of the eavesdropper as follows: （20） ； Among them, the probability of missed detection It is the probability that a legitimate user actually transmitted information, but the eavesdropper determined that no information was transmitted. Describing the degrees of freedom as The chi-square distribution function, This indicates the detection threshold for eavesdroppers. ; Step 5f: Based on obtaining the optimal detection threshold, calculate the false alarm probability of the eavesdropper as follows: （21）； Among them, false alarm probability The probability that a legitimate user did not transmit any information, but an eavesdropper mistakenly believed that information was being transmitted. ; Step 5g, calculate the false detection probability of the eavesdropper: （22）； To characterize the stealth performance of ground users, the false detection probability of eavesdroppers is required. ,in .

7. The method according to claim 6, characterized in that, Step 6 describes the optimization of the phase matrix of the smart reflective surface. To maximize the achievable rate for ground users under the error detection threshold, including: Step 6a, calculate the first intelligent reflective surface Phase offset of each reflecting unit: （23）； Step 6b: Calculate the optimal phase for each reflective element of the smart reflective surface: （24）； in The function guarantees that the phase of each reflection unit is within the range of the reflection unit. Within the range.

8. The method according to claim 7, characterized in that, Step 7, predicting the future The drone's motion and communication performance in each time slot is analyzed. The action that maximizes the cumulative discounted concealment rate is selected from a set of candidate accelerations, thereby updating the acceleration, velocity, and drone position coordinates, including: Step 7a, set the prediction time slot length The candidate directions for acceleration are set to 16, and they are evenly distributed, i.e. The magnitude of the acceleration is Therefore, the candidate acceleration is Initialize optimal acceleration And the best reward is ; Step 7b, set the number of loops. Enter the loop iteration: Step 7c, in each candidate Set initial rewards Then, the concealment rate for each time slot is calculated separately. and cumulative rewards : Step 7d: Using Euler's formula, update the speed and position of the UAV in each time slot: （25）； （26）； Step 7e: Based on the predicted drone coordinates, calculate the distances from the drone to the ground user and the eavesdropper as follows: （27）； （28）； Simultaneously, there exists a reflection link from the drone, reflected by the smart reflector, to the ground user and the eavesdropper, and the distance from the drone to the smart reflector, and from the smart reflector to the ground user and the eavesdropper, is... （29） Step 7f: Predict the line-of-sight transmission probability and path loss / channel gain between air-to-ground links, where the air-to-ground line-of-sight transmission probability... Satisfying the following formula （30） Among them Both are constants, especially in dense urban environments. , These represent the ground user, the eavesdropper, and the intelligent reflector, respectively. ; Therefore, the path loss and channel gain between air-to-ground links are （31） （32） in and These represent the path loss for line-of-sight and non-line-of-sight transmission, respectively. Let these represent the path attenuation factors for line-of-sight and non-line-of-sight links, respectively. , , Let these represent the line-of-sight and non-line-of-sight channel gains at a reference distance of 1 meter, respectively. , , This represents the small-scale fading channel gain between the drone and ground users and eavesdroppers. These represent the small-scale fading channel gain vector from the UAV to each intelligent reflection unit; Step 7g, predict the user's reachability rate: （33）； in Indicates the noise power of ground users; Step 7h: Use the Sigmoid function to predict the probability of false detection of the eavesdropper. （34）； in ; Step 7i, predict the user's concealment rate, when hour, ,otherwise ; Step 7j, calculate the cumulative reward with discount factor: （35）； in Discount factor; Step 7k, update the drone's acceleration, when , （36）； （37）； in It is the cumulative reward value. That is the optimal reward value. It is the candidate acceleration. It is the optimal acceleration; Step 71: Update the drone's speed and position based on the optimal acceleration. （38）； （39）。 9. The method according to claim 8, characterized in that, Step 8: Update the drone's transmit power for each time slot. To maximize the average reachability of ground users under the error detection threshold, including: Step 8a, set accuracy requirements and number of iterations ; Step 8b, enter the loop iteration, set ,in and These are the initial minimum and maximum transmit power values. and It is the first The minimum and maximum transmit power in each iteration; Step 8c, set the temporary drone transmit power, and Calculate the achievable rate and the probability of false detection of the eavesdropper: （40）； （41）； in , ; Step 8d, update the drone's transmit power, when and , （42）； （43）； otherwise （44）； in It is the first The optimal transmit power for the next iteration; Execute the loop iteratively until the maximum number of iterations is reached or The loop ends.

Images